Main Street USD

On the weekend of June 20, 2026, a stablecoin called msUSD (Main Street USD) lost most of its value after the dashboard proving its reserves went dark. 
Its issuer, Main Street, said the token “remains fully backed” and called the outage a “reporting issue, not a solvency issue.” Accountable, the firm behind the dashboard, ended its work with the issuer and said it “did not retain an ongoing, source-level view” of the reserves.
That left holders with an issuer vouching for the reserves and a verifier saying it could not see them. Neither statement settled whether the money was there, and the market declined to wait for an answer.
Altura, a vault that peaked at $39 million, said it “never had any exposure to Mainstreet,” although it did share a verifier. Within a day, it paid out more than $8.5 million in redemptions and announced a wind-down.
That event was just another example of crypto’s reserve proofs breaking in public. An early, prominent case followed the collapse of FTX in November 2022, when other exchanges rushed to publish proof-of-reserves reports from accounting firms. Mazars, which produced them for Binance, KuCoin and Crypto.com, paused the work a month later, citing “concerns regarding the way these reports are understood by the public.  
A month later, it paused the work, citing “concerns regarding the way these reports are understood by the public.”
Even the PCAOB, the US audit regulator, told customers to “exercise extreme caution,” that these engagements “are not audits,” that they described assets “at one point in time,” and that they “likely do not address the crypto entity’s liabilities.” The reports might also miss assets an exchange had borrowed to look well funded on the day of the check.
TL;DR: an exchange could pass on a Tuesday, but that wouldn’t tell its customers anything about its status on Wednesday.
With those complaints in hand, the next generation tried to fix onchain reserve reporting. Accountable bolstered its services to be the antithesis of “self-reported claims, delayed audits, and static PDFs,” instead offering a continuous, privacy-preserving view of assets and liabilities across wallets, custodians, exchanges, and banks.
That answered the snapshot problem and most of the liabilities problem. But private accounts stayed private, so holders still relied on the attestor’s word as to which accounts counted in its analysis.
Neutrl USD supplied the harder lesson in proof of reserves’ shortcomings. 
In May, Accountable said its dashboard provided “continuous cryptographic proof that NUSD reserves matched the protocol’s liabilities.” Less than three months later, Neutrl stopped holders from minting or cashing out NUSD, with about $53.6 million of the token outstanding. A DeFi researcher then found that Accountable’s page for Neutrl was no longer working, leaving holders to rely on Neutrl’s own disclosures. Two weeks into the freeze, Neutrl reported about $27 million in assets it could readily sell, roughly half of what was outstanding, plus other positions it could not. When it finally let holders cash out in September, the price was 51 cents per token. Neutrl has not said how much of the rest holders might recover.
Nothing public shows the May proof was false, and holders were stuck anyway. Nearly four years of better proofs had brought the industry back to the question it started with: could people get their money out?
Across both generations, the proofs tended to be accurate about something narrow and were read as saying something broad. A balance existed, so the product must be safe. A badge said verified, and nobody scrutinized the verifier or its methods.
Verified by llamas
DefiLlama launched Verified by DefiLlama on Nov. 2, with Frax Finance’s frxUSD as its first record. The latest snapshot shows $99.34 million of reserves against $96.05 million of frxUSD, or about $1.03 behind every token.
You can check most of that yourself, too. Six Ethereum addresses hold about 98% of the reserves, and the register lists each one with its units, price, and token contract. Another 10.48 million frxUSD sit locked in the Fraxtal bridge, with matching supply on Fraxtal, so the numbers are counted only once.
For offchain assets, DefiLlama takes read-only access to banks, custodians, and fund administrators, then checks balances at the source against token supply. Ultimately, it can see the money, and cannot move it. For example, our dashboard shows that Frax holds  $2.01 million at Erebor; readers rely on DefiLlama’s access there. If a source misses an update, the program’s label is “Stale.” If a figure cannot be independently checked, the label is “Unverified.”
This service has been in the works for some time, and developers have taken extra care to set customer expectations about what proof DefiLlama accepts to verify reserves.
The timing of this launch lines up with some important regulation. The GENIUS Act, due to take effect by Jan. 18, 2027, will require monthly reserve reports examined by a PCAOB-registered accounting firm. Frax says it is applying for this license. Those reports carry legal weight a dashboard cannot, and they arrive twelve times a year. Verified covers the days in between, on a site where the market already checks TVL and stablecoin supply.
The old limits still apply, since the record covers only the accounts that issuers provide. The record cannot say how fast fund shares turn into dollars. DefiLlama’s own page makes clear it is “not a statutory audit or a guarantee of future solvency.”
More records are coming in the near future, and those will show whether other issuers publish as much as Frax did.
The frxUSD register is public: https://defillama.com/verified/frxusd
Issuers can apply at https://defillama.com/verified

On the weekend of June 20, 2026, a stablecoin called msUSD (Main Street USD) lost most of its value after the dashboard proving its reserves went dark. 
Its issuer, Main Street, said the token “remains fully backed” and called the outage a “reporting issue, not a solvency issue.”

On the weekend of June 20, 2026, a stablecoin called msUSD (Main Street USD) lost most of its value after the dashboard proving its reserves went dark.  Its issuer, Main Street, said the token “remains fully backed” and called the outage a “reporting issue, not a solvency issue.”

What the last two generations of reserve proofs missed

It's official: Democrats will control both chambers of Congress from January 3, 2027. 
That's a clear shift from the last two years, when Republican majorities ran both the House and the Senate and steered crypto's market structure bill through the House, though the Clarity Act failed at a mid-September vote in the Senate.
That shift could change things for crypto. 
Democrats have generally pressed for stricter consumer protection, ethics rules on officials' crypto income and tougher anti-money-laundering controls, while Republicans have generally led the push for dedicated crypto laws that give the industry clear rules, faster.
The line isn't that clean, though. 
The GENIUS Act will take effect anyway, despite the ‘blue wave’ from these midterm elections
The CLARITY Act did pass the House with votes from both parties
The SEC, the CFTC and the veto pen stay with the president until January 2029. So, the shift could be smaller than the headline suggests. Let's break it down.
Democrats now hold Congress
Democrats won both chambers after the November 3 vote, and by more than expected. 
In the Senate, they reached 54 seats after flipping seven Republican-held seats: Iowa, Kansas, Maine, North Carolina, Ohio, South Carolina and Texas. 
In the House, Democrats  won well over the 218 seats needed for a majority.
The House flip was largely priced in but maybe not at that scale. 
Kalshi traders put a Democratic House near 90% in mid-October but a Democratic Senate at 63%, while Kansas and South Carolina were not among the races seen as closest before the vote. 
In short, the US midterm elections have resulted in a so-called “blue wave”.
Blue Wave: How did markets react?
The immediate reaction in crypto was muted. 
BTC traded near $78,000 on election evening, fell to about $75,700 overnight as results came in and Nasdaq futures dropped about 2%, then recovered toward $77,000 by morning. 
Still, prices in the coming days will depend on more than the election, with the recent security incidents, the Fed's rate path, new SEC guidance on decentralization and the conflict with Iran all in play at the same time.
Democrats take over Congress, but not crypto regulators
The new Congress, which begins its term in January 2027, inherits a crypto rulebook that's half built. 
The piece already in place is GENIUS, the stablecoin law signed in July 2025, which sets the rules for issuing and distributing dollar stablecoins in the US. This one won’t be impacted and will take effect on January 18, 2027.
The biggest missing piece is CLARITY, the bill that would set the rules for crypto trading in the US, covering which regulator oversees which tokens and how exchanges get licensed. It passed the House 294 to 134 in July 2025, then failed in the Senate on September 15th, when a vote to advance it got 49 votes, 11 short of the 60 it needed.
However, the regulators have kept going without Congress: 
The SEC, which oversees securities, proposed Regulation Crypto Assets in August, a set of exemptions that would let crypto projects raise up to $75M a year without a full securities registration. 
The CFTC, which oversees futures and other derivatives, is consulting until December 14 on rules for leveraged retail crypto trading and a new type of regulated crypto exchange.
Democratic Congress can slow the crypto agenda, not repeal it
From January 2027, the committees that write crypto law will be led by some of its sharpest critics:
Elizabeth Warren is in line to chair Senate Banking
Maxine Waters could return as chair of House Financial Services, though she faces a challenge for the role
Shontel Brown is in line to lead House Agriculture.
Either Michael Bennet or Sherrod Brown (the latter just won back a seat in Ohio) could lead Senate Agriculture, which oversees the CFTC.
Warren, Waters and Shontel Brown carry an "F" rating from Stand With Crypto, and their stated objections center on conflicts of interest, consumer protection and illicit finance rather than on crypto existing at all.
New crypto legislation unlikely to pass
A comprehensive new crypto law before 2029 is now unlikely, for reasons that stack up. 
Committee chairs decide which bills get a hearing and a vote, so a bill like CLARITY, which the incoming chairs have criticized, wouldn't get out of committee. 
The bill Democrats would rather pass, with strict ethics rules covering the president's crypto income, is one the president would likely veto, and overriding a veto takes two-thirds of both chambers. With 54 seats, Democrats remain well short of the 60 votes most bills need in the Senate, so any compromise would still require Republican support, and the 2028 campaign leaves little time to find one.
Slowing down regulation
The bigger lever is the Senate's power to approve regulators. 
The SEC and the CFTC, the two agencies writing crypto rules, had seven empty leadership seats in early October, and every new commissioner needs a Senate vote. 
A Democratic Senate can't cancel a rule, but it can slow the agencies down by holding up nominees. 
Committee chairs can also call crypto executives and agency officials to testify in public hearings and force them to hand over documents, which adds scrutiny of crypto firms and of the president's own crypto businesses without changing any rule.
What can't be stopped?
The GENIUS Act
GENIUS, the US stablecoin law, takes effect on January 18, 2027 whatever the result.
Because the date has already been written into the law, it can’t be changed by the new Congress.
GENIUS requires dollar stablecoins offered in the US to be backed one-to-one by cash and other safe, liquid assets and issued by approved firms. 
It also sets conditions for foreign issuers: 
from January 18, US platforms can only offer a foreign stablecoin if its issuer can comply with US legal orders
from July 2028, only coins from approved foreign issuers can be offered. 
USDT and USDC make up about 82% of all stablecoins and only one of them is issued in the US, so how issuers adapt to keep US access will shape the dollar liquidity traders use everywhere. 
Tokenized US Stocks
Tokenized US stocks are the second thing a Democratic Congress can't stop. 
They trade under a five-year exemption the SEC granted on September 17, which lets venues offer tokenized versions of US-listed shares under set conditions. 
Because it's an SEC decision, Congress could only overturn it with a law the president would have to sign, so it runs until 2031 unless a future SEC revisits it. 
Can CLARITY still pass before January 3?
On paper, yes. Senator Thom Tillis's motion to reconsider lets the Senate bring the bill back without starting over, and urgency is highest in this scenario: once the House flips on January 3, the bill dies and Democrats write the next draft. Cynthia Lummis, its lead Senate negotiator, is serving her last weeks, and outgoing members vote with less electoral pressure.
The bar is still high. Supporters need 60 Senate votes, 11 more than in September, then identical text in both chambers and the president's signature, all within about 22 Senate session days. The issues that blocked the bill haven't moved: ethics rules on officials' crypto income, consumer protection, illicit finance and state enforcement powers.
It matters because three important things come from this law: licensing for spot exchanges and brokers, customer priority if an exchange goes bankrupt, and Bank Secrecy Act coverage for platforms. 
Put simply, the rules will keep coming from the SEC and the CFTC rather than from Congress, more slowly than before, and anything adopted before 2029 can be rewritten by the next administration. 
Key factors to watch
Beyond the initial market reaction, what matters now is a short list of events that will shape the US rulebook over the coming year.
The lame duck, to January 3. The outgoing Republican Senate has one final opportunity to advance its legislative agenda. A last-minute vote on CLARITY remains possible, though unlikely. Confirming SEC and CFTC nominees, or advancing narrower legislation such as crypto tax reforms, appears more achievable.
The new committee chairs. Early statements and hearings from Warren at Senate Banking, alongside the incoming chairs of the House and Senate Agriculture Committees, will offer the first indications of the new Congress's regulatory priorities.
GENIUS kicks in January 18, 2027. The first stablecoin provisions take effect just two weeks after the new Congress convenes. Attention will turn to the OCC's final rules and how foreign issuers adapt to preserve access to US platforms. Their decisions could have implications well beyond the US, particularly for the dollar-denominated liquidity underpinning global crypto markets.
The pace at the SEC and the CFTC. A Democratic Senate could slow both agencies by delaying key appointments. Whether the SEC finalizes Regulation Crypto Assets, and how the CFTC responds to its consultation closing December 14, will provide an early indication of whether regulatory progress can maintain its current momentum.
Durability before 2029. Rules introduced without legislative backing, including the SEC's proposed exemption for tokenized US equities, could be revisited by a future administration after January 2029. For market participants, the distinction between enacted legislation, finalized regulations, and proposals will be critical in assessing the long-term regulatory certainty of their products and business models.
Conclusion
A Democratic Congress slows the US crypto-regulation pipeline  without entirely stopping it. New laws are unlikely before 2029, the regulators keep writing rules at a pace the Senate can now influence, and GENIUS goes live on January 18 regardless. 
In the short term, crypto prices will keep reacting to factors beyond Congress: Fed rates, security incidents, SEC guidance and the conflict with Iran all move the same charts. 
For the crypto community, the bigger question isn't how prices react today, but whether the rules shaping the industry's future are built to last.

It's official: Democrats will control both chambers of Congress from January 3, 2027. That's a clear shift from the last two years, when Republican majorities ran both the House and the Senate and steered crypto's market structure bill through the House. That shift could change things for crypto.

It's official: Democrats will control both chambers of Congress from January 3, 2027. That's a clear shift from the last two years, when Republican majorities ran both the House and the Senate and steered crypto's market structure bill through the House. That shift could change things for crypto.

Blue Wave, Same Regulators: What a Democrat-led Congress Means for Crypto Regulations

Rand Hindi

Ryan Celaj

Kyle Cai

Zama

Following our report on Zama, “A Look Into Confidential DeFi,” Ryan Celaj, our head of research, and Kyle Cai, our former senior research analyst, sat down with the company’s CEO and co-founder, Rand Hindi. They discussed Zama’s unexpected foray into crypto, its competitors’ approach to encrypting blockchain data, the “AI apocalypse,” and much more. This conversation has been edited for length and clarity. You can find the full interview here.
1. A confidentiality layer for onchain finance
Ryan Celaj: For anyone new to Zama, what are you building and why does it matter?
Rand Hindi: My background is in AI. I did a PhD at 21, built one of the first AI companies, and sold it in 2019. Since 2020, I've been building Zama with my co-founder Pascal Paillier, a pioneer of Homomorphic Encryption. I've also been in crypto for over a decade.
Zama is a confidentiality layer for onchain finance. It lets you pay, trade, invest, and put money in vaults on public blockchains like Ethereum and Solana without everyone seeing what you're doing. 
A good analogy is, what HTTPS did for the internet, we're doing for onchain finance.
2. Who confidential vaults are for
Kyle Cai: You started with a Steakhouse Financial confidential USDC vault and have since expanded to 16 vaults across five curators: Steakhouse Financial, Armitage by Wintermute, Flowdesk, Rockaway, and Bitwise.
Ryan Celaj: Who are those vaults built for? Retail or institutions?
Rand Hindi: The people who need privacy most onchain are those with large positions. This isn't about hiding what you buy with your credit card. It's about making sure competitors and other funds can't see how you trade or allocate your portfolio.
Vaults are one piece of the puzzle. People put money in vaults, but they also want to trade and build strategies, and they need confidentiality for all of it. That's what we're building out now.
3. Why finance needs its HTTPS moment
Ryan Celaj: Interest in privacy has grown a lot over the last two years. Is that because institutions are coming onchain?
Rand Hindi: Pretty much. We didn't start as a blockchain company. We started as a cryptography company working on Fully Homomorphic Encryption. Exploring use cases, we concluded that if finance moves onchain, confidentiality becomes necessary. Today, only people who have no choice or don't care about exposure use public blockchains. Most of finance never will without confidentiality.
We've seen this before. In the early internet, every message, website visit, and credit card number was visible to anyone watching the network. That's where public blockchains are today. HTTPS was created largely so people could pay online safely, and it ended up enabling far more than payments.
We're starting by making financial transactions confidential. Eventually, that opens a design space for applications that couldn't exist otherwise.
4. FHE vs. ZK
Ryan Celaj: What is FHE, and how is it different from ZK? You've called Zcash "encrypted gold" and Zama "encrypted Ethereum."
Rand Hindi: FHE lets you compute on encrypted data. If you add two encrypted numbers A and B, decrypting the result gives you A plus B. You never see the data you're computing on.
For smart contracts, that means both onchain state and transaction data can stay encrypted while the code still runs and remains publicly verifiable. That's why it fits blockchains so well.
FHE also gives you auditability. With some shielded-pool designs, it's hard to verify in real time that a bug wasn't used to create funds. With FHE, we can verify the encrypted supply. And if FHE were ever broken, the main impact would be on privacy rather than on the integrity of the supply.
We launched confidential vaults at the end of June. Within about three months, shielded TVL passed 130 million. For onchain finance, FHE is the right tool.
5. RWAs and the scale of TradFi
Kyle Cai: What surprised me most was Apex Group targeting $100 billion of tokenized assets on T-REX by 2027, with Zama for confidentiality. [Ed. note: In March 2026, Zama became the native confidentiality layer for T-REX Ledger, the RWA infrastructure backed by Apex Group.] What does that partnership mean?
Rand Hindi: It shows how different the scale is. Across all privacy protocols today, shielded value is measured in hundreds of millions to a few billion dollars. Even a single-digit percentage of what T-REX aims to tokenize would be larger than that.
You can read that as "that's amazing," or as "we're very early." DeFi has been talking in billions for five or six years. TradFi thinks in trillions. As those assets move onchain, I believe most of them will eventually be confidential, just as most internet traffic is encrypted today. That's a multi-year journey.
6. From one vault to millions
Kyle Cai: Beyond TVL, what drives Zama's next stage of growth?
Rand Hindi: Vaults. Take a private bank. You ask your banker for something better than T-bill yields. They come back with a strategy, say market-neutral, with your money locked for three months. But you can't use that position as collateral elsewhere, or combine strategies from different banks into one portfolio.
Now imagine each strategy is a vault onchain. You can compose vaults into a meta-strategy no single bank could offer. That's a new paradigm of composable investment products. I expect many more vaults over the next few years: tokenized stock strategies, tokenized funds, permissioned and permissionless. Confidentiality is essential for that to happen.
7. Compliance by design
Ryan Celaj: How do you think about compliance as privacy comes onchain?
Rand Hindi: There are two sides. First, whether a product should be regulated is the issuer's decision, not Zama's. Whoever launches a vault or an RWA decides whether it needs to be permissioned or KYC'd.
Second is privacy itself. We settled on a middle ground: nobody onchain sees your amounts or balances, but the service provider can. A confidential stablecoin issuer, for example, can see transactions using its stablecoin so it can meet AML and counter-terrorism obligations.
We're not protecting you from issuers. We're protecting you from competitors and front-runners. We deliberately don't hide sender and recipient, only amounts and balances, which keeps traceability possible if there's an issue.
Zama is built for compliant financial infrastructure, and institutions want tools designed to support compliance. The only way to make most onchain transactions confidential is to do it legally.
8.  What's next
Ryan Celaj: Anything you can tease for the DefiLlama audience?
Rand Hindi: Two things matter right now. The first is security. With AI and, eventually, quantum, cybersecurity has never mattered more, and I think it's the biggest risk for all of crypto in the coming months. We invest heavily in audits and AI-assisted security research. The second is utility. Beyond growing TVL, we want shielded assets to be used: confidential swaps, more vaults, more integrations. Anything you can do financially onchain, you should be able to do confidentially, with the same or better execution.
Get started
Read DefiLlama's report: A Look Into Confidential DeFi
Explore Zama's confidential vaults: https://app.zama.org/earn 
Follow Zama on X for updates
Confidential onchain finance is just getting started.
Disclaimer: This interview is for informational purposes only and is not investment, financial or legal advice, or an offer or solicitation to buy or sell any crypto-asset. Views and statements are the speaker's own. Figures are as of the interview date.

Following our report on Zama, “A Look Into Confidential DeFi,” Ryan Celaj, our head of research, and Kyle Cai, our former senior research analyst, sat down with the company’s CEO and co-founder, Rand Hindi.

What HTTPS did for the Internet, Zama is doing for Onchain Finance

The Tokenized Stock Market
In just over a year, tokenized stocks have gone from a niche experiment to a real market. DefiLlama's tracked tokenized-stock market cap grew from about $63M in August 2025 to $1.66B as of 14 October, a 26x increase over 14 months. As of that date, DefiLlama tracks 1,284 instruments across 18 chains. Instrument counts need care. In August 2026 the tracked count rose from 574 to 1,235 (+115%), but much of that reflects DefiLlama widening its coverage, not 661 new products launching. Market cap is the cleaner growth signal, though it is also affected by which instruments were tracked at each date.
Our first Spotlight showed that Bitget  has become one of the main venues for trading tokenized stocks, with strong retail demand and growing volume. But volume only tells you how much people are trading. It doesn't tell you how well the market handles that activity: how tight prices remain, how easily large orders can be filled, or how much  liquidity is available outside regular e US trading hours.
This piece asks the next question: now that the market has real size, how good is the liquidity behind it? To find out, we took a live snapshot of four names (AAPL, TSLA, GOOGL and SPY) across perps on Bitget, Binance, OKX and Bybit, and spot on Bitget, Binance, OKX, Bybit and Hyperliquid. The snapshot was taken on September 29, 2026 at 04:47 UTC, outside US market hours. 
Here's what we found. All findings come from this single snapshot, so they describe that moment rather than a standing ranking:
Bitget recorded the deepest perpetual order books on all four instruments tested (AAPL, TSLA, GOOGL and SPY), despite never leading on 24-hour volume. By displayed order-book depth within 1% of the price, Bitget ranked first among the four perp venues on every name. By 24-hour volume it ranked second or third.
Onchain books aren't comparable yet. Hyperliquid's spot books had almost no orders near the price, with spreads in the thousands of basis points. We report on-chain and centralized books separately.
Large-order execution revealed significant differences between venues. Bitget and Binance were the only spot venues with sufficient displayed liquidity to fill a simulated $100,000 buy order across all four instruments, although execution costs varied by assets. 
Every ranking and figure below is a single reading from September 29, 2026 at 04:47 UTC and should not be read as an average or a standing ranking.
Where Tokenized Stocks Trade
Before comparing execution quality, it's worth being clear about what's actually being traded, because "tokenized AAPL" isn't one product. Perps on Bitget, Binance, OKX and Bybit are cash-settled derivatives that track the stock price. They're created by each exchange, so there's no underlying share or issuer behind them. Spot tokenized equity products provide exposure to underlying equities through different issuance and custody structures. Depending on the product, tokens may represent claims on underlying shares or other forms of economic exposure, with investor rights varying by issuer. 
Venue
Product
Symbol convention
Issuer / brand
Bitget
Spot
r-prefix (rAAPL, rGOOGL, rSPY)
Reality, a Bitget subsidiary — branded "Reality Tokenized"
Binance
Spot
B-suffix (AAPLB, TSLAB,
GOOGLB, SPYB)
B-Tech Holdings Limited (confirmed for TSLA, GOOGL, SPY; not found for AAPL)
OKX
Spot
X-prefix (XAAPL, XTSLA,
GOOGL, XSPY)
Not identified
Bybit
Spot
X-suffix (AAPLX, TSLAX,
GOOGLX)
Not identified (tickers match other venues' products, so not confirmed as Bybit's own)
Hyperliquid
Spot (on-chain)
Ticker as-is (AAPL, TSLA, GOOGL)
Not identified
For spot products, the issuer is shown where it could be identified. "Not identified" means we could not confirm one.
We chose AAPL, TSLA, GOOGL and SPY because they're the most widely listed tokenized stocks across the venues we tested. Every venue offers at least one of them, and three of the four are available on spot at all five venues. SPY is the only index product in the set, which lets us check whether liquidity looks different for a broad-market ETF than for single stocks.
The venues match those covered in our first Spotlight: Bitget, Binance, OKX and Bybit for perps, with Hyperliquid added on spot as the on-chain comparison. Using the same list across both markets keeps the comparison like-for-like.
How we measured
All figures in this piece come from one live order-book snapshot, taken on September 29, 2026 at 04:47 UTC, outside US market hours. We covered four instruments (AAPL, TSLA, GOOGL and SPY) across the venues listed above. Spreads are the gap between the best bid and best ask at that moment, in basis points. They are a single-moment reading, not an average over a trading day, so they can look wider or tighter than the typical spread at other times, especially on individual names. Depth is the dollar value of resting orders within a set distance of the price. Slippage is simulated: we walked each order through the book as displayed, so it shows the cost the book implied at that moment, not an actual executed trade.
The Busiest Venue Isn’t The Deepest
Most trackers rank venues by trading volume. But volume only shows how much is being traded, not how much liquidity is sitting in the book. To test whether the two line up, we ranked each venue by 24-hour perp volume and then, separately, by order-book depth within 1% of the price.
Ticker
Venue
Volume rank (of 4 venues; OKX not shown)
Depth rank (of 4 venues; OKX not shown)
Pattem
AAPL
Binance
1
2
Aligned
AAPL
Bybit
2
3
High volume, thin book
AAPL
Bitget
3
1
Low volume, deep book
TSLA
Binance
1
2
Aligned
TSLA
Bitget
2
1
Aligned
TSLA
Bybit
3
4
Aligned
GOOGL
Binance
1
2
Aligned
GOOGL
Bybit
2
3
High volume, thin book
GOOGL
Bitget
3
1
Low volume, deep book
SPY
Binance
1
2
Aligned
SPY
Bitget
2
1
Aligned
SPY
Bybit
3
4
Aligned
The clearest pattern is Bitget. It never ranks first by volume, yet it has the deepest order book on all four names. On AAPL and GOOGL the gap is widest: Bitget ranks third of four venues by volume but carries the most resting liquidity.
This matters because volume and depth measure different things. Volume shows how much has already traded over the past 24 hours. Depth shows how much the book can absorb right now without moving the price. A trader placing a large order cares far more about the second number, and on that measure Bitget had the deepest book on every name, despite trailing on volume each time.
That's easy to miss on a volume leaderboard, which is how most tokenized-stock trackers rank venues today. This snapshot shows the difference in displayed depth, but it doesn't establish what drives it.
Bybit shows the opposite. On AAPL and GOOGL, it ranks second by volume but third by depth, so heavy trading sits on a thinner book. These are the same two names where Bitget looks quiet but deep, while Bybit looks busy but shallow. On TSLA and SPY, Bybit ranks low on both, so the mismatch doesn't apply there.
Binance sits between the two. It leads on volume for every name and ranks second on depth each time, just behind Bitget. It shows a venue can be strong on both measures, but even the most active venue here didn't have the deepest book.
What It Actually Costs To Trade
Rankings help compare venues at a glance, but they don't say what a trade actually costs. This section puts numbers on it in three ways: spreads, depth at different distances from the price, and slippage at different order sizes.
Hyperliquid is included in the spot data but shown separately. Its spreads run in the thousands of basis points, compared with single digits on centralized venues, so averaging it in would produce a "typical" figure that no real venue matches.
Spreads
These are single-moment readings from the September 29 snapshot, taken outside US hours.
The 0.83 bps reported in our earlier research was calculated across a different dataset and sampling period. The figures below reflect a single order-book snapshot taken on September 29, 2026, outside US market hours, and are therefore not directly comparable. . A median across many readings smooths out moments when a book is thin or wide. A single reading does not, so the two numbers measure different things and shouldn't be compared directly. Perp spreads in this snapshot ranged from about 0.13 to 1.75 bps. 
On perps, Bitget isn't the tightest venue. Its spreads are wider than Binance and OKX on TSLA (1.4 bps vs 0.28 bps) and GOOGL (1.75 bps vs 0.29 bps), and roughly in line with the field on AAPL and SPY.
On spot, Bitget's spreads are wider than Binance's on every name. On GOOGL, for example, Bitget quoted 9.06 bps against Binance's 0.88 bps. But the spread only shows the price of the very first order. For larger trades, what matters is how much size sits behind it.
Depth
Spot books are shaped differently from venue to venue. Bitget's book keeps adding size the further you go from the price. On AAPL, its depth grew from $42.7k within 0.5% of the price to $122.1k within 1% and $208.6k within 2%, with a similar build-up on GOOGL and SPY. Binance and OKX, by contrast, concentrate more of their size close to the price, with less added further out. On AAPL spot, Binance's book was deeper than Bitget's at every distance shown; Bitget's depth lead in this snapshot was on perps.
This matters for larger orders. A small trade only touches the orders closest to the price, while a large one works through them and relies on size further out. In this snapshot, Bitget's spot liquidity was spread more evenly across price levels, while Binance's was concentrated close to the price. The snapshot doesn't tell us why the books are shaped this way.
Slippage
Slippage figures here are simulated execution costs. We walked each buy order through the order book as displayed at the time of the snapshot, so they are not actual executed trades. Real fills would also depend on how the book moves while the order executes.
At $1000, simulated slippage was below 5 bps across almost all , centralized venues tested, with differences becoming more pronounced at larger order sized. 
Slippage is the extra cost of a buy order relative to mid price / best ask, found by walking the order through the displayed book. At $100k, only Bitget and Binance could fill a simulated order on every ticker. Binance was cheaper on AAPL (11.8 bps vs Bitget's 96.2 bps), TSLA (3.4 vs 7.7) and SPY (2.8 vs 4.7). Bitget was cheaper on GOOGL (6.4 vs 9.5). The AAPL gap is an outlier, and because it reflects one snapshot of the displayed book, we would not read it as a standing difference. Neither venue wins across the board.
Perps vs spot
We did not compare perp and spot execution costs like-for-like, so this piece makes no claim about which is cheaper. The two are different products. Perps are cash-settled derivatives with no underlying share, and their costs include trading fees, funding, margin and liquidation risk, none of which are covered here. Spot slippage above is simulated from the displayed book for Bitget and Binance only, because the other venues could not fill a $100K order.
Conclusion: Volume Alone Doesn’t Show Where Liquidity Sits
Tokenized stocks now have real scale, and this piece set out to test whether the liquidity behind them matches the headline numbers. In this snapshot, volume was a poor guide to where displayed liquidity sat. Bitget never led on 24-hour perp volume, yet it had the deepest perp book on all four names we tested. It was also one of only two spot venues, alongside Binance, that could fill a simulated $100k order on every ticker.
Perp depth was Bitget's clearest strength in this snapshot. On spreads, it quoted wider than Binance on several names, and it cost more on most simulated $100k spot fills. The gaps are small on most names, though AAPL was a clear outlier on $100k spot fills.
There's a lot this snapshot can't tell us. It can't show what drives the differences in displayed depth. We didn't compare fees, which affect where liquidity providers post size. And because this was a single snapshot taken outside US trading hours, we don't know whether these rankings hold when the US market is open or across other market conditions. A US-hours comparison is a natural next step.
As tokenized equities expand across trading venues, liquidity and execution quality are becoming important measures of market competitiveness. This analysis shows why trading volume along cannot capture those differences. In the September 29 snapshot, Bitget recorded the deepest perpetual order books across all four instruments tested, while spot execution costs varied across venues and order sizes.

Now that the market has real size, how good is the liquidity behind it? To find out, we took a live snapshot of four names (AAPL, TSLA, GOOGL and SPY) across perps on Bitget, Binance, OKX and Bybit, and spot on Bitget, Binance, OKX, Bybit and Hyperliquid.

Tokenized Stocks Have Volume. Do They Have Liquidity?

Gate's biggest repositioning since its 2013 founding: what Gate Money offers, the exchange behind it, and how it compares with peers.
For most of crypto's history, an exchange had one job: match buyers and sellers. Gate is now betting that a trading venue is no longer enough. On October 7, founder and CEO Dr. Han took the main stage at TOKEN2049 Singapore to launch Gate Money, a single account for holding, moving and spending money across crypto, fiat and stocks.
Gate Money is the centerpiece of Gate's "all-in-one platform" brand upgrade, under the theme "One Gate, Everything Money." Traditional finance splits everyday money tasks across separate accounts, apps and systems. Gate Money brings them into one, built on a principle Dr. Han set out on stage: "Accounts should not be the boundary of funds."
This report covers what Gate Money offers, the scale of the exchange behind it, how the market reacted, and how it compares with similar moves by OKX and others.
Key takeaways
One account for everything. Gate Money combines crypto, fiat, stocks, a bank account in the user's own name and a payment card. It is live now for eligible users.
Built on a top-tier derivatives venue. Gate now ranks 3rd by derivatives volume, behind only Binance and OKX, at $12.77B a day. Its share of derivatives volume across five major exchanges rose from 9.3% to 11.2% month-on-month.
GT rallied before the launch. GT rose about 70% from its August low to around $11, and has held near that level since the keynote.
Gate leading the race. Others like OKX are building similar products. Gate leads with a live bank account.
Latest Proof of Reserves. Gate's latest Proof of Reserves report shows total reserves of $8.68B and an overall reserve ratio of 118%.
What Gate Money is
Gate Money is a single account that connects:
Digital assets and fiat currencies: Users hold multiple fiat currencies and digital assets in one account system, with a consolidated view of balances and fund movements. Fiat accounts can be managed separately by currency. Fiat and on-chain deposits, withdrawals and transfers are supported.
Stocks: Access to stocks and other traditional financial assets, now organized in a clearer account structure.
A bank account in the user's own name: Eligible users can upgrade to one, to receive funds and make withdrawals with their personal account details.
Payments: Gate Card works online and in stores with fiat currencies or digital assets and can be linked to Apple Pay and Google Pay. In certain countries and regions, users can also access ATM withdrawals and local QR-code payments.
Convert and Earn: Convert swaps between assets. Earn offers earning products, with eligible balances such as USDT earning up to 3.6% and no lock-up.
Why now: money has stopped fitting in boxes
Stablecoins, tokenized real-world assets and blockchain payment rails now give more kinds of assets digital liquidity and instant settlement. If a stock, a gold position and a stablecoin can all settle quickly, there is less reason to keep them in different apps. 
Holding, growing and spending an asset used to be three separate steps. Gate Money is designed to make them one.
Gate Money is backed by licenses across several jurisdictions:
a MiCA crypto-asset service provider license and a PSD2 payment institution license in Malta
a full virtual asset service provider license from VARA in Dubai
a registration with AUSTRAC in Australia as a digital currency exchange
a money services business registration with FinCEN and 37 state-level money transmitter licenses in the US, held through Gate US and covering 47 US jurisdictions
local entities in markets such as Japan, covering 81 regulated jurisdictions in total
These are supported by partnerships with payment institutions, banks, custody providers and clearing institutions.
Gate serves more than 61 million users and offers 5,300+ digital assets and 12,800+ stock assets, plus metals, forex, indices and commodities.
The foundation: a derivatives-heavy exchange
Gate Money is built on an exchange that is already large. Of the 88 centralized exchanges tracked by DefiLlama, Gate sits in the top tier on derivatives and reserves.
Metric
Gate
Rank
Derivatives volume, 24h
$12.77B
3rd
Spot volume, 24h
$1.60B
2nd
Reserves excluding its own token
$6.19B
6th
Total tracked reserves (including its own token)
$7.49B
n/a
24h volume figures as of October 20, 2026. The spot rank excludes Bitrue, whose spot volume carries a data-quality warning.
By derivatives volume, Gate is now third, behind Binance and OKX. Binance leads by a wide margin, at $50.24B a day against OKX's $18.85B.
Gate's derivatives volume is about 8 times its spot volume ($12.77B against $1.60B). That is the same shape as most large exchanges, where perpetuals dominate. It also shows why Gate Money matters strategically. Trading volume is cyclical and cheap to switch away from. An account that holds incoming funds, stocks and a card is stickier.
How Gate's derivatives volume has moved
The chart shows the average daily derivatives volume by week for Gate against OKX, Bybit and Bitget since mid-June. Binance is left off because its volume is several times larger and would flatten the other lines.
Monthly averages of daily derivatives volume, from the same data:
Month
Gate
OKX
Bybit
Bitget
Binance
June (from the 14th)
$10.99B
$21.49B
$12.20B
$8.28B
$46.25B
July
$9.51B
$19.34B
$10.01B
$6.66B
$45.38B
August
$8.87B
$20.05B
$11.88B
$7.36B
$50.67B
September
$13.15B
$23.27B
$14.42B
$7.62B
$59.69B
October (1st to 11th)
$11.23B
$19.19B
$12.24B
$7.08B
$51.63B
Gate's September average was its strongest of the period. August 19 is excluded, as Gate's volume is recorded as zero that day, a likely data gap.
Among these five exchanges, Gate's share of derivatives volume rose from 9.3% in the 30 days to September 11 to 11.2% in the 30 days to October 11. This is share of the five-exchange set, not of the whole market. Gate's average daily volume over the same two windows rose from $10.97B to $12.7B.
The launch has not lifted volume so far. In the five days from October 7 to 11, Gate averaged $11.1B a day, against $11.9B in the week from September 30 to October 6. That is a short window, but the September rise came before the announcement, so it should not be credited to Gate Money.
What Gate holds
Gate's latest Proof of Reserves report, released on 21 October, shows total reserves of $8.68B and an overall reserve ratio of 118%.
The two sources in this section measure different things. Gate's Proof of Reserves is its own report, comparing the assets it holds with what it owes users. DefiLlama's figures, used for the rest of this section, only count on-chain assets in wallets DefiLlama tracks for Gate, priced at market, and can leave out assets held elsewhere. They do not include what Gate owes users, so they show what Gate holds on-chain, not how fully user balances are covered.
On Gate’s latest PoR data, its tracked reserves are diversified, led by Bitcoin.
BTC is the largest tracked holding at $2.12B.
GT, Gate's own token, comes second at $1.17B. This is the main reason Gate's total tracked reserves ($7.49B) are higher than its reserves excluding its own token ($6.19B).
ETH and ETH-linked assets follow, with ETH at $1.08B.
Stablecoins and yield-bearing dollars appear healthy, including USD1, USDT, USDC, GUSD and more, at an aggregated value of $1.9B.
Gate's tracked clean assets in reserves come to $6.2B, fifth among exchanges.
Exchange
Tracked clean assets in reserves
Binance
$146.5B
OKX
$29.6B
Bybit
$15.5B
Robinhood 
$15.0B
 Gate
$6.2B
 Bitget
$5.5B
 Gemini
$5.49B
Gate's on-chain footprint is small
Gate Money is a centralized-finance product, and that is where Gate's weight lies. Its Gate Layer network holds $3.36M of TVL, up 8.8% over 30 days, and generated about $120K of app fees over 30 days. That is tiny next to the exchange's $6.19B of reserves, so the on-chain side is not what this launch rests on.
Strategy: three business lines
At a press Q&A on October 8, Dr. Han said "the core of growth is still user demand." As users' needs extend from trading to investing, payments, transfers and fund management, Gate plans to widen its services to match.
He set out three business lines for the next three years:
Digital asset trading. Still the core business, with continued investment across spot, derivatives and on-chain markets.
Gate Money. Fund management, payments, transfers and multi-currency accounts.
Stocks and other asset services. More ways to access and manage different asset types on one platform.
Beyond these, Gate will keep watching opportunities in AI and the on-chain ecosystem.
Gate Money against its peers
Gate is not the only exchange trying to turn itself into an everyday finance app. In recent months, the biggest exchanges have all announced versions of the same idea. The differences are in which pieces they lead with: bank accounts, payments, real stocks or tokenized stocks.
The peer set, product by product
Gate
OKX
Coinbase
Branding
"One Gate, Everything Money"
OKX Money app (select markets)
"Everything Exchange"
Bank-style account
Bank account in the user's name for eligible users, live now (Oct 7)
Fund with local currencies and hold USDG, USDC or USDT (Oct 6)
None identified
Card and payments
Gate Card; Apple Pay and Google Pay; ATM and QR in some regions
Card in the OKX Money app
Coinbase One Card
Stocks
Access to stocks in the account
Tokenized US stocks via partners such as Ondo and xStocks
Tokenized stocks and options announced as coming
Standout feature
One account for banking, payments and stocks
Digital-dollar balances with a card
Borrowing against staked Solana
Source: company announcements, press releases and media coverage, as of October 12, 2026.
How Gate's approach differs
Gate's account is live; Bybit's is still being built. Gate's bank account is available now for eligible users. As of September, Bybit held an electronic money institution license in Austria, expected a MiFID license, and planned a banking-style app with its own accounts and IBANs.
The "use assets without selling" idea is not unique to Gate.
Bitget lets 100 stock tokens serve as margin or collateral inside its unified account.
Coinbase has announced borrowing against staked Solana.
At the keynote, Dr. Han said stocks held in Gate Money can be used for daily spending without selling them first.
Gate has not yet published the details. Limits, haircuts and collateral rules will decide how useful it is.
Timing is crowded. OKX launched its OKX Money app on October 6, the day before Gate's keynote. Bybit began offering US stocks and ETFs on October 12. Gate's launch sits in a cluster of similar announcements, but it is the only one of these with a bank account already live.
The scale comparison
Product breadth only matters if there is trading and balance-sheet scale behind it. Here is where the closest peers stand in DefiLlama's data.
Exchange
Spot volume (24h)
Derivatives volume (24h)
Reserves excl. own token
Binance
$8.16B
$40.55B
$148.0B
OKX
$1.55B
$15.54B
$30.1B
Bybit
$1.33B
$9.71B
$15.7B
Gate
$1.30B
$8.82B
$6.19B
Bitget
$0.60B
$5.24B
$5.70B
All figures from DefiLlama, October 12, 2026, 08:01 UTC.
On October 12, Gate ranked fourth of these five on combined volume. Its spot plus derivatives volume was $10.1B, against $11.0B for Bybit, $17.1B for OKX and $48.7B for Binance. Gate was well ahead of Bitget at $5.8B.
Gate maintains a substantial reserve base of $6.19B excluding its own token, placing it above Bitget and supporting the platform as it expands beyond trading into account-based financial services. As users keep more balances on-platform, reserve transparency becomes increasingly important.
Derivatives dominate everywhere. Derivatives volume is between 5 and 10 times spot volume across all five exchanges, and 6.8 times at Gate. The all-in-one pitch is partly an attempt to broaden revenue away from that dependence.
Coinbase and Kraken are excluded, as comparable current data is not available for either.
The wider picture
Large exchanges are pushing past trading into payments, stocks and wealth products. What stands out about Gate Money is that it is built as one account with a bank-style front end, not as a set of separate products inside an exchange app.
That framing has real strengths. Gate already operates in derivatives, spot, a card and stocks, so Gate Money is mostly about connecting existing pieces. It also targets a real gap, because most crypto users still move between exchange, bank and broker accounts to do everyday things.
For Gate, Gate Money is the bid to turn a place users trade into a place they live financially. The launch is early, and the numbers that will say whether it worked, such as activations, card spend and balances held, are not yet public.
Data as of October 12, 2026, unless stated otherwise. Sources: DefiLlama for exchange, reserve and price data; Gate for product, license, company and Proof of Reserves figures; company announcements, press releases and media coverage for peer details.

Gate Money is the centerpiece of Gate's "all-in-one platform" brand upgrade, under the theme "One Gate, Everything Money." This report covers what Gate Money offers, the scale of the exchange behind it, how the market reacted, and how it compares with similar moves by OKX and others.

Gate Money: From Crypto Exchange to All-in-One Financial Account

In late 2025, Kalshi still worked like a classic prediction market. Nearly all of its trades happened on single markets, and the most traded prices sat around 50¢, where an outcome has even odds.
Combos have changed that picture. As we highlighted in our first Kalshi spotlight, notional volume hides how little cash actually enters the platform and in September, a contract implies less than an 8% chance of paying out.
Kalshi has kept adding products and leverage since. Single markets have traded at prices down to 0.1¢ since late July 2026, and combos down to 0.01¢ since September 3. On top of that, 15-minute crypto markets let traders bet on price moves that settle every quarter of an hour. The lower the price, the bigger the payout for the same stake: a combo bought at 0.01¢ pays 10,000x.
That raises a simple question: is Kalshi still a prediction market, or is it becoming a long-shot platform? The data is clear: behavior has changed, and by trade count, we lean toward the second.
To test this, we looked at two dimensions, cash and flow, across several metrics. Cash tells us where the money goes, and flow tells us where the trades go. Together, they give a clear picture of what Kalshi is becoming.
Combos pull in cash faster than trades, while singles do the opposite
Without surprise both Singles and Combos have grown in both number of trades per quarter and cash per quarter. Though, what is key here is to note that the growth in cash is higher on combos compared to the growth of trades. This is not the case for singles and that hints at something: that not only combos grw in number of cash it also grew in cash per trade while single has grown in trades but they carry less cash.
The fill sizes by price zone make this split clear. In single markets, the average fill across all price zones fell from $127.31 in October 2025 to $39.63 in September 2026, and the median fell from $14.78 to $4.59. Both ends of the curve shrank by about three-quarters. At 90¢ and above, the average fill went from $319.62 to $82.81, and under 10¢ it went from $15.62 to $3.82. Single tickets got smaller everywhere, from near-certain outcomes to long shots.
Combos moved the other way. Their average fill rose from $37.78 to $61.69, while the median remained more or less stable around $9.30. The rise came mostly from one end of the curve. At 90¢ and above, the average combo fill went from $202.64 to $512.82. Under 10¢ it stayed at about $12.50 although its median fell from $7.17 to $4.68.
The two moves inverted the balance. A year ago, a combo fill carried less than a third of the cash of a single one: $37.78 against $127.31. By September 2026 it carried more, at $61.69 against $39.63.That’s a striking shift, and it hints at three things. 
First, singles keep growing but are losing ground in bettors’ wallets. Single trades grew 19x in a year, yet each one carries less cash, and singles’ share of Kalshi’s cash fell from 99.7% to around 80%.
Second, combo bettors put big money on near-certain outcomes. The heaviest combo fills now sit at 90¢ and above, mostly on the side that bets a parlay will fail.
Third, combo bettors also take more risk at the bottom of the curve. In September, an average combo fill under 10¢ carried more than three times the cash of a single one, at $12.55 against $3.82, despite lower odds of winning. 
The lower end: the hidden growth everyone misses under the 10c
Trades priced under 10¢ went from 9.2% of all Kalshi trades in October 2025 to 18.6% in September 2026. Put differently, nearly one trade in five now carries a price-implied chance of success under 10%. In absolute terms, that’s 1.48 million trades in October against 64.2 million in September, a 43x increase while all trades grew 21x.
One step further down, the picture is sharper. Trades under 5¢ went from 4.4% to 12.3% of all trades, or from 0.71 million to 42.5 million. Trades at 1¢ or less, which our data tracks from March 2026, went from 0.5% to 4.3% of all trades, reaching 14.9 million in September.
Combos are the obvious suspect. Around 55% of combo trades priced under 10¢ with significant growth in trades at 1¢ or less which went from 6.0% of combo trades in March to 22.0% in September. 
Interestingly, Singles move the same way from a lower base. Their under 10¢ share went from 8.7% to 13.2% of single trades, and their under-5¢ share from 4.0% to 7.8%. Trades at 1¢ or less grew from 69,000 in March, or 0.08% of single trades, to 5.2 million in September, or 1.71%. 
Put simply, the riskiest end of the curve is growing in both products, for two different reasons.
On combos, the entry price is the main driver. Every leg pushes the price toward zero, and on September 3 Kalshi lowered the floor from 0.1¢ to 0.01¢. The cheapest combos are mostly built on sports legs.
On singles, finer price grids since late July play a part, but the bigger driver is the 15-minute markets and their wave of small, frequent trades. They went live on BTC, ETH and SOL by January 2026 and now cover more than a dozen assets, with gold and silver added on August 3. By September, they made up half of single trades and nearly half of Kalshi’s trade growth over the year.
The higher end: what the 90¢+ growth really is
The cheap end isn’t the only part of the curve that is growing. In single markets, trades priced at 90¢ and above went from 1.2 million in October 2025 to 46.5 million in September 2026, or from 7.5% to 15.4% of single trades. In combos, the same zone went from about 1,400 trades to 1.7 million, or from 0.9% to 3.9% of combo trades. Both products reach this zone for different reasons, and those reasons explain the fill sizes we saw earlier.
On singles, the 15-minute markets again do most of the work. They account for 58% of single trades at 90¢ and above. This is most likely mechanical: as the window closes, the price of the nearly decided side climbs toward 100¢ and trades stack up there. This type of trade largely explains why single ticket size collapsed. Fifteen-minute trades are small, with a median fill of about $2.80, and they now make up most of the 90¢+ zone. 
On combos, the 90¢+ zone is almost entirely one trade: buying the NO side of a parlay, which is a bet that it fails. NO buys make up 94% of the zone, with 1.6 million trades in September.
In both cases, the 90¢+ growth is a bet on near-certainty, at the safe end of the curve. On singles it follows the 15-minute product, and on combos it comes from large favorite-side purchases. 
Kalshi, still a prediction market?
All of the above shows that, through the design of its products, Kalshi is pushing trades toward the two extremes of the price curve, but not symmetrically. On combos, 14 trades sit under 10¢ for every one at 90¢ and above. On singles, the 15-minute markets push traders either toward a last-minute safe bet that earns a little or toward the life changing long shot.
So the answer to our question is clear. Kalshi’s trades say less and less about what people think will happen. Instead, they increasingly answer two other questions, where the safest quick gain lies and which long shot pays the most, and the more trades those questions draw, the less Kalshi looks like a prediction market.
Methodology
This analysis uses Kalshi fills for full months from October 2025 to September 2026. Combos are fills in combo series, and everything else is a single market. Cash is the price the taker paid on the traded side, and notional counts each contract at $1. Price zones use the traded-side price, and the 0–1¢ zone includes fills at exactly 1¢.

To test this, we looked at two dimensions, cash and flow, across several metrics. Cash tells us where the money goes, and flow tells us where the trades go. Together, they give a clear picture of what Kalshi is becoming.

Prediction or roulette? Kalshi’s trades are moving to the edges of the price curve

Ethereum

Bitcoin

Tether

USDC

On September 24, attackers drained $387.5M from Bitget’s hot wallets. Within a day, part of the funds were already being bridged and swapped across numerous assets, including XRP, ETH, BTC, USDT and USDC, through venues such as  THORChain,  Uniswap, Chainflip, NEAR Intents, 1inch  and LayerZero.
Following the hack, Bitget CEO Gracy Chen published the attacker addresses and formally asked THORChain to refuse them service. THORChain declined. Its position is that the protocol is open to anyone by design and doesn’t filter who uses it, and the emergency halts that protect the protocol can’t be used to freeze particular funds. It added that the attackers behind its own May exploit, which took $10.7M from its pools, were never blacklisted either and kept swapping freely.
The refusal reignited the debate around DeFi, neutrality, decentralization, and what protocols should or shouldn’t do in these situations. What stands out is that no side is clearly right or wrong, it’s more of a stance.
That makes this a good moment to revisit the values behind decentralized finance, starting with Satoshi’s whitepaper, and show why they now operate more as a spectrum. Protocols sit at different points along it, and each point carries a cost. We then look at how the venues involved in the Bitget hack acted relative to their stated ethos and promises. MistTrack’s tracking of the funds lets us draw a clear picture of the situation as it stands two weeks later.
For protocols built on a decentralized ethos, neutrality is credible only if no single party holds the power to change the rules. For those that retain discretionary control, having the power to act doesn’t guarantee effective enforcement: responses remain slow, coverage incomplete, and alternative routes accessible.
The four principles behind DeFi must be seen as a spectrum
In 2008, Satoshi Nakamoto published Bitcoin: A Peer-to-Peer Electronic Cash System. Its ambition was to let people send value directly to one another without relying on a financial institution to process the payment, replacing trust in an intermediary with a network that validates transactions according to shared rules. That proposition laid the foundation for four concepts that have since become central to crypto: decentralization, permissionlessness, censorship resistance, and neutrality.
The four are often used interchangeably, but three describe what users can expect from a system and one describes how it’s built.
Permissionlessness concerns access, meaning anyone can use the system without approval. The whitepaper’s abstract describes exactly that: online payments that move straight between two parties, with no financial institution in the middle.
Censorship resistance concerns execution, the ability of a valid transaction to go through despite attempts to block it. In the whitepaper, every new transaction is shared with all the nodes on the network, so no single participant decides whether it gets through.
Neutrality concerns rules, and whether they apply consistently regardless of who the user is or what interests are at stake. The whitepaper identifies participants only by public keys that don’t need to reveal who is behind them, so the rules can’t depend on identity.
Decentralization is architecture: it concerns how control is distributed across the parties that run the system. The whitepaper describes nodes that need no identification, coordinate only loosely, and can join or drop out whenever they like. The more decentralized a system is, the more censorship-resistant it tends to be, and the better positioned it is to stay permissionless and neutral. The degree of decentralization varies from one protocol to another
These properties reinforce one another, but they don’t necessarily coexist. A system can deliver one without the others, which is why the distinction matters when judging how a protocol behaves under pressure.
Decentralization, for example, can make a system resistant to unilateral change while keeping it fully permissioned. Canton Network is one case: it’s operated by multiple independent institutions but restricts who can take part. Conversely, a centralized entity can offer permissionless access to its products while keeping the power to censor, as Circle does with USDC: anyone can hold and send it, but Circle can freeze an address. At the far end, a system can be decentralized, neutral, censorship-resistant, and permissionless all at once, like THORChain.
What does that mean in practice? That these properties don’t fit a binary description. Protocols and entities position themselves freely on a spectrum with four dimensions, one per principle, with plenty of nuance in between.
Brand
Permissionless
Censorship resistance
Neutrality
Decentralization
THORChain
Full
Full
Full
Full
Chainflip
Strong
Partial
Partial
Strong
NEAR Intents
Partial
 Limited
Limited
Partial
Circle
High
Minimal
Limited
Minimal
Tether
Strong
Minimal
Minimal
Minimal
Uniswap
Full - Partial
Full - Strong
Full - Limited
Full - Minimal
1inch
Full - Limited
Full - Partial
Full - Limited
Strong - Minimal
In practice, the Bitget funds crossed every point on the spectrum
The Bitget hack put that spectrum to the test. According to MistTrack, the attacker moved the $387.5M through cross-chain swap protocols, bridges, and DEXs, and two weeks later 86.2% of the stolen value sits in BTC. Along the way, the funds met venues at every point described above, from those that can’t refuse a transaction to those that can freeze it at the token level.
THORChain followed its ethos
THORChain was the attacker’s main route, processing approximately $280M in swaps. This represented roughly 36% of the $780M in cumulative volume recorded across all venues used by the attacker.
That lead follows directly from its design. THORChain has no mechanism to refuse an address: swaps are executed by its node operators according to the protocol’s rules, and none can single out a user. Its only emergency tool is a halt, which stops an entire chain for every user. A halt can be triggered automatically, as it was during THORChain’s own exploit in May 2026, or manually by node operators (a single operator can pause the chain, but any other operator can lift that pause, so a halt only holds if the operators collectively let it stand).
Critics, including OKX founder Star Xu, argue that a network able to stop when its own funds are at risk can’t claim to be like Bitcoin when it refuses to stop for anyone else’s. But halts of this kind have precedents on major chains. In 2010, Bitcoin’s developers and miners coordinated to discard the blocks created by a bug that minted 184 billion BTC. In 2022, BNB Chain’s validators halted the chain after an exploit minted roughly $570M of BNB, though its validator set is far from the most decentralized. In each case, the goal was to protect the chain's integrity, not blacklist a user.
That’s also why THORChain halted in May: had the exploit continued, the protocol could have been drained to the point of having nothing left to serve its users. The halt only prevented more funds from being stolen and did nothing to recover the funds that had already been taken. THORChain also applied the same permissionless principle to itself, since the attackers behind that exploit were never blacklisted and remained free to swap on the protocol.
THORChain’s response, both to Bitget and to its own exploit, therefore matched its architecture and its stated principles, whether one likes it or not. The cost of that consistency shows in the numbers: sitting at the far end of the spectrum meant carrying most of the stolen funds. 
Freezing powers stopped 0.3% of the stolen funds
Across every venue, $1.26M has been frozen so far, 0.3% of the stolen funds. The power to freeze exists at several points along the funds' path, from stablecoin issuers to protocols and exchanges, yet each caught only a fraction.
Stablecoin issuers have the most direct lever with the ability to freeze at the address level. About a fifth of what was stolen, $75.47M, was in USDT, USDC and USDT0, and the attacker kept using them. $55.21M moved through Circle’s CCTP and $59.27M through USDT0 on LayerZero. Circle froze $568K and Tether $239K, about 1% of the stablecoins stolen. 
Most of the remaining funds, $298M in XRP, ETH, ZEC, BNB, AVAX, and TRX, were in assets no issuer can freeze. NEAR Intents, which screens swap requests against AML databases and can hold funds during execution, froze $266K. Unit, which issues the BTC traded on Hyperliquid, froze $170K. Exchanges such as KuCoin and Bybit stopped small amounts sent to their deposit addresses, about $14K in total.
This gap can further be explained by three constraints. The first is speed: an asset can only be frozen while someone still holds it, and the attacker swapped out faster than the freezes arrived. The second is selectivity: the funds were split across more than 1,000 addresses, each of which had to be identified. The third is coordination: Circle’s policy on freezing is tied mainly to legal obligations, threats to the stablecoin’s security, and urgent requests from authorities. All of which take time to set in motion. None of this means the issuers or protocols failed, but highlights the limitations of today’s means.
Chainflip rejected the funds, then processed them through CoW Protocol
Chainflip sits closer to THORChain on the spectrum, with one difference that matters here: the brokers who open deposit channels can reject a deposit before it’s swapped. MistTrack reported that Chainflip did exactly that with one deposit tied to the hack, which was refunded to the sender rather than frozen.
Attacker-linked funds still went through Chainflip by another door. SlowMist found an automated script creating CoW Protocol orders whose receiving address was a pre-prepared Chainflip deposit address, so once CoW settled an order, Chainflip swapped the assets into BTC. MistTrack counts around  $31M transiting through Chainflip in total.
Why the CoW-routed deposits got through isn’t documented. One hypothesis is that Chainflip saw CoW’s settlement contract as the depositor rather than a flagged wallet. Another is that the channels were opened through a broker that doesn’t screen, since any account can become one. SlowMist’s founder described the broader pattern: by splitting the funds and retrying across different bridges, the scripts could switch routes whenever a transfer was rejected, and in his view, AML checks are falling behind this kind of automation.
The problem therefore runs deeper than one broker’s filter. The real question is whether protocols that choose to intervene can build tools that follow funds across several hops or whether entry-point screening will always be one door behind.
Screening rules change with the product, and the funds went through anyway
After THORChain, DEXs carried the largest share of activity, with $273.98M in attacker-linked swaps. Uniswap accounts for $114.92M, CoW Swap for $65.48M, MetaMask Swaps for $56.85M, 1inch for $20M and SunSwap for $15.07M.
Several of these brands advertise screening, but the rules can change significantly depending on the product a user uses. Uniswap Labs blocks flagged wallets on its app, wallet and extension, while the contracts underneath stay open to any other interface. 1inch screens wallets through TRM Labs on its dApp and requires KYC for its Business API and Aqua product, but there’s no restriction on the classic router. In other words, the same brand can turn a wallet away on one product and still serve it on another.
That raises two questions: if a brand presents itself as protective but leaves gaps across its product suite, where does it actually sit on the spectrum, and what is it accountable for? And if screening covers only part of the product suite, does it really protect its users?
Every position on the spectrum comes at a price
Most of the debate around Bitget has been framed as a binary choice. Still, the more useful lens is the spectrum: DeFi offers a range of products and positions, and every type of user can find the one that fits how they want to transact.
However, each position on that spectrum carries its own price. For the most open protocols, those closest to the ethos of Bitcoin’s whitepaper, it means illicit funds can pass through. It also means legitimate users can’t be subjectively judged on their identity, country of residence or wealth. For the protocols that keep control, the trade-offs run the other way. Legitimate funds can be frozen by mistake, causing stress or financial loss, and users can be cut off because of where they live or because of shifting political and regulatory pressure. More importantly, their controls are also limited by what’s technically feasible, as the Bitget case showed.
In the end, with or without the power to block, freeze or screen, almost all of the stolen funds went through. The questions worth debating are therefore far more numerous than the one put to THORChain.
Issuers. How can those able to freeze funds react as fast as attackers? That means better tools and more efficient cooperation with law enforcement, but probably also stricter freezing policies, which can degrade the experience of legitimate users.
Screening protocols. How can a protocol screen effectively when it’s reached through aggregators, solvers and multiple hops? One option could be to screen the origin of funds rather than the last sender. Another could be to make screening a condition of integration.
Brands. Many teams screen the apps and APIs they operate while the contracts underneath stay permissionless. When funds move through those open products under a brand’s name, how much responsibility should sit with the team behind it?
Neutral protocols. What level of decentralization is enough for a protocol to defend its neutrality and keep the legitimacy to withstand pressure? The number of independent operators, how evenly stake is spread among them and how many of them are needed to change the rules all play a part, and the bar is likely to rise as the pressure grows.
In the end, on a spectrum where every position comes at a price, the real choice for each protocol is which price it’s willing to pay, and whether it pays it consistently.

On September 24, attackers drained $387.5M from Bitget’s hot wallets. Within a day, part of the funds were already being bridged and swapped across numerous assets, including XRP, ETH, BTC, USDT and USDC, through venues such as THORChain, Uniswap, Chainflip, NEAR Intents, 1inch and LayerZero.

Decentralization, Permissionlessness, Neutrality and Censorship Resistance: DeFi’s Founding Principles Put to the Test

Flare

Flare design starts from the challenge of giving smart contracts reliable access to information that lives outside the chain. This report looks at how that architecture works, where it is already producing results, and what would show it scaling.

How Flare uses data to expand what capital can do

Ethena Staked USDe

Ethena USDe

USDtb

Introduction
Ethena  spent much of the past year proving that a synthetic dollar backed by a derivatives trade can scale. This year it has started proving something else: that the same delta-neutral playbook can be applied to almost any market with a funding rate to earn. That expansion has taken three forms. Ethena pushed USDe into consumer payments and everyday banking. Its Risk Committee approved, and then launched, a version of the core basis trade built on tokenized US stocks rather than crypto. And a public company, StablecoinX Inc., has become one of the largest holders of ENA, with close to a third of the token's total supply on its balance sheet.
That expansion is happening against a mixed backdrop. Ethena's total value locked stands at $5.33 billion today, up 18.2% over the past 30 days but down roughly 63% from the $14.5–15 billion it held a year ago. The 30-day gain is a rebound from an August low near $4.08 billion, not the start of a new growth trend. Over the full year, the picture is a sharp contraction followed by a partial recovery.
The more important number sits underneath that TVL. Over the past 30 days, Ethena's basis trade generated $20.1 million in fees, but only $16,141 of that was kept as protocol revenue. That's a split of roughly 1,250 to 1 in favor of USDe holders and stakers over the protocol itself.
That gap frames everything that follows. Ethena shrank sharply, is now recovering, and has so far chosen to pass almost all of its yield to holders rather than keep it, just as it extends the same trade into new markets. It's also the backdrop for Ethena's recent fee switch vote. This piece works through each part of that expansion in turn: how the core trade works, where it's being extended, the governance question over who captures the upside, and the public company that has turned ENA into a Wall Street position.
The Core Engine: USDe and Basis Trading
How it works
Ethena USDe    is backed by a delta-neutral trade rather than cash reserves or crypto loans. Ethena holds crypto collateral such as BTC, ETH and SOL, and at the same time opens an equal short position in the matching perpetual futures market. The two positions cancel out price risk: if the collateral rises in price, the short loses money, but the gain on the collateral offsets it, and the reverse is also true. What's left is the funding rate, the regular payment that traders on one side of a perp market pay the other to keep the perp price close to the spot price. When markets are bullish or neutral, funding is usually positive, meaning shorts like Ethena get paid. That payment is what funds USDe's yield, and it's the same mechanism Ethena is now taking into new asset classes.
Where the TVL sits
Ethena's $5.33 billion in TVL is split across three products, each doing a different job.
USDe, the core basis-trading product, makes up about 92% of total TVL and accounts for nearly all of the recent growth. Ethena USDtb, which is backed by tokenized real-world assets and offers lower yield for lower risk, is growing much more slowly. A smaller yield wrapper, tsUSDe, is tiny by comparison and shrinking. In other words, almost all of Ethena's growth is flowing through the same basis trade it's now extending into stocks. The TVL base isn't diversified; it's concentrated in the highest-yielding, highest-risk part of the product line.
Where the yield goes
Over the past 30 days, Ethena's tracked fees came to $20.1 million while protocol revenue was just $16,141. "Fees" here means the full yield the basis trade generates, almost all of which is passed straight to USDe holders and sUSDe  stakers. "Revenue" is the small slice the protocol keeps after that. Put simply, Ethena today works more like a yield-distribution engine for its users than a business that earns fees for itself.
That split shows up clearly in valuation. Ethena's price-to-fees ratio is a reasonable 10.4x, but its price-to-revenue ratio is roughly 13,000x. That extreme figure reflects how little revenue the protocol keeps, not necessarily that the token is overvalued. It's also exactly the gap the ENA fee switch proposal, covered below, is meant to address.
New Bets and Latest Proposals
Ethena Pay
Ethena Pay is a self-custodial "neobank" app, a dedicated consumer product rather than a payments feature added onto USDe. It launched in beta on September 1, 2026, using Avalanche as its only settlement network, with an initial rollout across about 48 countries and 400 early-access users, expanding weekly through the month. The app combines savings of up to 6% APY on USDe balances, a card with cashback paid in AVAX (4% on the standard tier, up to 5% on higher tiers), free international transfers and fiat on-ramps in multiple currencies, all inside one self-custodial wallet. The US and EU are excluded from launch while regulatory approval is pending. Ethena Pay Ltd. is incorporated in Malta and says it doesn't hold customer funds; fiat payments run through licensed banking partners, and users keep control of their wallet keys. With the product still in beta, there's no public data yet on adoption or transaction volume, so for now it's a development to watch rather than one we can measure.
Equity basis
Ethena's Risk Committee approved a framework for running the core basis trade on tokenized US stocks: buy a tokenized stock, short the matching stock perpetual, and earn the funding between the two. The trade is live, not just proposed. Binance is the first venue, using its bStocks tokenized-stock product (reported at over $610 million in value, per Token Terminal data cited in coverage) as the long side, hedged with Binance's USDT-settled stock perpetuals.
According to figures Ethena has shared with the press, which we haven't independently verified, the equity basis has averaged about 3.56% annualized over the past six months, and open interest on Binance's stock perpetuals (around $2.9 billion) has grown at roughly 105% per month through 2026. Founder Guy Young called it the most significant expansion of USDe's funding mechanism since launch, pointing to the size of global equity markets as the real opportunity rather than the initial trade size.
Binance has also lowered the auto-deleveraging priority for eligible delta-neutral accounts, including Ethena's. That matters: if one side of a hedged position is forcibly closed during a volatile move while the other stays open, the trade is suddenly exposed to the price risk it was designed to remove. The equity basis gives USDe a second source of yield that should behave differently across market cycles, though for now it's a single trade on a single venue, running alongside the crypto basis trade rather than replacing it.
Converge: the bet that stalled
Not every Ethena initiative has shipped on schedule. Converge, announced in March 2025 as a joint settlement chain with tokenization firm Securitize, was designed as a settlement layer for traditional finance and real-world assets. It would use USDe and USDtb as gas tokens, rely on ENA-staked validators for security, and run on Arbitrum technology with Celestia for data availability. Mainnet was targeted for Q2 2025. It reportedly missed that timeline, and Ethereal, an Ethena-incubated perpetuals exchange that was meant to settle on Converge, reportedly cancelled its own launch by November 2025 as a result. It's a reminder that Ethena's announcements haven't always matched its delivery.
Governance
The main recent governance item is the ENA fee switch proposal, which ran on Snapshot from August 27 to September 2, 2026, with 88 voters. Its timing lines up directly with the fees-vs-revenue gap covered earlier, since a fee switch is the mechanism through which Ethena could start keeping more of the yield it generates. Beyond that, recent governance activity has mostly been routine, such as the Risk Committee's fifth-term election (August 3–9, 2026, with 71 voters).
The DAT: StablecoinX
Holdings and share of supply
StablecoinX Inc. (Nasdaq: USDE) is the main public-market vehicle for ENA exposure, and its position is one of the largest single-entity token holdings among digital asset treasury companies. It holds 3.03 billion ENA, worth $761.9 million, or about 30% of ENA's total supply. That's a striking concentration for a governance token: nearly a third of Ethena's token sits on one company's balance sheet, held as a long-term treasury position rather than for trading.
Trading below its own ENA, but the gap is closing
StablecoinX's realized mNAV, which compares the company's market value with the value of the crypto it holds, stood at 0.61x as of September 25, 2026. In plain terms, the market values StablecoinX at about 61 cents for every dollar of ENA it owns.
That 0.61x is actually the best level in its tracked history, going back to July 2025. The company's mNAV hit a low of 0.036x in September 2025, meaning it was valued at about 3.6% of its crypto holdings, and spent long stretches through late 2025 and into 2026 between 0.08x and 0.18x before recovering. So the more accurate picture isn't a company stuck at a discount, but one that spent more than a year deeply discounted and is now closing the gap without reaching parity. Even at its best point, the market still values StablecoinX below what its ENA is worth. That could reflect doubts about ENA's price, a discount for tokens that can't be freely traded, or broader skepticism about the treasury-company model itself. The data doesn't tell us which matters most.
Risks
The first risk is funding rates, which now matter on two fronts instead of one. Both the crypto and equity basis trades only earn yield when funding is positive. If funding falls or turns negative in crypto or stock perps, fee generation drops directly, and if both markets turn at the same time, the diversification from adding stocks weakens exactly when it's needed most.
The second is thin protocol revenue. At $16,141 over 30 days against $20.1 million in fees, the protocol keeps almost none of the yield it generates. That's a design choice rather than a temporary lag, and it's what the fee switch vote is about.
The third is a possible cut to token incentives. Ethena reportedly plans to reduce USDe-linked token incentives to zero by the end of September 2026. If confirmed, that would remove a source of demand for USDe just as the protocol is trying to grow through Ethena Pay and new basis-trade venues.
The fourth is concentration. With about 30% of ENA's supply held by StablecoinX alone, any forced selling, redemption pressure or change of strategy at that one company could move ENA's price far more than it would with a wider holder base.
The fifth is the valuation of StablecoinX itself. Its mNAV has recovered from 0.036x to 0.61x, but the market still values it below its own treasury. That's a signal about how investors view treasury companies as a category, and it makes raising more capital through the same vehicle harder.
Finally, there's delivery risk. Converge's stalled timeline shows that not every confidently announced Ethena project ships, which is worth remembering when judging newer bets like Ethena Pay and the equity basis.
Takeaway
Ethena's basis-trading engine is what makes everything else in this piece possible. Over the past year it has been through a sharp contraction, with TVL falling from roughly $14.5–15 billion to a low near $4.08 billion in August, followed by a partial recovery to $5.33 billion today. On top of that recovering base, Ethena is running three expansions at once: a consumer payments app on Avalanche to widen USDe's everyday use, an equity basis trade on Binance that diversifies where USDe's yield comes from, and a public company, StablecoinX, that has made ENA directly investable on Wall Street, even if the market still values that position at a discount.
What ties it all together is one tension that runs through every section: the trade generates real and growing yield, but the protocol keeps almost none of it, passing nearly everything to token holders while pushing into new markets. The fee switch vote is where that tension gets decided, and its outcome is arguably the most important open question in Ethena's current direction, more than the payments app, the equity basis trade or StablecoinX's discount. Converge, meanwhile, is a reminder that Ethena's ambition has at least once run ahead of its delivery, which is worth keeping in mind as its newer bets are judged.

Ethena's total value locked stands at $5.33 billion today, up 18.2% over the past 30 days but down roughly 63% from the $14.5–15 billion it held a year ago. Over the full year, the picture is a sharp contraction followed by a partial recovery.

Ethena's Expansion Playbook: Payments, Basis Trading, and Wall Street's ENA Bet

Satoshi Nakamoto subtitled the Bitcoin white paper “A Peer-to-Peer Electronic Cash System.” But most cryptocurrencies have functioned more like tech stocks than cold, hard cash. 
Two products are quickly changing crypto from something you HODL into something you spend: stablecoins and crypto-based credit and debit cards. 
Stablecoins have been around for years. But cards are just now seeing the kind of explosive growth that suggests crypto may yet embed itself in people’s daily lives, as Satoshi had envisioned. 
“We’re seeing [that] stablecoin-linked cards are in hyper-growth-mode,” Cuy Sheffield, Visa’s head of crypto, said on CNBC last month. “There are new issuers coming into our network, launching stablecoin-linked cards every week.”
Since April 2024, money spent using stablecoin cards has grown every month, without fail. In September, people spent a record $1.2 billion in stablecoins using cards, according to paymentscan.xyz. That’s a 203% increase from the same period a year earlier, when people spent just $403 million. 
Transactions and addresses both hit all-time highs last month as well. In fact, though addresses appeared to have plateaued in March, they began rising again in August. Each of the past five weeks saw record highs. 
As more and more people flock to these cards, they’ll need a way to decide which ones fit their lifestyle and budget. They’re in luck. 
DefiLlama’s latest tool, Card Compare, makes it easy to figure out which crypto cards are available in a given country and how those cards stack up.
Consolidating far-flung information 
Card Compare has its origins in the MiCA dashboard that launched earlier this year, according to Patrick Scott, DefiLlama’s chief revenue officer. The dashboard’s unexpected success suggested people weren’t just seeking information that could help them invest — they also wanted information that could help them choose between myriad crypto applications.
That insight led to the development of Find my Exchange, released September 15, and Card Compare, which followed on September 30.
The availability, fees, rewards, and features of crypto cards vary widely between providers and countries. Worse yet, that information is scattered among providers’ websites. Transferring an entire stablecoin balance into fiat and using a standard credit or debit card is far easier than scouring dozens of websites to find and apply for the perfect crypto-based card.
Fees, perks, and more 
Card Compare lets users discover and compare crypto cards in one place.
The dashboard includes cards’ issuers, the regions in which they’re available, the cash back they offer, the annual fees they charge, their foreign exchange fees, and their requirements. It also tracks the perks that cards offer, such as airport lounge access; the ATM fees they charge; the cryptocurrencies they let users spend; their integrations with Apple and Google Pay; and spending limits. 
Providers that want their card to stand out can become sponsors; their cards are featured above the dashboard, allowing providers to display the cards’ key features. 
Take Plasma, one of our featured partners. Its spotlight at the top of the Card Compare page highlights their Platinum card’s 10% cashback on AI and flights and 4% cashback on all other purchases.  
The Plasma Platinum card is, of course, in the Card Compare database itself. The tool notes the card belongs to the Visa network, is issued by Third National Bank, and comes in two versions: a US-available credit card and a globally available spend card. Both feature no foreign exchange markup from Plasma, though Visa foreign exchange fees may apply. Both require that users lock 100,000 XPL for a 12-month period.   
Despite the growing popularity of crypto-based cards, DefiLlama developers found some red flags as they built the Card Compare. For example, they found some companies’ cashback claims misleading.
“Up to 12%” often hides tiny base rates, caps, token requirements, and FX fees that can erase rewards,” one said. 
Another found the process frustrating — with a caveat. 
“I couldn't sign up for anything, [a popular card] gave my data to a fake government authority, it was really tough to find accurate details, sometimes separate docs would conflict,” they said. “I will say for the right user, the setup is pretty seamless — someone in an allowed location can get fully signed up with a ready-to-go card in a few mins.”
To learn more about crypto cards, their regional availability, cashback, fees, requirements, and much more, visit Card Compare today. To become a featured partner, contact us at sales@defillama.com.

The dashboard’s unexpected success suggested people weren’t just seeking information that could help them invest — they also wanted information that could help them choose between myriad crypto applications.

Card Compare dashboard tracks crypto cards’ cashback, fees, and more

Ask a DeFi user about sandwich attacks today and most of them probably won’t be too concerned. They have good reason to think so as most wallets now use private routing while the number of sandwich attacks has collapsed.
However, this overlooks one important point. Private routing is ultimately a fixed layered on top of a problem that still exists at the protocol level. And that fix has already shown cracks, from data leaks to attacks executed through private channels, which means that part of the problem has simply moved from the public mempool to private infrastructure.
Even the Ethereum Foundation is now exploring ways to address toxic MEV directly at the protocol level. THORChain offers an interesting case study: its swap-ordering mechanism has operated for years, making conventional sandwich attacks at the protocol level economically unviable while protecting more than 1.1 million unique wallets that have swapped on the network.
Sandwich attacks, a “market” in decline but still extracting. 
On Ethereum, the share of blocks containing at least one sandwich trade ran near two thirds through 2024. By mid-2026, it sits closer to one fifth.
EigenPhi tracked monthly sandwich extraction falling from nearly $10 million in late 2024 to around $2.5 million by October 2025, with the average attack clearing just over $3. At that level, the gas and priority fees needed to win a position can exceed what the attack returns, and close to 30% of active bots lost money over the period. What followed was a consolidation into a handful of operators, with jaredfromsubway.eth alone accounting for roughly 70% to 85% of Ethereum sandwich-related priority fee spending by mid-2026 (Greenfield Capital, August 2026).
Despite the decline, the sector remains extremely active: sandwiching still represents 35% to 51% of MEV-related priority fees, while the number of wallets being sandwiched on Ethereum remains above 12,000 per week. 
More interestingly, we could assume that sandwich attacks mainly target volatile assets, but the data shows a different picture. In late 2025, EigenPhi reported that roughly a quarter of attacks involved ETH, BTC wrappers and LST pairs, while around 12% involved stablecoins. This has changed further in 2026, with stablecoins now representing more than 50% of activity and major assets and their wrappers accounting for around 15%.
The pattern observed on Ethereum is also visible on other chains. BNB Chain experienced a net decline in the number of sandwich attacks following the introduction of stronger MEV protection at the wallet level. But as activity resumes across the broader market, the share of attacks also starts to increase again. On Solana, sandwich activity fell from more than 15,000 attacks per day to roughly 1,000 between February and June 2025, before climbing back toward 5,000 per day during summer 2026.
The fix moved the attack
There are several reasons that can explain this decline, including the end of the memecoin mania, lower trading activity across the board, and the Fusaka upgrade, which significantly reduced transaction fees. However, there is one trend that cannot be ignored: order flow leaving the public mempool.
Indeed, private order flow reached roughly 92% of Ethereum DEX volume by August 2026, while private transactions represented closer to 30% of broader Ethereum activity only two years earlier. Part of this migration came from users themselves, moving toward solutions such as CoW Swap or manually adding a defensive layer on top of their transactions. But a large part also came directly from the trading infrastructure, with MetaMask, OKX Wallet, Uniswap Wallet and several major BNB Chain wallets taking decisive steps to protect users from sandwich attacks by default.
However, what can be seen as a solution is ultimately a fix that introduces another layer of trust and risk. In April 2023, Flashbots disclosed a flaw in its Protect infrastructure that inadvertently exposed metadata from pending private transactions for a week. A study led by Mancino and Rezzoli later identified 2,932 sandwiches involving transactions that were not observed in the public mempool across two months in late 2024, with roughly 65% traced to a single operator. On October 22, 2025, Jito banned 15 validators linked to sandwich activity after researchers documented multi-slot attacks affecting roughly 222,000 victims and extracting 529,000 SOL over a year.
More recently, in March 2026, CoW Protocol investigated evidence that a transaction submitted through a private RPC may have leaked into the public mempool before inclusion, potentially enabling significant MEV extraction around a $50 million swap.
This leads to an obvious conclusion: the fix is ultimately a stack of intermediaries sitting on top of a base protocol that still permits the attack. In June 2026, the Ethereum Foundation’s chief strategy advisor described toxic MEV as a structural threat to network neutrality and pointed to reducing the ecosystem’s reliance on private order flow as part of the solution. In other words, the longer-term objective is to remove, or at least significantly reduce, the ability to sandwich directly at the protocol level rather than simply hiding transactions from public view.
And that is exactly where THORChain becomes a useful case study. THORChain settles native cross-chain swaps between assets such as BTC, ETH, USDT, USDC, TRX and, soon, XMR without relying on wrapped tokens or external bridges. And all of these trades can happen without sandwich risk, because the way swaps are ordered at the protocol level prevents an attacker from profitably placing one transaction before a victim and another immediately after.
Ordering that can't be bought
THORChain's original design: ordering by slip-fee
A sandwich needs three things in sequence: the attacker's buy, the victim's trade, and the attacker's sell. On Ethereum, that sequence can be purchased because priority fees determine transaction position.
THORChain has never relied on a priority fee auction. Swaps are collected into a queue, scored and sorted before execution, so their position is determined by the economics of the trade itself.
What decides that position is the liquidity fee. THORChain charges a slip-based fee calculated from the size of the trade relative to the pool depth it consumes. It scales with the square of the trade, so a swap twice as large pays roughly four times as much, and the fee accrues to the pool rather than to a block producer.
At the end of each block, queued swaps are scored on that fee and sorted from highest to lowest. An attacker front-running a targeted swap therefore needs a first leg large enough to rank ahead of it and a second leg small enough to rank below it. Selling back less than they bought means sitting on inventory, and both legs pay the slip fee, sized according to the targeted trade rather than the expected profit. 
The combination of fee-based ordering and the slip-based fee is what made sandwich attacks unprofitable, but it came with a trade-off: the quadratic curve can quickly become expensive, which pushed the protocol to rework how swaps execute while keeping the protection intact.
Note: Anyone willing to dig further into the math can check this example and the relevant documentation. 
Fee logic has changed but protection still holds
Streaming swaps were the answer. A large order gets split into smaller sub-swaps, executed a few blocks apart with arbitrageurs rebalancing the pool in between. By doing so, the slip-fee drops by roughly (n-1)/n, where n is the number of sub-swaps. That left pools earning less, which is why a minimum fee was introduced a few months after. It currently sits at 10 basis points, charged on every sub-swap.
None of these changes touched the underlying logic. The sub-swaps are still scored based on the liquidity fee and executed from highest to lowest. Streaming swaps only makes things harder for an attacker, because sandwiching a full order now means front-running and back-running every sub-swap in turn.
Streaming swaps also shrinks the potential value to extract. Sub-swaps exist to limit pool impact, and the protocol enforces a floor on how thin they get. With the current setting, a fully streamed order moves the pool by roughly 5 bps per slice.
The maths then become relatively straightforward. To sandwich a trade, the attacker has to successfully position trades before and after each sub-swap, pay 10 bps per leg and keep a part in inventory to extract around 5 bps of price movement. The profitability simply doesn’t hold.
Swappers should remain cautious, no matter the protection
Private routing has clearly reduced the visible sandwich attack surface but it remains a fix layered on top of a base protocol where the attack is still possible. And as the Flashbots, Jito and CoW cases show, the protections are not absolute.
This is also why the Ethereum Foundation is now looking at toxic MEV at the protocol level rather than relying entirely on private routing. And considering that the assets being targeted are mainly the ones users actually move at scale, including stablecoins, ETH and BTC-related assets, it is worth looking at alternatives in the meantime.
THORChain addressed this problem years ago at the protocol level and offers a robust anti-sandwich mechanism where protection comes from the way swaps are ordered. For users looking to reduce the risk of receiving less value than expected during a swap, THORChain is worth considering.

THORChain offers an interesting case study: its swap-ordering mechanism has operated for years, making conventional sandwich attacks at the protocol level economically unviable while protecting more than 1.1 million unique wallets that have swapped on the network.

Sandwich attacks and the limits of private routing: what THORChain does differently

Firepan

In April 2026, DefiLlama records show Lazarus hit three crypto projects in eighteen days. On April 1, it took $295M from Drift through a hijacked proxy upgrade. On April 14, it drained $100,000 from Zerion's hot wallet after spending weeks working on employees over Telegram, LinkedIn, and Slack. On April 18, it took $293M from Kelp by spoofing a crosschain message.
Only one of those attacks has a public record of AI use, and it was the smallest. DefiLlama's incident record says the Zerion attackers used AI-assisted social engineering to get the employee credentials. It is also the only attacker-side AI case Firepan found after reviewing all 1,236 incidents in DefiLlama's hacks database, which runs from June 2011 to August 2026.
One case in fifteen years says more about what attackers disclose than about what they use. Whether AI could do the work is a more useful thing to measure. Firepan went back to each incident, to the code as it stood before the exploit, and asked whether a current code-analysis agent could have found the vulnerability and built a working exploit.
For onchain DeFi, the answer is about two in three.
Two in three
Of 1,040 onchain DeFi incidents, 68% were rated both AI-discoverable and AI-exploitable. Another 12.9% were partially reproducible: the flaw was visible in code, but the exploit needed something an agent could not supply on its own. The remaining 19.1% sat outside the reach of code analysis entirely.
Across all 1,236 incidents, including centralized exchanges, wallets and chain-level failures, the fully reproducible share falls by only 8 percentage points, to 60%.
Firepan scored each incident on four separate questions:
Could a current agent identify the vulnerability from the pre-hack code and state?
Could it construct the working exploit path, including the transaction or contract?
Would AI materially reduce the skill or time needed to execute, even without acting alone?
Is there evidence AI was used in this specific incident?
The "reproducible" verdict requires a yes on the first two.
Firepan assigned scores by vulnerability class, working from the technique recorded in each DefiLlama incident and falling back to its broader classification, so the same technique always receives the same score. That makes the method auditable and most verdicts unambiguous.
Overall, 79% of the reproducible verdicts carry "High confidence."
Where the money went
Reproducible incidents make up 60% of the dataset, but only 23% of the $20.12B lost. Non-reproducible incidents make up 27% of the dataset and 46% of the money lost.
Reproducible hacks were also smaller, one by one. Measured against DefiLlama's loss figures, the median hack Firepan rated reproducible lost about $545,000. The median non-reproducible hack lost $3.3M, roughly six times more. More than half of all reproducible incidents (57%) lost under $1M.
Losses are also concentrated at the top. The ten largest incidents account for 44% of every dollar in the dataset, and most of them were key and signer failures:
Bybit ($1.4B, signer phishing)
Ronin ($624M, validator keys)
Coincheck ($534M, hot-wallet keys)
Mt. Gox ($470M, private keys)
FTX ($450M, phishing)
The single largest loss on record, LuBian's $3.5B in December 2020, came from weak private key generation. On its own, it makes up more than half of the partially reproducible dollar total.
A contract-reading agent has nothing to read when the vulnerability is a signer who will eventually click the wrong link.
One attacker owns the other column
DefiLlama tags a threat actor wherever attribution is public. 38 incidents are attributed to Lazarus or other DPRK-linked groups. Together they account for $4.61B, or 23% of all losses in the dataset.
Firepan rated two of those 38 reproducible, worth $81M combined. 33 were rated not reproducible and account for $3.88B. A single state-linked actor accounts for 42% of every dollar lost to attacks that code analysis could not have caught.
The most expensive attacker in crypto's history has made most of its money from the people who hold the keys.
What reproducible looks like
Several technique classes were rated reproducible in every recorded instance:
improper access control (137 incidents)
spot-price oracle manipulation (124)
reentrancy (58)
share-accounting errors (42)
arithmetic errors (30)
donation attacks (15)
The large cases in this group read like a list of DeFi's worst weeks. Poly Network lost $611M to an access control flaw in 2021. Cream Finance lost $130M to spot-price manipulation the same year. Mango Markets lost $115M to oracle manipulation in 2022, and Euler lost $197M to a donation attack in 2023. Cetus lost $223M to an arithmetic error in 2025, and Balancer V2 lost $128M to a rounding error in November of that year.
The partial bucket is where 2026 went
Partially reproducible incidents are those where an agent could spot the flaw but couldn't own the full attack.
Bridges and governance dominate the bucket: Nomad ($190M, forged proof), Portal ($326M, signature verification), and Binance Bridge ($570M, proof verifier bug) all needed coordination across chains.
This bucket has grown sharply this year: DefiLlama classified 25 incidents in 2026 through August as bridge and cross-chain attacks, against 3 in all of 2025. Counting every incident involving a bridge or multichain application, the tally runs 38 to 11. Incidents Firepan rated partially reproducible rose from 17 in 2025 to 49 so far in 2026, accounting for about 65% of the $1.3B lost this year.
Drift and Kelp were the two largest losses of 2026 when this dataset closed on August 25. A month later, Bitget lost about $352M, a figure since revised toward $390M, after attackers compromised a backend system that processes its wallet withdrawals and used it to push fraudulent transfers through the exchange's own approval process. Bitget has pointed to suspected North Korean hackers; If that attribution holds, North Korean groups are behind the three largest hacks of 2026, worth close to $1B combined. More on this further in the article.
Same mix, more incidents
If AI were changing which bugs get exploited, the reproducible share would change as well, yet it has not. Among onchain DeFi incidents, it was 76% across 2020 and 2021 and 67% across 2024 through 2026. Code-visible exploit classes have made up most of DeFi's incidents for as long as DeFi has existed.
DefiLlama logged 231 incidents in 2026 through August 25, already more than any full year in the dataset. Monthly incidents climbed from 17 in January to 44 in May, then held at 36 and 38 in June and July. The median loss per incident this year is about $536,000, down from $3.8M in 2021.
More incidents, each one smaller, drawn from the same vulnerability classes. That is the pattern one would expect if finding a known class of bug were getting cheaper and smaller targets were becoming worth the effort. The data cannot prove it, and some of the increase likely reflects better coverage of small incidents.
By Firepan's count, 99.8% of the incidents in DefiLlama's dataset carry no documented evidence of AI involvement. That is an attribution result. Attackers rarely publish their workflows, and an empty field in a database says nothing about what was running on the attacker's machine.
Firepan's other flag runs the opposite direction. Moonwell's February 2026 incident traced back to an AI-co-authored commit that set an incorrect cbETH oracle configuration, and the attacker did not need AI to find or use it. In DefiLlama's records, AI shows up in a defender's commit two months before it shows up in an attacker's playbook.
Two security problems
All three of Lazarus's April attacks started with people. Drift's attackers tricked Security Council members into pre-signing transactions that handed over admin control. Kelp's compromised LayerZero's RPC nodes through social engineering, then exploited a bridge configured to trust a single verifier. Zerion's worked on employees over messaging apps for weeks. Only the smallest of the three has documented AI use, and it cost $100,000. The other two cost $588M.
Bitget followed the same pattern a month after we rounded up and ran this data. On September 24, the exchange lost about $352M, a figure since revised toward $390M. According to CEO Gracy Chen, no private keys were taken. An attacker who had compromised a backend system in Bitget's wallet infrastructure spoofed transaction data, and the exchange's own authorization process approved the transfers. Bitget says the losses were limited to parts of its hot and warm wallets, and that it suspects North Korea. There was no contract bug for an agent to find. The weakness sat in the machinery that decides whether a withdrawal is real, and securing it means auditing the systems around the code as well as the code itself. 
The first problem lives in contract code. It covers roughly two-thirds of onchain DeFi incidents, most of them small, and current agents can already work through it from discovery to exploit. The second lives in keys, signers, and infrastructure. It covers fewer incidents, most of the dollars, and much of one state actor's haul. That ground belongs to key management, signer controls, and penetration testing.
For most of DeFi's history, finding the bugs in the first group took a scarce specialist, and few specialists had reason to spend their time on a $500,000 target. The 2026 numbers, more incidents and smaller losses drawn from the same bug classes, fit what happens when that barrier comes down.
The agents are available to auditors too. Firepan rated 74.8% of historical onchain DeFi incidents AI-discoverable, and 68% fully reproducible, from discovery through a working exploit. Those figures cover the bugs attackers found and used, a narrower set than every flaw sitting in deployed code. Each was readable in the code before the exploit, so the same agent pass, run earlier, would have had the same code in front of it.
On any given contract, the finding goes to whoever runs them first. For a protocol team, perhaps the only way to be first every time is to run them on every code change, before it reaches mainnet. 
Methodology and sources
Incident data comes from DefiLlama's hacks database: 1,236 incidents from June 19, 2011 to August 25, 2026, totaling $20.12B in reported losses. The onchain DeFi subset contains 1,040 incidents and $9.37B in losses. Incident descriptions, techniques, classifications, loss figures, dates and threat actor attributions are DefiLlama's. Comparisons of Firepan's labels against DefiLlama's loss, date and threat actor fields (medians, year-over-year counts, DPRK-attributed losses) are DefiLlama Research's analysis.
Firepan mapped each incident to four capability labels (AI-discoverable, AI-exploitable, AI-amplified, AI-attributable) based on its vulnerability class, and assigned a High, Medium or Low confidence rating reflecting how cleanly the technique maps to a known vulnerability family. Incidents with an unknown technique (48) were rated not reproducible. The reproducible verdict reflects whether a current agent could perform the technical work from pre-hack code and state. It does not claim that AI was used in, or would have prevented, any specific incident.
Dollar figures are nominal USD at the time of each incident as recorded by DefiLlama. Year-to-date 2026 figures run through August 25, 2026.

One case in fifteen years says more about what attackers disclose than about what they use. Whether AI could do the work is a more useful thing to measure. Firepan went back to each incident, to the code as it stood before the exploit.

What percentage of DeFi hacks could AI reproduce today?

Zama is DeFi's universal entry point for shielding, transferring, earning, and swapping digital assets confidentially.

A Look Into Confidential DeFi

Lido DAO

DefiLlama’s Universal Token Ratings dashboard is now a month old. When it first launched, we called it a “rating you have to keep earning.” That’s because the UTR is a living rating system, tracking token performance in real-time to give users an up-to-the-minute assessment of a token’s liquidity, bid-ask spreads, volume, exchange coverage, derivatives conditions, market maker adherence, and more. 
In the weeks since, we’ve also described how the tokens generally score far better on the Performance axis than they do on the equally important Disclosure axis. The latter is an albatross, weighing on the scores of tokens that would otherwise earn stellar ratings. In other words, token issuers’ willingness to be transparent — or lack thereof — serves as a kind of bottleneck. 
But it doesn’t have to be that way. Superior performance is hard to manufacture; transparency is a choice. This week, we want to highlight the way that new data can affect token ratings by taking a close look at Lido’s governance token, LDO. 
As of Tuesday, LDO has seen an 87% month-over-month increase in its UTR score. Here’s why.
LDO in the limelight 
Last week, we called out LDO as an example of a high-profile token with a middling overall score. 
Its performance scores were exemplary, a testament to its large market capitalization and interest among crypto traders.
But it had poor marks on the Disclosure axis. Lido’s blog and documentation provided an incomplete view of its leadership team (7 out of 10) and organizational structure (4 out of 10)  and revealed almost nothing regarding insiders’ wallet addresses (2 out of 10) or allocations to external parties such as investors, exchanges, or influencers (3 out of 10). Nor was it immediately clear who controlled its treasury wallet (2 out of 10).
That meant it earned an overall score of 32.4 and a BBB rating. Not terrible. But not great, either. 
That changed after Lido provided additional information that had been scattered across posts in its governance forum, YouTube videos, social media, employees’ LinkedIn profiles, and more. 
LDO now has an AA rating and an overall score of 59.55, good for second in the UTR dashboard. It was lifted by a stellar Disclosure score of 8.63, the second-highest in the UTR.
Lido furnished executives’ profiles and pointed to its quarterly calls, hosted for the benefit of tokenholders. Those calls offer regular updates on the makeup of its leadership team. Lido also provided the address from which pseudonymous contributors are compensated, which gives observers a detailed, if imperfect, picture of exactly who the organization is currently paying. It also corrected the record regarding its vesting schedules, shared limited information regarding its sole market maker engagement, clarified the manner in which token holders exercise control over the Lido treasury, and provided a detailed breakdown of its relevant legal entities.  
LDO now has a perfect score on eight out of 13 sub-categories within the Disclosure axis, though it still has room to improve in terms of exchange engagement, external allocation, and insider wallet disclosures. 
All of which is to say, this highlights the motivation behind building the UTR. Token ratings systems predate the UTR, though few are as comprehensive. Few are updated so frequently. Moreover, DefiLlama is among the most-visited and most-trusted sources of information within the DeFi ecosystem. The UTR was meant to provide issuers an incentive to do better. We hope that additional teams will come forward, filling out gaps in publicly available (our easily accessible) information, making it easier for investors to make informed decisions as they navigate the wild world of crypto.

Superior performance is hard to manufacture; transparency is a choice. This week, we want to highlight the way that new data can affect token ratings by taking a close look at Lido’s governance token, LDO.

Ratings in Motion

With monthly notional volume jumping from $6.57B in December 2025 to $40.0B in August 2026, more than 6x in nine months, Kalshi’s growth looks spectacular. But this figure hides another reality.
A reality very similar to TVL, which was once DeFi’s default activity metric before the market realized it didn’t always reflect actual usage or revenue. For Kalshi, it all comes down to the word notional.
When a contract settles, Kalshi counts $1 of notional volume per contract, regardless of the price actually paid by the taker. This means that whether a contract changes hands at $0.25 or $0.75, it still contributes $1 to reported notional volume.
Actually, notional reported volume combines two things: the trader’s cash and the collateral posted by the counterparty, which most of the time comes from a market maker rather than another trader. Said differently, notional includes the liquidity supporting the trade, not just the capital actually deployed by traders.
And by looking at trader activity, the picture is clear: in November 2025, every $100 of Kalshi notional carried $47 of trader cash (a standard figure we would expect). By September 2026, it carried just $23.
The ratio has declined throughout the year, and nothing in the data or recent developments suggests it has reached a floor. Is that a problem? Not necessarily. However, relying on notional alone can overstate underlying activity and give investors and users the wrong picture on who is actually leading prediction markets.
Same dollar, six times less cash behind it
The explanation for that decline comes down to product mix. Combos, the multi-leg contracts Kalshi launched in late 2025, grew from just 1.7% of monthly notional in November to 46.5% in August 2026. The trend has only accelerated since, with combos reaching 60% of notional by mid-September, meaning they now account for more volume than single-leg contracts.
However, the two products don't turn cash into notional at the same rate. Indeed, on a single-event contract, trades happen somewhere in the middle of the price range, so the taker pays close to 50 cents for every $1 that gets reported. On the other hand, a combo multiplies several probabilities together, which pushes its price toward zero, since the odds of all the events landing together get smaller with every leg added. Hence, instead of being close to 50 cents for $1, the taker's part drops significantly, reaching 8.3 cents in August. This part, the actual cash brought in by the trader, is called the premium.
TABLE - PREMIUM/NOTIONAL for single and combo
Taking the premium of each product against total notional shows clearly what is happening. The notional represents less and less the money actually traded. The rise of combos into Kalshi's dominant product pushed the ratio down to the point where for every $100 of notional declared, only $23 comes from a trader as of today, against $47 right before combos launched.
Hence Kalshi's notional volume no longer reflects real trading activity the way it did in 2025, and combos inflate the figure enough to make cross-venue comparisons on notional misleading.
Combos don't pay… yet
Using the fee rate, anyone can estimate the fees generated by the taker’s activity. The conclusion everyone reaches is that while the mix is close to 50/50 on notional, the imbalance on fees generated is clear: in August, around 83% came from single markets against only 17% from combos.
(note: fee rates vary with some exceptions here and there, and for simplicity we applied the standard taker rate across the board)
Any analyst would quickly get to the usual conclusion that combos print volume and don't pay for themselves. But that overlooks three things a closer reading would catch.
First, the capital behind combos is more productive. Measured against premium rather than notional, combos have carried the higher take rate in every month since launch, though the gap has narrowed from 1.9 times in February to 1.4 times in August.
Second, Kalshi opened a new monetization stream on August 20 by introducing maker fees on combos at half the taker rate, while makers remain fee-free across most of the exchange. Including those maker fees (not captured in the chart above) raises combos’ share of total fees from around 17% to 24.3% in late August and nearly 32% in September.
Third, combos likely generate flow elsewhere. A market maker holding combo exposure would hedge by trading the individual legs, which are single markets, deepening the liquidity available there and indirectly supporting volume on the other products. This can't be measured from public data, since the hedge may be executed on Kalshi, on another venue, or netted internally against the rest of the book.
Notional is the new TVL, and Kalshi is fine with that
Notional counts contracts and collateral, which makes it the right metric for settlement exposure and the wrong one for measuring actual economic activity. DeFi went through the same phase when TVL was treated as the default measure of activity.
Combos dilute the headline dollar and generate only a fraction of the fees their volume implies. Most analysis stops there, but that misses what Kalshi is actually building.
The strategy is straightforward: build the product that acquires users fastest, accept thin taker fees while it scales, and let the hedging activity it creates deepen liquidity. Then reinforce the economics, which we are starting to see with maker fees on combos (August 20), followed by sub-cent ticks on combos (September 3).
Those two changes push the metrics in opposite directions: volume becomes cheaper to print while combos become more expensive to trade.
Press and analysts may criticise the headline notional volume as “fake,” but Kalshi has little reason to care. What matters more to the business is how effectively it can monetise combos through fees, and Kalshi is currently executing a strategy in that direction.

When a contract settles, Kalshi counts $1 of notional volume per contract, regardless of the price actually paid by the taker. This means that whether a contract changes hands at $0.25 or $0.75, it still contributes $1 to reported notional volume.

Kalshi’s “Fake Volume” hides a real fee engine

Do your own research. 
You hear it often in crypto. It’s an admirable directive, encouraging self-reliance and trust in one’s own judgement. But it can feel impossible given the cacophony of chains, protocols, tokens, and exchanges. 
DefiLlama has made it easy to understand and compare chains, protocols, and tokens. Now, we’re making it easy to understand and compare exchanges. 
Our brand-new Find my Exchange tool lets you see which exchanges list the tokens you own, which exchanges work in your country, the fees they charge, and much more.
Real rankings 
Find my Exchange was inspired by the unexpected popularity of DefiLlama’s MiCA dashboard, which lists the exchanges subject to the European Union’s Markets in Crypto-Assets regulation. 
“There was demand not just to use DefiLlama for research, but really to use it as a discovery platform for different types of crypto apps,” Patrick Scott, DefiLlama’s chief revenue officer, said. 
“What we learned was, people were interested in using it to discover things to use, not just to buy. Historically, most of the site has revolved around metrics [used] to decide whether or not to invest in something.”
Find my Exchange lists more than 50 exchanges, centralized and decentralized. They’re ranked in tiers, with scores based solely on metrics DefiLlama could measure or verify.
DefiLlama isn’t the first to rank exchanges. But most earlier attempts rank exchanges on volume, a self-reported and easily manipulated metric.  
Find my Exchange uses what we’re calling the REAL ranking. It scores exchanges on:
Reserves on chain or in filed accounts
Execution from tick-level fills
Activity, in the form of volume and open interest our own order-book snapshots could support 
Liquidity, in the form of resting depth we observed
The final score is the weighted mean of those four pillars. Reserves and execution each account for 30% of an exchange’s score.  Activity and liquidity both count for 20%.
The S-tier required a rating of at least 85, a bar only Binance cleared due to its exceptional activity and liquidity scores. Hyperliquid was the highest-rated decentralized exchange. It was also the only A-tier DEX. 
For the most part, Find my Exchange does not punish exchanges for metrics we can’t measure. For example, if we couldn’t measure an exchange’s reserves, we didn’t give it a zero in that category. Instead, its weight was redistributed over the pillars we did measure.
Traders might find the execution data especially useful. Among other things, it considers the all-in cost of a $10,000 order at the exchange’s base fee tier and actual slippage at fixed order sizes.
On-ramps and off-ramps 
Find my Exchange also provides valuable data on the most basic aspects of exchanges’ operations, including the cost of converting fiat to Bitcoin or Bitcoin to fiat; the rails the exchanges rely on; the fiat currencies they accept; and the types of documents and verification steps they require.
Take Coinbase. In the US, converting $1,000 to Bitcoin costs $7.62. Converting $1,000 in Bitcoin to USD costs $6.11. On-ramp rails include ACH, FedWire, SWIFT, and cards. To use the exchange, you need to provide a government ID and a selfie. 
Maybe you don’t live in the US, though. Maybe you live in Germany. There, a $1,000 fiat-to-Bitcoin round-trip on Coinbase will cost $26 (almost double what it costs in the US). On-ramp options are limited to SEPA and cards, and using the exchange will require a proof of address in addition to a selfie and one’s government ID. 
Maybe you find that a bit expensive or burdensome. In Germany, that same round-trip will cost almost $8 less on Kraken, which accepts additional fiat currencies and requires fewer documents to use. 
Maybe you’re just interested in a single token. A token finder tool lets you find the exchanges that list that token, filtered by country. 
Find my Exchange is now the home for some preexisting DefiLlama dashboards, including centralized exchange reserves and exchange’s markets.

Find my Exchange lists more than 50 exchanges, centralized and decentralized. They’re ranked in tiers, with scores based solely on metrics DefiLlama could measure or verify. DefiLlama Find my Exchange uses what we’re calling the REAL ranking.

Our brand-new Find my Exchange tool lets you see which exchanges list the tokens you own, which exchanges work in your country, the fees they charge, and much more.

There’s only one A-Tier decentralized exchange

With the launch of MotoSwap on Ethereum, one important question is whether the platform can build meaningful and lasting market share in an already competitive DeFi environment.
We take a closer look at some of the most notable attempts by DeFi protocols to attract liquidity and users from established competitors, often called ‘vampire attacks’ in DeFi, and assess which strategies succeeded and resulted in sustained adoption. We then examine MotoSwap’s approach and how its design compares with the lessons from these previous launches.
How MotoSwap plans to bootstrap liquidity
After 18 months on Bitcoin L1, MotoSwap launched on Ethereum Monday, generating over $6 million in volume and adding $5.7 million TVL in the first 24 hours. MotoSwap will use incentives to encourage Uniswap V2 LPs, which currently hold $790 million in TVL, to participate in this new EVM-based DEX.
Liquidity mining and LP migration
The core of MotoSwap’s liquidity bootstrapping campaign is a two-step acquisition strategy. First, it targets Uniswap V2 LPs with 30 million MOTO (3% of supply) and will distribute these tokens over a 2-week blitz campaign. During this period, LPs can stake their existing Uniswap V2 LP tokens into the MotoSwap contract to earn MOTO rewards alongside their Uniswap V2 fees.
This lowers the friction to participate because LPs do not need to remove liquidity from Uniswap immediately, and can instead keep their existing position intact while testing the incentive program. After the 14-day campaign, MotoSwap will convert these LPs into MotoSwap-native liquidity through a one-click migration process that moves their positions into the same pair on MotoSwap.
The second phase aims to retain that liquidity. After migration, MotoSwap will distribute an additional 70 million MOTO over 90 days through liquidity mining on MotoSwap-native pools. In total, MotoSwap allocates 10% of the supply to liquidity acquisition and retention: 3% to attract Uniswap V2 LPs, and 7% to retain them post-migration.
This strategy first attracts liquidity on Uniswap V2, simplifies migration, and then offers LPs a larger incentive to remain on MotoSwap.
MotoSwap’s fee structure adds another retention layer. Where Uniswap V2 charges a 0.3% swap fee, with 0.25% going to LPs, MotoSwap plans to charge 1%, with 0.3% to LPs. The remaining fee is split between MOTO stakers, trader rakeback, buyback and burn, the treasury, and creator fees.
The trader rakeback is one of the more novel parts of the model. MotoSwap returns 0.2% of every swap to the trader who generated the volume, distributing rewards through weekly epochs. That turns traders from pure fee payers into fee loop participants. For example, TRUMP launched in January 2025 and cleared roughly $120 billion in volume that month alone. Run that through MotoSwap's model and about $240 million would have gone back to the wallets that traded it.
This can also support retention: traders must return to MotoSwap to claim their rakeback, creating another touchpoint with the DEX after the initial swap.
Modernizing the token flywheel
The same fee structure also feeds into a broader MOTO flywheel. Beyond trader rakeback, MotoSwap combines buybacks and burns with direct revenue distribution to MOTO stakers.
This differs from pure buyback models, where protocol revenue can end up absorbing tokens from unlocks or sold by early investors or insiders. Indeed, MotoSwap’s design creates market demand for MOTO and directs part of the value directly to users who remain staked.
Moto.fun adds another layer to that design. Upon bonding, the curve triggers MOTO buying to create MOTO/ETH liquidity, generating additional trading and arbitrage opportunities. The same bonding event also triggers MOTO buy-and-burn, and under the current parameters, a token reaching roughly $1 million in market cap results in around $50,000 worth of MOTO being bought and burned.
This creates a broader flywheel: more launches drive MOTO demand and burns, new pairs generate more trading activity, and that activity produces fees that flow back to stakers and further buybacks.
What historical vampire attacks show
MotoSwap is not the first DEX to use token incentives to attract liquidity and users from an established platform. Comparable strategies have appeared throughout DeFi, most notably through vampire attacks that directly targeted the liquidity or users of incumbent protocols.
These earlier examples provide a useful benchmark for MotoSwap’s liquidity bootstrapping strategy. We compare eight notable vampire attacks to assess which approaches translated initial incentives into lasting market share and which failed once those incentives began to fade.
We compare each attacker’s position at the time of the attack with its position three months later, with one exception that uses a different evaluation window. Protocols that captured more than 50% of the incumbent’s market were classified as successful, while those that failed to do so were classified as failures.
Of the eight vampire attacks analyzed across DEXs and NFT marketplaces, four succeeded and four failed. This shows that launching a vampire attack does not guarantee success, but it can provide a meaningful opportunity to grow relative to the incumbent in terms of TVL or trading volume.
Attacker
Incumbent
Launched
Evaluation Window
Position at Attack
Position at Window End
Outcome
SushiSwap
Uniswap
Sep 2020
3 months
68.4% of incumbent TVL
60.9% of incumbent TVL
Succeeded
Swerve Finance
Curve
Sep 2020
3 months
30% of incumbent TVL
0.5% of incumbent TVL
Failed
Trader Joe (LFJ)
Pangolin
Aug 2021
3 months
14.7% of incumbent TVL
531.9% of incumbent TVL
Succeeded
0xDAO
veDAO
Jan 2022
2-day governance snapshot
37.7% of incumbent TVL
550% of incumbent TVL
Succeeded
LooksRare
OpenSea
Jan 2022
3 months
10% of total NFT volume
6.9% of total NFT volume
Failed
X2Y2
OpenSea
Feb 2022
3 months
4% of Total NFT volume
9% of total NFT volume
Failed
KLEX Finance
KlaySwap
Aug 2022
3 months
8.9% of incumbent TVL
6.0% of incumbent TVL
Failed
Blur
OpenSea
Oct 2022
3 months
11.5% of incumbent volume
101.6% of incumbent volume
Succeeded
DEXs
The most well-known DEX vampire attack was SushiSwap’s attack on Uniswap. SushiSwap launched in August 2020 with SUSHI incentives for users who deposited Uniswap LP tokens, then migrated that liquidity to SushiSwap on 9 September 2020. The attack was initially successful, pulling roughly $1.08 billion of liquidity from Uniswap and briefly allowing SushiSwap’s TVL to surpass Uniswap’s.
However, Uniswap responded by announcing its UNI airdrop. Every Ethereum address that had interacted with Uniswap V1 or V2 contracts received 400 UNI. The launch of UNI also allowed Uniswap to introduce token incentives of its own, helping it win back market share.
After 3 months, SushiSwap was still able to hold 60.9% of Uniswap’s TVL. This made the vampire attack clearly a success for SushiSwap, as it briefly overtook Uniswap and then settled into a comfortable position.
Shortly after SushiSwap’s launch, Swerve Finance attempted a vampire attack on Curve by launching a fork of the protocol and distributing its SWRV tokens through liquidity mining. Initially, this appeared successful. At launch, Swerve captured TVL equal to 30% of Curve’s TVL and at one point reached $800 million in TVL. 
Swerve ultimately failed because its incentives dried up, and the protocol did not offer a differentiated product. After three months, its TVL had fallen to just 0.5% of Curve’s TVL, ultimately an unsuccessful vampire attack.
Last, 0xDAO represents a vampire attack centered around a specific event. Solidly was Andre Cronje’s ve(3,3) DEX on Fantom, and a snapshot allocated its initial governance power to the protocols with the highest TVL.
veDAO launched first, offering aggressive APYs to users who deposited liquidity into its farms. It quickly grew to become the protocol with the highest TVL. 0xDAO was then launched by a coalition of Fantom-native protocols, including SpookySwap, Scream, Liquid Driver and Tarot, which sought to pull liquidity back because they did not want veDAO to win the snapshot.
Conclusion
The main lesson from previous vampire attacks is that incentives alone are not enough to produce lasting market share. The strongest examples paired rewards with mechanisms that encouraged users to remain active after the initial incentive period.
MotoSwap has incorporated several of these lessons into its liquidity bootstrapping strategy. Its LP incentives and one-click migration process resemble SushiSwap’s approach to attracting Uniswap liquidity, while its airdrop and points program more closely mirror Blur’s strategy of rewarding the users and activity that matter most to the platform. By targeting both LPs and traders, MotoSwap is attempting to build liquidity and trading activity at the same time instead of relying on liquidity mining alone.
This also means MotoSwap does not need to capture the majority of Uniswap V2’s liquidity for the campaign to be meaningful. With Uniswap V2 currently holding around $790 million in TVL, attracting just 10% would already provide MotoSwap with roughly $79 million in liquidity and establish a substantial base from which to build trading activity.
For MotoSwap to be considered successful, we would want to see a high share of staked Uniswap V2 LP tokens actually migrate to MotoSwap after the 14-day campaign, and for that liquidity to remain after the 90-day liquidity mining program starts to mature. More importantly, that liquidity needs to become productive, with sustained trading volume and meaningful fee generation without relying entirely on emissions.
The main risk is that MotoSwap attracts mercenary LPs and incentive farmers who leave once rewards decline. Its 1% swap fee might seem high to cost-conscious traders, but it is comparable to fees charged by traditional in-wallet swaps and certain aggregators. Pons, for example, applies a 1% standard trading fee to tokens launched on Robinhood Chain.
The difference is where those fees go. MotoSwap redistributes part of them to traders through rakeback, while also supporting MOTO stakers and buybacks. The real test will therefore be whether MotoSwap can turn its initial attack into persistent liquidity, real trading volume, and recurring user activity.

With the launch of MotoSwap on Ethereum one important question is whether the platform can build meaningful and lasting market share in an already competitive DeFi environment. We examine MotoSwap’s approach and how its design compares with the lessons from these previous launches.

MotoSwap’s playbook for building liquidity and trading activity

Meteora

Uniswap

Three weeks have passed since we released Universal Token Ratings. The UTR, built in partnership with Forgd, is meant to give investors a simple way to evaluate tokens in real-time, with up-to-the-minute data.
So far, 168 tokens have been scored by the UTR. Scores range from an abysmal 5.6 out of 100 to 61.5. Tokens are also given a familiar, Moody’s-style letter grade. As of Tuesday, 26 tokens earned a CCC. Only two tokens earned a AAA. 
But the UTR isn’t just for investors. It’s also for token issuers. Ideally, it will motivate them to do better, disclose more, and work harder to ensure that their tokens are deeply liquid. 
That begs the question: how, exactly, do tokens earn top marks? And what can the issuers of middling tokens do to improve their score? 
In this Spotlight, we examine Meteora’s MET token, which earned a AAA letter grade and an overall score of 60.2, placing it second in our database. 
To see our earlier reports on the UTR, see here and here. 
Meteora: a case study  
Tokens are scored on two axes, Disclosure and Performance. Scores for each axis range from 0 to 10, and a token’s overall score is the product of their Disclosure and Performance scores — not an average, and not a percentage. In other words, if a token scores 7 out of 10 on both axes, its overall score is 49 out of 100, not 70.
Each axis is comprised of categories that are further divided into sub-categories. Tokens earn a score between 0 and 10 on 32 separate sub-categories, such as whether token issuers have disclosed the identities of insiders or published their vesting schedules. Performance-related sub-categories consider a token’s volume, bid/ask spreads, recent price performance, open interest, exchange coverage, market capitalization, and more. 
Here’s the key: each sub-category is weighted. The Disclosure axis features 13, but they aren’t created equal. Identifying the people who control relevant multi-signature wallets is worth 10% of a token’s Disclosure score. Transparency regarding market maker engagement, deals with exchanges, and allocations to external parties each count for 9%. 
Meteora earned the fourth-highest Disclosure score in the UTR, just edging Uniswap’s UNI token — the only other token to earn a AAA rating. Interestingly, if the overall Disclosure score was the sum of its sub-categories, UNI would have topped Meteora here. But that isn’t how the UTR works. The sub-categories are weighted, and Meteora earned top marks where it mattered. 
Consider how each token scored on its disclosure of multi-signature wallets. Meteora scored an 8 out of 10, meaning it had provided the address of a primary treasury wallet, confirmed that it was controlled by a committee of signers, and detailed how many of those signers would need to agree in order to withdraw funds. Meteora had also disclosed partial information about secondary multisig wallets. It did not, however, disclose signers’ roles or identities. 
Uniswap, meanwhile, scored a 6 out of 10. That means it had shared the address of a primary wallet and its signer threshold but did not disclose secondary wallets.
Meteora also scored a perfect 10 out of 10 on exchange agreement disclosures. It had disclosed which centralized and decentralized exchanges it was listed on and it disclosed whether any listing fees, token allocations, or incentive arrangements were part of the listing agreement. 
Uniswap’s 5 out of 10 in this sub-category means that it had listed its major trading venues but offered no comment on its listing arrangements. 
On the Performance axis, Meteora scored highly in four of the five most important sub-categories: top-tier exchange coverage, trade volume as a percentage of market capitalization, order book depth as a percentage of market capitalization, fully-diluted valuation, and the market cap to FDV ratio.
Exchange coverage is the most important, accounting for 15% of a token’s Performance score. Meteora trades on Binance, OKX, Coinbase, and Upbit, and earned a perfect 10 out of 10. 
It also earned perfect marks on trade volume and order book depth. Meteora’s token saw $53 million in average daily trade volume over the past month, roughly a quarter of its market capitalization. That put it well above the threshold for a perfect score: trade volume as 10% of market cap. 
But it lost points on fully-diluted valuation. The value of all Meteora tokens is just $204 million, earning it a 5 out of 10. Only projects with an FDV over $1 billion earned perfect marks. 
How to improve 
Our previous research found that Disclosure is the true bottleneck. For three out of every four rated tokens, the Disclosure score is lower than the Performance score. Rated tokens’ mean Disclosure score sits at 4.98 out of 10. Their mean Performance score sits at 6.07.
For most issuers, that means doubling down on transparency is the easiest thing they can do to improve their tokens’ scores.
Let’s take the example of two high-profile tokens with middle-of-the-pack scores: Ethena’s ENA and Lido’s LDO, which earned overall scores of 34 and 32.4, respectively. Both earned BBB ratings. 
Their performance scores were exemplary, a testament to their large market capitalizations and interest among crypto traders. They are, after all, governance tokens that give holders a say in the management of multibillion-dollar protocols. 
But Ethena is not, as it touts on its website, an example of “unparalleled transparency.” Specifically, it has disclosed little about its core contributors. It earned a 3 out of 10 on the “insider identities’” sub-category. That means that most of the people who built Ethena and who continue to exert outsize influence over it are pseudonymous. Its founder, Guy Young, is well-known in the industry. Beyond that, however, Ethena doesn’t appear to offer a detailed org chart identifying all founders, key developers, and executives. 
Lido fares even worse despite its status as the world’s largest crypto protocol. Most of its insiders have been identified (it earned a 7 out of 10 in that sub-category) but its organizational structure is poorly defined (4 out of 10) and it has not divulged insiders’ wallet addresses (2 out of 10). It has shared little information on who controls its treasury wallet (2 out of 10) and it has not disclosed allocations to external parties such as investors, exchanges, or influencers, beyond vague references to the manner in which it raised capital (3 out of 10). 
For tokens that want to boost their Performance scores, the task is more daunting: every single founder wants to manufacture investor interest and, in turn, high market capitalization and trade volume. Very few succeed. 
Issuers would do best to focus on securing a listing on a top-tier exchange. It is the single most valuable thing an issuer can do to boost its Performance score. Unlike other metrics in the UTR’s Performance axis.

The UTR isn’t just for investors. It’s also for token issuers. Ideally, it will motivate them to do better, disclose more, and work harder to ensure that their tokens are deeply liquid. In this Spotlight, we examine Meteora’s MET token, which earned a AAA letter grade and an overall score of 60.2.

In this Spotlight, we examine Meteora’s MET token, which earned a AAA letter grade and an overall score of 60.2, placing it second in our database.

Anatomy of a Highly Rated Token

Morpho's fully diluted value sits 27% above Aave's. It holds less than half the deposits, generates half the fees, and pays out nothing, because the fee switch written into its contract has never been flipped.
That gap is a bet on growth and we are testing this bet against the three things that would have to show up in the data: liquidity, fees, and revenue. 
Our findings? Only the first one holds.
The scoreboard
Comparing both projects on key metrics gives us a straightforward picture.
On valuation, Aave is larger by 5% on circulating market cap, but the situation completely reverses once you account for the tokens still to be unlocked. Morpho has only 65% of its supply circulating, meaning the protocol is 40% larger than Aave on FDV.
On liquidity, Aave holds more than twice the deposits and shows a higher utilization ratio, 43.5% against 34.8% for Morpho. Yet the market prices Morpho nearly 3x higher per dollar of deposits on FDV, 17.5% against 6.5%.
On financials, Morpho earns half of Aave's fees on 47% of the liquidity, so Morpho’s book monetizes well. But those fees never become revenue: Morpho doesn't take a share (fee switch not activated as of today), while Aave collects 13.2%.
All of this brings us to the ratios that matter. Market cap to revenue can't be compared today, since Morpho has no revenue, though we'll project it later in this piece. On price to fees, where both can be measured, Morpho trades at 1.7x Aave on market cap and 2.5x on FDV. That's a clear premium, and it's a bet on growth. But is that growth actually materializing this year?
Fact 1- Deposits back the bet
The two books have moved in opposite directions since Q4 2025. Morpho sits at 154 and Aave at 67, meaning Morpho's real deposit base has grown 54% while Aave's has shrunk by a third.
Morpho's growth is also accelerating. Month over month since May: +0.4%, +2.3%, +7.0%, then +10.5% in August, the strongest month of the year.
Aave held above its Q4 base through February, peaking at 109, then fell for three straight months: 102 in March, 86 in April, and 59 in May, pulled down by the April 18 KelpDAO rsETH exploit and the withdrawals that followed. What matters here is that Aave hasn't recovered: it's 13% off the May low and still a third below where it started.
So the first leg of the bet holds. Morpho is taking real deposits at an increasing rate, while its largest competitor is smaller than it was nine months ago.
Fact 2 - Fees don’t follow
 
Morpho's fees sit below their Q4 2025 level and have been choppy all year, ranging from 57 to 95 and averaging 80. Nothing there qualifies as growth.
Morpho does hold up well relative to Aave, whose fee base has collapsed from 86 in January to 34 in August, most of that fall coming after the April hack. That gap could look like a reason for the premium but it would be based on competition underperforming rather than growth.
Another way to measure relative growth is how much in fees each dollar of deposits produces. Here the result is striking: the two protocols are broadly the same, 58 for Morpho and 51 for Aave, and both sit near half their Q4 2025 level.
Morpho gathered deposits and held its fees, Aave lost both, and yet each deposited dollar earns the same on either side. That reflects the structure of lending markets, where rates are driven largely by system-wide borrowing demand rather than by the protocol itself. So if the premium is meant to reflect growth in productivity per dollar deposited, the bet doesn’t hold: Morpho has gone backwards on that measure, and only marginally above Aave.
Fact 3 - Even the maximum switch falls short
Morpho currently returns all borrower interest to lenders, but the Morpho Blue contract allows governance to retain up to 25%. That creates a useful way to test the valuation: what would Morpho’s revenue multiple look like if it monetized at Aave’s current take rate, or pushed the fee switch to its maximum?
At Aave’s 13.2% take rate, Morpho would trade at 63x revenue on market cap and 96x on FDV, 1.8x and 2.6x Aave’s respective multiples of 35.9x and 37.2x. In other words the valuation gap would still be there.
At the 25% maximum, the gap narrows sharply. Morpho would trade at 32.8x on market cap, slightly below Aave’s 35.9x. On FDV, however, it would still trade at 50.0x versus Aave’s 37.2x, a premium of roughly one-third.
But this remains a theoretical exercise. Morpho’s circulating market cap will rise as tokens unlock, and neither its fee base nor TVL will necessarily be the same when a fee switch is activated. More importantly, monetization itself could affect the economics of the protocol. Morpho currently passes 100% of borrower interest to lenders; retaining up to 25% would reduce lender returns and could, in turn, affect deposits.
A premium already narrowing?
Across the three metrics, the picture is mixed. On liquidity, the bet has worked: Morpho grew while Aave contracted. On fees, the case is weaker: Morpho monetizes deposits slightly better, but fee growth has stalled and productivity per dollar remains close to Aave’s. On revenue, the maximum fee switch is needed to close the gap on market cap only, and not FDV.
For now, Morpho’s premium is therefore clearly supported by deposit growth, but much less so by the economics that sit on top of it.
And the market may already be adjusting. Morpho continues to trade at a premium, particularly following the Apollo announcement, but that premium has narrowed materially. On market cap, the fee premium fell from 2.9x to 1.3x, while the deposit premium declined from 2.8x to 1.8x. 
The key takeaway is that Morpho has earned its growth premium on deposits. Keeping that premium requires those deposits to translate into fees, and eventually into revenue. Until then, the valuation remains a speculative bet on future monetization and the market may already be starting to reprice that bet.

Morpho's fully diluted value sits 27% above Aave's. It holds less than half the deposits, generates half the fees, and pays out nothing, because the fee switch written into its contract has never been flipped. That gap is a bet on growth and we are testing this bet against the three things that wou

Morpho's fully diluted value sits 27% above Aave's. It holds less than half the deposits, generates half the fees, and pays out nothing, because the fee switch written into its contract has never been flipped. That gap is a bet on growth.

Morpho Trades Above Aave on Zero Revenue. A Rational Bet on Growth?

Solana

Dogecoin

The tokenized-equity conversation over the past year has been one of magnificent growth. Stock, ETF, and equity-index perpetual open interest is up ~78x since DefiLlama first captured reliable tracking in November 2025, from $114M to a peak of $9.99B on August 18, 2026.
As of the most recently completed trading day, DefiLlama tracks $31.6B in 24h volume and $14.6B in open interest across 3,132 RWA-perp pairs on 32 venues. Stock, ETF, and equity-index perps specifically — the segment this piece focuses on — account for $15.6B of that volume (49.2%) and $8.5B of that open interest (57.9%) , making equities the single largest slice of the RWA-perp market, ahead of commodities, forex, and private-equity/venture perps combined.
For context on the other side of the market: on-chain tokenized equities (the spot, custody-backed side — Ondo Global Markets, Backed/xStocks, bTech/bStocks, and others) currently carry $3.79B in active on-chain market cap across 1,639 tracked assets . Stock-perp open interest alone is now roughly 2.2x larger than the entire on-chain tokenized-equity market  — a pattern we covered in our earlier piece, Stocks On The Blockchain - Do They Have Product-Market Fit?
Where Gate Fits In
Adoption of stock perps on the CEXes has mostly been a listings story: how many stocks does an issuer offer, how fast is on-chain market cap growing, which venue has the broadest catalogue. 
Gate  has listed the most tokenized equity perpetuals, by far - standing at 385, beyond all the other exchanges. This is impressive, and we want to go one layer deeper to diagnose whether a venue can actually support leveraged trading in these instruments at scale, without the mechanical failure modes (bad fills, funding-rate whipsaws, cascading liquidations) that kill adoption in thin markets?
We're drawing on two independent data sources to do that:
Exchange order-book depth data supplied by Gate and independently re-verified against the complete Aug 25–Sep 7, 2026 window — covering crypto spot and perpetual pairs (BTC, ETH, SOL, XRP, DOGE) and equity-linked perpetuals across six venues (Gate, Binance, Bitget, Bybit, OKX, Hyperliquid), on an RPI basis. Binance spot has no RPI book, so its public order book is used instead. This is vendor-supplied; DefiLlama does not independently track order-book depth or bid/ask spreads for these instruments, so these figures are disclosed as exchange-reported.
DefiLlama's own RWA-perp tracking — volume, open interest, and funding-rate data for perpetual contracts referencing stocks, ETFs, and equity indices, current through today, across every venue we track.
As such, we will be going into the weeds to look at how Gate handles their flow, given they have the most number of perpetuals listed.
Gate — Mechanism, Core Lquidity, and RWA Footprint
While Gate isn't necessarily the largest stock-perp market by volume, it's the broadest equity-perp catalogue we found in this data — 385 listed contracts, more than any of the other five venues compared here — which makes it the highest-stakes test of the thesis above: a wide catalogue is exactly what "listings-story" adoption looks like, so if depth doesn't hold up behind it, that's where it would show.
Why Depth Matters More For Perps Than Spot
A thin spot market mostly costs you a bad fill. A thin perp market compounds that: shallow books widen the gap between mark price and where you can actually enter or exit, which pushes funding rates around more violently as small trades move the mark, which in turn increases the odds that a leveraged position gets liquidated on a move that a deeper book would have absorbed. 
For a still-young product category like stock perps, that failure chain — bad fills, funding-rate whipsaws, cascading liquidations — is exactly what determines whether traders come back after their first large order. 
Liquidity Evidence
Gate  lists 385 distinct equity-linked perpetual contracts as of September 7, 2026 — the most of the six venues in this comparison, ahead of Bitget  (275), Bybit  (227), Binance  (180), OKX  (168), and Hyperliquid  (105) — and added the most new listings over the over the preceding 10 days (+12, vs +8 at Bybit and OKX),
Breadth alone proves availability, not tradability. So we checked the order book on the five most-traded names — SNDK, SKHYNIX, SPCX, SOXL, and MU, which together account for 60% of tracked equity-perp volume, averaged across the fourteen daily samples — at every depth band from top-of-book out to 2%.
At the top of the book, it isn't close: Gate has the deepest resting size on four of the top five contracts by volume — SNDK, SKHYNIX, SOXL, and MU. The exception is SPCX, where Bybit's top-of-book ($90.6K) edges out Gate's ($80.3K).
Moving one band out to the 0.1% level, Gate does sweep all five names, with margins ranging from 5% to 28%.
That lead isn't a single-day artifact. Tracking SOXL's and MU's 1%-depth daily over the full Aug 25–Sep 7 window, Gate held the top spot every single day on both names (14/14), with its lead never dropping below 19% on SOXL and 7% on MU. SNDK is more mixed by comparison — Gate led on only 10 of the 14 days, with its edge thinning noticeably in the back half of the window.
Moving out to 1% depth, the picture stays strong but gets more textured: Gate is deepest on four of the five names, losing only SPCX, where Bitget's 1% depth ($25.0M) runs roughly double Gate's ($10.6M).
Looking at the combined top-5 book across every band tells the fuller story. Gate leads clearly at the extremes — top-of-book and the wide 2% band — but the middle of the book is level rather than Gate-led: at the 0.5% mark, Bitget is ahead on three of the five names, with the two of them tied within 1% of each other on the remaining two. That's a real, disclosed pattern, not a rounding error — Gate isn't uniformly deepest through the whole book, it's deepest at the extremes.
Spreads confirm the same "strong but not uniformly best" pattern: on median bid-ask spread across the five most actively-traded contracts, Gate sits at 0.141 basis points — tighter than Bybit (0.187), Bitget (0.348), and Hyperliquid (1.096), but looser than Binance (0.106) and OKX (0.106), which are effectively tied for tightest.
If this were the only evidence, it would be a solid but mixed story. It reads more convincingly once you check whether the same firm's other order books look the same way. They do:
Across Gate's core crypto spot markets (Aug 25–Sep 7 average), Gate has the deepest top-of-book on four of the five major assets we checked — BTC ($237.0K), ETH ($117.4K), XRP ($19.5K), and DOGE ($10.0K) — ahead of Binance, Bybit, OKX, and Hyperliquid on each of those. SOL is now the exception: OKX holds the deepest top-of-book at $57.0K, ahead of Gate's $30.4K. The margin varies: comfortable on ETH (1.58x the next-best book, OKX) and DOGE (1.52x), thinner on BTC (1.27x); XRP sits at 1.51x.
The same shows up on the perp side: with Bitget added to the comparison set, Gate has the deepest top-of-book on all five majors — BTC, ETH, SOL, XRP, and DOGE — ahead of Binance, Bitget, Bybit, OKX, and Hyperliquid on every one. At 1% depth the picture is more mixed: Gate holds the deepest book on BTC and SOL, while Bitget edges ahead on ETH, XRP and DOGE. This isn't offered as proof that Gate's equity books are deep — the equity data above already establishes that directly. It's evidence that the equity-book result is a general trait of how Gate runs its order books, showing up independently in a second, unrelated product line, rather than a one-off snapshot.
Rwa-Perp Footprint And Stability
On the full equity-perp universe, across the 93 equity perpetuals tracked over the 14-day average, Gate ranks 3rd of 6 on 24h volume ($1.15B, against Binance's $8.34B) and 3rd on open interest ($738M, against Binance's $2.51B and Hyperliquid's $1.69B). Breadth and depth are not the same thing as scale — Gate isn't the biggest book here.
Narrowing to DefiLlama's own RWA-perp tracking, restricted to the stock/ETF/index scope specifically, Gate accounts for roughly $888.5M in daily volume and $778.1M in open interest on the most recent complete day.
However, despite Gate losing out in market size, it makes up for in funding-rate volatility. Across 87 stock tickers listed identically on Gate, Binance, Bybit, and OKX, Gate's funding-rate volatility over the trailing 30 days (0.00245 percentage-point standard deviation) is a near-dead-heat with OKX (0.00248) — and meaningfully tighter than Binance (0.00478, ~1.95x Gate's) and Bybit (0.01132, ~4.6x Gate's). That gap holds despite Binance carrying roughly 2.9x Gate's average open interest on those same 87 names over the period — bigger books, less stable funding. 
Put together, the picture that emerges isn't "Gate is the deepest book everywhere." It's more specific than that, and more useful: Gate leads the top of the book on four of its five most-traded equity contracts and sweeps all five once you move out to the 0.1% band, leads at the wide 2% end of the book too, and shows the identical pattern on its core crypto markets — but the middle of the book (0.5% depth) is level rather than Gate-led, and it sits mid-pack on raw spreads. Scale tells a third story again: Gate isn't the largest stock-perp venue by volume or open interest, trailing Binance by a wide margin on both.
What ties those three uneven results together is the one number that isn't a snapshot of order-book size at all: funding-rate volatility on identical instruments. There, Gate is essentially tied with OKX for the steadiest funding in the group, and meaningfully steadier than Binance and Bybit — despite running a fraction of Binance's open interest. Depth at the right points in the book, not raw size, is what keeps a leveraged market's funding from whipsawing. It also shows its not always about who lists the most, but also, who can actually hold a book together under size. 
Comparisons & Gate’s Product Positioning
Before comparing venues head-to-head, it's worth being upfront about what's actually being compared. Trading in this segment is concentrated: across the 93 equity perpetuals tracked in the vendor data, the five largest contracts account for 60% of volume, with SNDK alone at 28.5%. That means most of what separates one venue's book from another's — in the depth comparisons above and the tables below — comes down to how each handles a small, identifiable set of names, not a broad average across the full catalogue.
Vendor equity-perp order-book comparison (Aug 25–Sep 7 average; listings as of Sep 7, 6 venues, full 385-instrument Gate catalogue):
Warehouse RWA-perp comparison (stock/ETF/index scope, latest complete day, DefiLlama's own tracking):
Read together, the two tables tell the same story from different angles: Gate has 2.3x Binance's pair count in the warehouse comparison, and 2.1x Binance's listings in the vendor comparison too — but Binance runs ~7x Gate's volume and 3.4x Gate's open interest. Breadth leader, not scale leader, in both.
The numbers point to one thing: deep, actively-managed order books. That capability shows up across two separate product lines (crypto and equity perps) and survives the one comparison: funding volatility, checked against identical instruments on other venues. None of that makes Gate the largest venue in either table above — it isn't, and the tables say so plainly. It's evidence for a specific, narrower claim: that Gate's liquidity is deep enough, and disciplined enough, to support the kind of leveraged trading this piece opened by asking about — not a claim to be the biggest market doing it.
Conclusion
Strip away the listings counts and the volume totals, and the question this piece opened with was never really "who has the most stocks" — it was whether any venue can actually hold a leveraged stock-perp book together at scale. That's an infrastructure and stability question, not a catalogue question, and it's the one the data above was built to answer.
Gate's answer is specific, not sweeping. It isn't the largest stock-perp venue by volume or open interest — Binance runs ~7x its volume and 3.4x its open interest in this data, and that gap is real and disclosed, not talked around. What it is: the broadest equity-perp catalogue of the venues compared here, the deepest top-of-book on four of its five most-traded names (and the deepest liquidity at the wide 2% band on four of the five, and the largest combined 2% book across them ($92.8M vs Bitget's $80.9M), the same depth pattern independently visible in its unrelated crypto markets, and — on the number hardest to manufacture — funding-rate volatility on identical stock contracts that's essentially tied with OKX for the steadiest in the group, despite carrying a fraction of Binance's open interest on those same names.
That last point is the one worth remembering after the tables are forgotten. A bigger book can still produce worse funding stability than a smaller, better-managed one, and for a trader deciding whether to keep leveraged exposure open overnight, stability is the number that determines whether they stay. Listing counts got tokenized equities onto exchanges. Whether traders actually stick around in size is going to be decided by exactly the kind of order-book discipline this piece measured — and on that measure, Gate's crypto liquidity and its stock-perp funding stability are two data points pointing at the same underlying capability, not two unrelated claims stitched together for a press release.

Gate has listed the most tokenized equity perpetuals, by far - standing at 385, beyond all the other exchanges. We want to go one layer deeper to diagnose whether a venue can actually support leveraged trading in these instruments at scale.

Deep Enough to Trade: Gate’s Case For Tokens And Stocks

Sentora just shipped a purpose-built vault that lets PYUSD holders earn yield by lending directly against Huma Finance's PayFi Strategy Token ($PST). This initiative represents one of DeFi's most direct efforts to bring real-world payment credit, as opposed to treasuries or crypto collateral, onchain.
Within its first 20 days of DeFiLlama-tracked history, the vault grew from $5.4M to a current $41.6M, and is still growing rapidly while maintaining a high utilization rate across the underlying Morpho markets.
The “What?” 
Put plainly: it's a stablecoin savings product where the yield comes from cross-border payment financing instead of the usual sources (staking, treasuries, crypto-collateralized borrowing).
Huma Finance V2 originates short-term loans to licensed cross-border payment companies, which require upfront capital to pre-fund settlement corridors. Think of a remittance business that must maintain dollar reserves in a corridor before a customer's transfer is cleared on the receiving end. Huma provides this capital in stablecoins backed by fiat deposits, and the loan is typically repaid within one to seven days following settlement and the return of cash flows. This structure constitutes the economic engine: real payment companies, real short-term working-capital loans, real repayment from real cash flow. Since its inception, Huma has originated more than $9 billion in financing, with zero credit defaults.
PST (PayFi Strategy Token) serves as Huma's receipt when money is deposited into that lending pool; if you hold PST, you have a claim on a group of the short-duration payment loans, and the 'yield' consists simply of the interest and the origination fees that those payment companies pay.
Sentora's vault provides DeFi users with exposure to this lending mechanism. Users deposit PYUSD (PayPal's stablecoin) into a vault on Morpho, a widely recognized permissionless lending protocol. Sentora, serving as the risk curator, designs the parameters that enable deposited PYUSD to fund loans collateralized by PST. In return, depositors receive a receipt token, senPYUSDPST, which accrues interest as the loans are repaid. Users do not hold PST directly nor interact with Huma; instead, they lend PYUSD against PST as collateral within a Morpho market, where Sentora manages risk parameters such as loan-to-value caps, exposure limits, and market allocation.
This is not the first time Huma’s PST is active on Morpho though.  PST was first listed as Morpho collateral on June 10, 2026, through a separate USDC vault curated by RockawayX. That launch attracted over $16 million in deposits within its first week. PST is also curated by Bitwise and Galaxy Digital.
The vault covered here is a second, distinct product: Sentora's own PYUSD-denominated vault, launched roughly two and a half months later in late August, using PYUSD specifically rather than USDC, and it has already far outgrown the previous PST vault.
The “Why?”
Firstly, Huma needs a way to make PST usable, not just held. A yield-bearing token that just sits in a wallet isn't very useful to institutional allocators; they want to be able to borrow against it without unwinding the position. Getting PST accepted as collateral on a major lending venue like Morpho turns a static receivable position into working capital. It's the difference between owning an asset and being able to use that asset.
Second, Sentora is broadening its curation business into a new asset class. The company already operates a suite of "Smart Vault" products, including RWA, ETH, BTC, and multi-stablecoin vaults, with over $2 billion in curated capital deployed across DeFi.
Payment receivables represent a new category within Sentora’s strategy, and this curation is integrated into the broader Embedded Earn distribution. It uses the same infrastructure to power Kraken's DeFi Earn program, which has grown to over 100,000 depositors and $800 million across existing vaults.
Consequently, the new PSTvault is not the final product; rather, it functions as a testing ground for a strategy that will later be introduced to larger, retail-focused exchange and fintech platforms. Depositor interest has accelerated due to attractive borrow rates in the underlying PST markets (approximately 7%, with further details provided below) and the opportunity for a highly appealing looping trade: deposit PST, borrow PYUSD against it, redeploy the PYUSD, and repeat the process.
The Performance
DefiLlama already tracks this specific pool: Sentora Huma PST Main, on Ethereum and wrapped using Morpho Blue.
Total value locked (TVL) has grown roughly 7.7x since DefiLlama-tracked history began on August 26, from $5.4M to $41.6M. Over the same 20-day period, APY remained relatively stable at 6.6%–9.2%, generally around 7%–8.5%. That stability is the stronger evidence for the underlying borrow-rate thesis: if reward emissions were fixed while TVL climbed this much, yield dilution would have shown up clearly, and it didn’t, even as TVL grew nearly 8x.
The Underlying Collateral Markets
The vault's PYUSD is actively deployed into two Morpho Blue markets where PST serves as collateral. Both markets are tracked by DefiLlama independently:
Utilization remains high but has eased slightly as supply has outpaced new borrowing: combined utilization across both PST markets now sits at about 70.6% ($54.9M borrowed against $77.7M supplied)  — still well above typical idle-liquidity levels for a market this young, even if a touch below the ~75% figure at an earlier snapshot.
Additionally, an 86% maximum loan-to-value (LTV) ratio against a receivables-backed, predominantly off-chain asset is considered aggressive by DeFi standards. In comparison, blue-chip crypto-collateral lending markets (such as ETH or BTC against stablecoins) typically maintain significantly lower LTVs due to the potential for sudden price declines in crypto collateral. PST's value, however, is considerably more stable than that of crypto assets, owing to its short-duration, cash flow-backed structure. This characteristic distinguishes both the DeFi sector and this vault specifically. 
How Big Is This?
At a current $41.6M, this single vault is well under 1% of Morpho's $9.92B in total TVL (about 0.42%). It's a much bigger share of the other two names in the stack, though: roughly 1.7% of Sentora's $2.46B curated book, and about 13.6% of Huma Finance's own $305.8M in DefiLlama-tracked TVL. This one vault is now a meaningful fraction of Huma's entire tracked footprint, which is arguably the more interesting framing given Huma is the origination engine behind the yield.
The Takeaway
For those unfamiliar with the product, the vault offers several potential advantages:
The vault provides a distinct source of yield. Most "real-world asset (RWA) yield" in DeFi has historically referred to tokenized Treasury bills, which generate returns linked to interest rates and behave similarly to other treasury products. In contrast, PST's yield is determined by the repayment behavior of licensed payment companies, introducing a fundamentally different risk factor. This yield is not correlated with Federal Reserve policy or crypto market sentiment. If this decorrelation persists, the vault could serve as a valuable diversification tool for on-chain portfolios, rather than functioning as another treasury wrapper with a different branding.
Demand is materializing rapidly from genuine usage. Approximately 70.6% combined utilization in the PST markets, following a brief ramp-up period, indicates that deposited liquidity is actively being borrowed and deployed.
Sentora's Huma PST Main vault exemplifies a small yet rapidly expanding DeFi primitive. Payment-receivables credit is being operationalized through a familiar and well-understood structure, namely a curated Morpho vault. Deposits are both genuine and increasing, while the underlying collateral markets demonstrate substantial borrowing demand rather than idle liquidity intended solely for TVL metrics. This level of substantive activity surpasses what many "RWA" launches achieve in their initial week.

Sentora just shipped a purpose-built vault that lets PYUSD holders earn yield by lending directly against  Huma Finance's PayFi Strategy Token ($PST). This initiative represents one of DeFi's most direct efforts to bring real-world payment credit, as opposed to treasuries or crypto collateral, oncha

Sentora just shipped a purpose-built vault that lets PYUSD holders earn yield by lending directly against Huma Finance's PayFi Strategy Token ($PST). This represents one of DeFi's most direct efforts to bring real-world payment credit, as opposed to treasuries or crypto collateral, onchain.

Sentora Brings Huma's Payment Credit to Morpho: A Spotlight on the New PST Vault

Executive Summary
A new DeFi category is forming around a simple premise: institutions want the composability of public blockchains without broadcasting their positions to every competitor and MEV bot watching the chain. Call it confidential DeFi — the use of Fully Homomorphic Encryption (FHE), zero-knowledge proofs (ZK), multi-party computation (MPC), or trusted execution environments (TEEs) to keep balances, trade sizes, and strategies private while settlement stays on a public, auditable chain.
DefiLlama currently tracks 38 protocols under its privacy category, with a combined $892.4M in TVL — up 24.4% over the past 30 days but roughly flat (+4.6%) over the trailing year. That category mixes two very different things, though: anonymity-focused mixers like Tornado Cash, which most institutional compliance frameworks explicitly exclude, and compliance-compatible confidential-DeFi infrastructure such as Railgun, Zama, and Privacy Pools. Stripping out Tornado Cash, the remaining tracked cohort grew from roughly $70.1M in combined TVL in March 2025 to $163.2M in September 2026 — a rise of about 133% over eighteen months. That is a genuine expansion of on-chain confidential balances, not merely a re-shuffling of an existing pool.
Live implementations today span all four cryptographic approaches, but one has moved decisively ahead on the institutional-yield use case: on September 15, 2026, Zama, Morpho, and five institutional curators (Steakhouse Financial, Armitage by Wintermute, Flowdesk, RockawayX, Bitwise) are expanding from a single confidential USDC vault to 16 confidential vaults spanning five asset classes (USDC, USDT, WBTC, AUSD, tGBP), alongside the launch of the Zama Swap Protocol for confidential swaps between shielded positions. This is the clearest evidence yet of institutional capital moving from single-vault experimentation toward a repeatable, multi-curator confidential-yield product line.
Why Now
Institutional capital's core objection to public DeFi has always been informational: every position, every rebalance, every unwind is visible in real time to anyone watching the mempool. A 2025 empirical study of Ethereum sandwich attacks ("Sandwiched and Silent," Mancino & Rezzoli) quantifies just how costly and persistent this exposure is — and, more importantly, shows that the industry's existing workaround (private transaction routing) is not a real fix. Between November and December 2024 alone, the paper confirms 117,581 public sandwich attacks and a further 2,932 private-path sandwich attacks affecting 3,126 victim transactions, with $409,237 in victim losses and $293,786 in attacker profit — and a single bot responsible for roughly 65% of all private-path frontrunning. Private routing itself concentrated 50.1% of all Ethereum transactions by February 2025, up from 31.8% a year earlier — meaning the industry's main defense against information leakage is itself becoming a centralized chokepoint rather than a genuine privacy guarantee.
For institutions, the stakes go beyond MEV: a visible position is a visible trading strategy, a visible cost basis, and a visible compliance exposure. Confidential DeFi reframes the problem at the settlement layer instead of the routing layer — encrypting the balance itself rather than hoping no one is watching the mempool.
Landscape: Four Competing Approaches
Confidential DeFi is not one technology; it is four cryptographic approaches converging on the same category, each with different tradeoffs:
FHE (Fully Homomorphic Encryption) — computes directly on encrypted data without ever decrypting it. Zama, the largest live FHE deployment, reports its underlying scheme rests on lattice cryptography, which is inherently post-quantum resistant unlike ZK or TEE-based approaches. 
The tradeoff has historically been computational overhead, but Zama's Q2 2026 reports that the protocol reached 1,000 confidential transfers per second on standard GPUs — a milestone arrived roughly a year ahead of its internal roadmap.
ZK (Zero-Knowledge Proofs) — proves a transaction is valid without revealing its contents, and underpins the largest current deployments by TVL: Railgun ($96.5M) and historically Aztec Connect  (now near zero TVL). 
ZK shielded pools face a structural auditability tradeoff: following a June 2026 vulnerability disclosed in Zcash's Orchard shielded pool, once assets are shielded under a pure ZK design, third parties generally cannot independently verify that shielded supply matches underlying deposits.
MPC (Multi-Party Computation) — splits computation and key material across multiple parties so no single party sees the full data. It underpins institutional custody/wallet infrastructure — Zama cites Bron, a self-custody wallet combining MPC with native confidential-token support, as a partner  — more than base-layer settlement today.
TEE (Trusted Execution Environments) — relies on hardware-isolated enclaves to process data privately. DefiLlama tracks Oasis Sapphire, a TEE-based confidential EVM chain, at just $175.5K in TVL, down 96% year-over-year — the weakest traction of the four approaches in tracked DeFi TVL today, though TEEs remain in active institutional use for custody and off-chain compute.
Use Case Deep Dive: Steakhouse's Confidential Prime USDC Vault on Morpho
The most advanced live implementation of confidential yield today runs on Morpho, curated by Steakhouse Financial — and it is now the template being replicated across four more curators. The Steakhouse Confidential Prime USDC vault, live since June 2026, uses the same collateral set and credit strategy as Steakhouse's existing public Prime USDC vault on Morpho (cbBTC, wstETH markets), with one difference: it accepts confidential USDC as the deposit asset. Deposits are batched every 24 hours before being routed into the underlying Morpho strategy; an observer can see the batch total but not any individual depositor's balance or position.
The Steakhouse Confidential Prime USDC vault's growth curve is the clearest read yet on institutional appetite for confidential yield. From a standing start at launch in mid-June 2026, the vault reached $25.5M in deposits by the end of the month — enough to place it among the top-10 USDC vaults on Morpho by TVL, paying 8–13% APY including incentives. Growth continued through the vault's 12-week incentive program, peaking near $40M roughly seven weeks after launch — consistent with the program's own design, which front-loaded the largest yield boosts into the first two weeks and tapered them steadily through a "mid pulse" and "late pulse" phase.
By September 11, 2026, DefiLlama's independent tracking put the vault at $33.5M in TVL, earning a 7.2% base APY — a pullback of roughly 16% from the ~$40M peak, in line with what the incentive structure would predict: as the boosted-yield phases wind down, the vault's marginal APY compresses back toward its native ~7% rate, and some yield-sensitive capital rotates out. The trajectory is a rise, a peak tied to incentive design, and a partial settling — not three competing numbers, but one consistent growth-and-normalization curve for the first live confidential-yield product at scale.
For context on the venue itself: Morpho  carries $9.6B in total TVL (+21.8% over 30 days) and Steakhouse Financial curates $3.0B across its vault suite — so even at its $40M peak, this vault represented well under 1% of either platform's TVL. The significance isn't scale yet; it's that a fully confidential vault, running the exact same credit strategy as its public counterpart, sustained real deposits through a full incentive cycle on infrastructure institutions already trust.
Metrics: Shielded TVL as a Category Benchmark
DefiLlama's privacy category is the closest existing proxy for a category-wide "shielded TVL" metric, currently standing at $892.4M in TVL across 38 tracked protocols. It is, however, a blunt instrument today: it mixes anonymity mixers with compliance-oriented confidential-DeFi infrastructure, and each protocol's own TVL methodology is not standardized to a shared "value currently shielded" definition.
Zama separately reports its own shielded Total Value Locked (TVL) metric, and the two-quarter trajectory is worth showing rather than a single snapshot: shielded TVL peaked at $123M in Q1 2026 (driven by a $120M ZAMA token public sale shielded within three days) but stood at a $39.6M snapshot at the end of Q2 2026, up 15% quarter-on-quarter from Q1's own end-of-quarter level. 
Underlying, more durable shielded balances (from vaults, stablecoins, and tokens) are growing steadily — Q2's cumulative shielded volume reached $415.5M year-to-date, up 44% quarter-on-quarter, with 10,900 wallets holding shielded assets, up 17% quarter-on-quarter.
Zama's Q2 report also breaks TVL down by asset class as of June 30, 2026: stablecoins ($14.1M), tokens ($6.8M), vaults ($18.8M), and RWAs ($52.2K) — which sum to roughly $39.75M.
Adoption
Several signals point to maturation rather than one-off experimentation. 
Curator breadth: Five curators — Steakhouse Financial, Armitage by Wintermute, Flowdesk, RockawayX, Bitwise — spanning stablecoin, RWA, and BTC-collateral strategies, describing confidential entry as a natural extension of existing curated strategies rather than a new product line. 
Standards convergence: ERC-7984, the confidential token standard, was developed jointly by Zama and OpenZeppelin. It now sits at the center of the Confidential Token Association (CTA) — an industry group whose stated mission is 'a common framework for encryption-based onchain confidentiality,' co-founded by three organizations: Zama, OpenZeppelin, and Inco.
Zama's Q1 2026 report notes the standard is already integrated into OpenZeppelin's Contracts library, the most widely used smart-contract toolkit in Ethereum development. 
Institutional participants: GSR (OTC market maker) settled an OTC trade confidentially via Zama's protocol, Bron executed what Zama describes as the world's first confidential payroll on Ethereum mainnet in December 2025, and T-Rex Ledger who chose Zama to become confidentiality layer for the upcoming tokenisation project for Apex Group  ; 
Dfns, enterprise wallet infrastructure, integrated the confidential token standard natively in Q2 2026. 
Distribution ecosystem: access points beyond the Zama app itself — Zerion, Utila, Yield.xyz — and incentive layers — Pendle, Merkl — are part of the same expansion. The category is being built with distribution partners from the outset rather than as a single-app product.
Risks and Open Questions
The category faces three structural risks that apply regardless of which cryptographic approach wins. 
Fragmentation: FHE, ZK, MPC, and TEE implementations are not interoperable — a confidential USDC balance on one system cannot move to another without unshielding first, fragmenting liquidity across incompatible privacy silos. 
Standardization risk: ERC-7984 is gaining adoption, but it is one proposal among several possible confidential-token designs, and its long-run dominance is not assured. 
Regulatory uncertainty: programmable/selective-disclosure compliance is a design goal across every approach in this category, but no jurisdiction has yet defined a clear regulatory treatment for onchain confidential assets, leaving institutional adopters to build on infrastructure whose compliance status could shift. 
A fourth, technical risk is auditability: Zama's own report argues that ZK-shielded systems, following the June 2026 Zcash Orchard disclosure, can lose the ability to verify that shielded supply matches deposits — a category-wide concern for any pure ZK shielded-pool design, though this framing comes from a competing FHE vendor and should be read as an interested-party claim rather than a neutral technical audit.
Outlook
Three triggers would move confidential DeFi from emerging to mainstream: 
Additional institutional entrants beyond the current OTC/custody/curator set;
A compliance standard actually converging across jurisdictions rather than merely being proposed;
A second major curator ecosystem — beyond Morpho — adopting confidential vaults at comparable scale. 
The expansion to 16 vaults across five curators is itself a strong signal on the first count, showing that curators beyond Steakhouse (Wintermute's Armitage, Flowdesk, RockawayX, Bitwise) will adopt confidential rails once one proof point exists on infrastructure they already trust. 
The next signal to watch is whether a curator ecosystem outside Morpho — a different lending/vault venue entirely — replicates this pattern within the next 12 months, and whether the swap layer (Zama Swap Protocol) sees comparable adoption to the vault layer once it has a full quarter of reported volume.
APPENDIX: Methodology
TVL and category figures for DefiLlama-tracked protocols (privacy category total, Railgun, Zama, Privacy Pools, Aztec Connect, Oasis Sapphire, Morpho, Steakhouse Financial, and the Steakhouse Confidential Prime USDC vault) are drawn from DefiLlama's warehouse as of September 11, 2026. Figures attributed to Zama's Q1 2026 and Q2 2026 Shielded Reports, and to the September 15, 2026 press release, are self-reported by Zama and its named partners and are labeled as such throughout; reconciliation deltas between these sources and DefiLlama's independent tracking were computed directly from the source figures and are marked as derived. The MEV/private-routing statistics in Section 2 are drawn from "Sandwiched and Silent" (Mancino & Rezzoli, 2025).
Data as of September 11, 2026. Source: DefiLlama, plus Zama's Q1 2026 and Q2 2026 Shielded Reports and the September 15, 2026 press release (self-reported, labeled), and "Sandwiched and Silent" (Mancino & Rezzoli, 2025).

A new DeFi category is forming around a simple premise: institutions want the composability of public blockchains without broadcasting their positions to every competitor and MEV bot watching the chain.

Confidential Onchain Yield: The Missing Infrastructure for Institutional Capital

Nexo

Andrés Ondarra

Andrés Ondarra is Nexo's General Manager for Argentina, leading local operations from the company's Buenos Aires hub. He brings more than 25 years of experience across traditional finance, fintech, and crypto in Latin America, including a background in Wall Street investment banking. Nexo is a leading digital assets wealth management platform designed to help clients grow, manage, and safeguard their crypto holdings. Active in more than 200 jurisdictions, with over 10 million users and more than $7 billion in assets under management (AUM), it is one of the largest crypto lenders globally. Founded in 2018, Nexo has made Buenos Aires its regional hub for Latin America, entering the market through its Buenbit acquisition, local partnerships, and savings products, with further LATAM expansion now under way.
Argentina moved $93.9 billion in crypto volume over three years and generates the highest per-capita crypto volume in Latin America. Out of every market Nexo could have picked for a LATAM hub, what made Argentina the obvious choice, and how much did the Buenbit acquisition shape the timeline versus the underlying market itself?
We chose Argentina because of how its users behave. According to Chainalysis's Global Crypto Adoption Index, Argentina leads Latin America in per-capita adoption, with close to 19.8% of the population using cryptocurrencies in some form, placing it within the global top 20 for adoption. That profile — retail, non-institutional, looking to protect against inflation and dollarize savings — is exactly the profile our value proposition is designed to serve: daily-interest savings, crypto-backed credit, and now the Nexo Card.
There's an additional point that doesn't always show up in volume statistics but was decisive for us: Buenbit already offered its more than one million users the ability to invest from very low amounts in U.S. stocks. For years, the average Argentine investor could only access fractional U.S. assets through CEDEARs, which trade in pesos and are exposed to the exchange-rate gap. Buenbit solved that friction with an account denominated directly in dollars. Finding that product already built, tested, and with real traction was a very clear signal that Argentina not only had the market best suited for us, but also the talent and infrastructure to scale it with Nexo's global platform.
On top of that, since 2024 Argentina's National Securities Commission (CNV) made registration as a Virtual Asset Service Provider (VASP/PSAV) mandatory to operate legally in the country. Buying a platform that was already registered and licensed — with a clean track record and an established user base — allowed us to enter much faster than if we had started the licensing process from scratch. In short: the market explains why Argentina; the acquisition explains why we were able to move at the speed we did.
Nexo entered Argentina by acquiring Buenbit rather than registering as a PSAV from scratch. Was that regulatory shortcut the deciding factor in choosing Argentina over other LATAM markets, or did the $93.9 billion in three-year volume and the $2,041 per-capita figure (highest in the region) matter more?
 One thing is why we chose Argentina, and another is why we chose to enter by acquiring rather than registering ourselves.
We were going to bet on Argentina regardless: the size of the market and, above all, the intensity of use — the enormous local habit of using stablecoins as a savings tool rather than for speculation — already justified it on their own. What determined the form of entry was purely regulatory. Registering a VASP from scratch involves meeting minimum net-worth requirements, custody policies, compliance and cybersecurity standards, and going through the CNV's approval timelines. Acquiring Buenbit, which was already registered and had a clean operating track record, saved us that time and that risk, and allowed us to start operating with the regulatory confidence already built in. But it wasn't just a licensing matter: the appeal we saw in Buenbit's platform, with a solid track record and a value proposition that resonated strongly with Argentine users, was equally decisive. That already-built product, together with the regulatory registration already obtained, helped us accelerate our expansion into a market that is so strategic and such a priority for Nexo.
Argentina has one of the most stablecoin-native user bases in the world, with the vast majority of exchange activity happening in dollar-pegged assets. How has building for Argentine users shaped the way Nexo thinks about its products globally? Is this a market that's pulling Nexo's roadmap forward rather than the other way around?
The influence runs in both directions, and that's actually the interesting part. Daily-interest digital-dollar savings and crypto-backed credit already existed on Nexo's global platform before we arrived in Argentina. What changed here was how we communicated it and what local rails we built so a user could move in and out of those products from a local bank account without friction: for example, in addition to being able to fund the account with pesos, we recently launched local dollar rails, which allow moving dollars between a local bank account and the platform instantly — something designed specifically for the logic of the Argentine saver.
But it's also true that Argentina is serving us as a testing ground. Our intention is, building on this experience, to expand into other markets with similar dynamics — high inflation, capital controls, the need for dollarization — and that includes both products and narrative. Argentina isn't rewriting our global roadmap, but it is accelerating and validating it under real conditions, with a level of usage that few markets in the world have. That learning is part of what we now want to bring to Peru and Mexico.
The digital dollar savings product is being positioned as an upgrade on the traditional plazo fijo. What was the reaction from Argentine users when Nexo introduced it, and what does that tell Nexo about how ready this market is for crypto-native savings tools compared to other countries it operates in?
We officially launched the product in March 2026, presenting it, without overstating it, as an alternative to the traditional fixed-term deposit: stablecoin savings with daily interest credited automatically and no minimum terms, with immediate availability of funds, and a yield that at the time stood around 13% annual in dollars — above the average for local fixed-term deposits and mutual funds, which were around 8% annual at the time. As part of the launch, we lowered the entry threshold for our loyalty program's welcome benefit (Platinum tier) to a deposit of US$1,000 within the first seven days, so that more users could access a benefit originally aimed at higher-capital users.
The reception was very positive and faster than in other markets where we operate: Argentine users don't need us to explain what a stablecoin is or why it makes sense to save in digital dollars — they already arrive with that habit. That makes Argentina, compared with other countries in the region, a particularly mature market for this type of product.
Nexo became the AFA's regional digital asset partner for what turned out to be a massive year for Argentine football. Looking back at the World Cup now that it's behind us, what did that partnership deliver for Nexo, and how did the AFA relationship change the way Argentines think about the brand?
In April 2026 we became the Official Regional Digital Assets Partner of the Argentine national football team for South America, a one-year agreement that included ticket giveaways for Argentina's World Cup matches, signed jerseys, and content with national-team players. We chose to partner with the Argentine national team because we share the same values: discipline, ambition, and the idea that you can always go further are the same principles we built Nexo on. Beyond the sporting result — Argentina reached the World Cup final, which remains a source of pride for the whole country — for us the value of this partnership never depended on a single result, but on something deeper: it allowed us to humanize and localize the brand, to show that Nexo understands and shares what matters to Argentine people, to show that behind the platform there's a local team, of Argentines, working every day so that other Argentines have better financial tools, and to associate ourselves with a symbol that brings together millions of people regardless of whether they already use crypto or not.
The partnership with the AFA was consistent with our local strategy: it wasn't generic advertising, but about building a long-term relationship with a country that has already chosen digital assets as part of its everyday financial life. That's why our campaign was built around that idea of continuity and progress — "always moving forward" — and the constant pursuit of the next generation of champions, both on the field and in the way we think about the future of digital finance in Argentina. We feel inspired by those who always go further.
Banks are expected to start offering crypto services in Argentina once the BCRA lifts its restriction. Does Nexo see that as validation of the market it's already built in, and what does it want its role to look like once that happens?
It would be great news for the whole ecosystem, and yes, we would read it as validation. If the Central Bank decides to lift the restrictions that currently prevent banks and payment service providers (PSPs) from offering crypto, it's because the market that platforms like ours built over years of restrictions has already shown demand solid enough for traditional banking to want to get involved. And that, ultimately, is good for the user: more access, more competition, and more opportunities for more people to benefit from what digital assets offer.
We don't see it as a threat to what we've built, but as the next stage in this market's maturation.
Buenos Aires is now Nexo's hub for expanding into Peru and Mexico. What is it about what's been built in Argentina that Nexo is most excited to bring to the rest of the region?
What excites us most is having proven, with real numbers rather than just hypotheses, that a digital-dollar savings product with a strong local narrative — positioned against traditional savings instruments people already know, like the fixed-term deposit — can drive mass adoption in a market under currency stress. That learning, together with the experience of building local rails in domestic currency to reduce onboarding friction, and a brand strategy with cultural grounding like the one we had with the AFA, is exactly the combination we want to replicate in Peru and Mexico.
These are markets with their own dynamics: Mexico, according to Chainalysis, is the region's third-largest market by volume, and Peru, though smaller, is growing strongly. In both cases the pitch is similar to Argentina's: access to better dollar rates, wealth diversification, and, over time, the ability to move money between countries in the region more simply. Buenos Aires functions as the center from which we coordinate that expansion, drawing on the team and the knowledge we've already built here.
If everything goes right for Nexo in Argentina, what does the company look like here in five years, and what role does Nexo want to have played in Argentina's crypto story by then?
 I think we're already seeing it, and that gives us a lot of confidence about where we're headed. In five years, we picture Buenos Aires consolidated not only as the hub toward Peru and Mexico, but as a relevant decision-making center for all of South America. We also picture much deeper integration between crypto and the traditional financial system, and Nexo as one of the reference players in the convergence of financial solutions that let users manage their assets through comprehensive offerings and access to more efficient products.
The challenge ahead of us is staying differentiated from other local and international exchanges competing for the same dollarized audience, and the way to achieve that is by keeping doing what got us here: listening to what the Argentine saver needs before anyone else, building products from that, and then bringing them to the rest of the region. If in five years people associate Nexo with having helped millions of Argentines preserve the value of their savings in a simpler and more transparent way, we will have fulfilled the role we set out to play.

"Argentine users don't need us to explain what a stablecoin is or why it makes sense to save in digital dollars — they already arrive with that habit. That makes Argentina, compared with other countries in the region, a particularly mature market for this type of product."

Why Nexo is betting on Argentina to lead its LATAM expansion

Order book DEXs, by design, can only do so much: they’re built to match buyers and sellers without ever taking a side themselves. That works well enough for crypto perpetuals, where a deep bench of market makers is comfortable managing risk on both sides of the book. But this fundamental structure hits a wall when you try to bring real-world assets or traditional finance (TradFi) instruments onchain. Dealers in those markets want to know who they’re trading with, and they prefer to manage risk on a know-your-counterparty basis, not anonymously.
This is a point Variational makes in its own documentation about its architecture: when order books are used in long-tail or off-chain markets, they break down because they try to create liquidity from scratch rather than drawing on the depth that already exists elsewhere.
Rather than building liquidity from scratch, Variational is designed to tap into the depth that already exists. Omni is Variational's cross-margined trading venue, spanning both crypto-native and RWA perpetuals, built on the Variational protocol. It runs on a request-for-quote system where a single liquidity provider, OLP, is the counterparty to every trade and hedges its own book off-chain, aggregating crypto liquidity from major exchanges alongside TradFi liquidity sourced directly from dealers.
That structure is why Variational can list something a CLOB exchange fundamentally cannot: Swaps, a new instrument type that routes trader exposure directly to TradFi dealer liquidity instead of aggregated crypto-native depth. Funding on Swaps is largely flat, benchmarked to USD borrow rates rather than the long/short imbalance that drives a typical perp funding rate. For example, on the gold and US100 markets when this article was written, longs pay 4-6% while shorts receive 2-3%.
A protocol precursor
Omni is the live proof of the OLP mechanism described above, not a separate product bolted on afterward. Every perp on Omni already routes through the same request-for-quote structure Swaps uses, just without the bilateral dealer hedge underneath. The real difference is where the liquidity comes from: Omni's perps source liquidity from crypto exchanges, while Swaps source liquidity directly from signed agreements with TradFi dealers. That makes Omni's trading history the closest thing to a track record Swaps had before launch.
That track record grew fast: Variational's RWA markets went from 6 listings and $188.4 million in open interest in May to well over 119 markets today and over $700M in TradFi OI. The listing count alone grew nearly 9x in a single month, from six to 53 in June. Growth since has been steadier, and open interest currently sits at its highest point.
Variational currently ranks third among tracked perpetual DEXs by open interest, at $1.77 billion, and 2nd in RWA open Interest. Every one of its 119 active RWA perp markets is still live, none delisted, carrying roughly $710 million in open interest.
Breaking open interest down by asset class reveals just how concentrated demand is: precious metals account for 42% of OI ($298.1 million) while covering only seven markets, while public equities, despite being spread across 83 tickers, make up just $116.5 million. Metals alone command nearly twice the open interest of all listed equities combined.
Seven gold and silver markets, carrying almost twice the open interest of 83 equity tickers combined, tell you where demand for on-chain TradFi exposure currently sits. It also tells you where swaps are likely to matter most first: metals and indices are the kind of deep, benchmark-priced underlyings where a dealer relationship beats an aggregated order book most visibly.
Top individual markets by OI:
Gold ($112.6 million)
Gold Swap (XAUS) ($71.8 million)
Tether Gold ($71.5 million)
Nasdaq-100 Index ($71.5 million)
S&P 500 Index ($57.6 million)
About Swaps
Variational co-founder Lucas Schuermann described the mechanics behind Swaps as “a bilateral contract where the trader faces a direct counterparty, in our case the Omni Liquidity Provider, rather than an exchange order book.”
Unlike perpetuals, which rely on variable funding rates, swaps have predictable carry costs. OLP hedges swaps directly on TradFi venues, so Variational can offer depth that isn’t capped by what crypto-native order books can support. Further, swaps often carry defined trading hours instead of trading 24/7 like perps, along with different index data, roll schedules, and hedging parameters underneath.
Put simply, the difference comes down to who's setting your costs.
A perpetual future is a synthetic position an exchange builds to track an asset's price without ever expiring, and the cost of holding it moves with the crowd, a funding rate that shifts depending on how many traders are long versus short at any given moment.
The kind of swap Variational is building works differently: it's a contract with one specific counterparty, a dealer, who prices your position off a reference rate instead of crowd behavior. The cost ends up looking less like a bet and more like a loan, a steady rate for the exposure, similar to what you'd pay to borrow money. Institutions favor that structure for this reason: the financing is a known, bookable number instead of one that can move against you unpredictably. 
On Variational, swaps and perps will sit side by side rather than replace each other. For example, search “XAU” and both XAU-PERP and XAU-SWAP will surface. A trader who wants 24/7 access and is comfortable with variable funding stays on the perp. A trader who wants tighter, dealer-backed liquidity and predictable carry moves to the swap, at the cost of trading hours that initially mirror the underlying’s own market.
Choosing swaps over perps means accepting limited trading hours in exchange for more predictable costs and dealer-backed liquidity; the kinds of decisions that traders in traditional finance already make, depending on their priorities.
Swaps’ cost structure
Variational is launching Swaps with some big commitments behind it: the team has secured over $1 billion in open interest capacity through signed agreements and is offering stable financing usually benchmarked to USD borrow rates at about 4.5% all-in.
Swap financing is also typically anchored to a transparent benchmark: carry is priced off SOFR, the standard USD reference rate, plus a spread.
The launch market for gold (XAUS) is already showing this methodology in action, with funding rates derived directly from live market conventions and visible in real time. At launch, XAUS opened with a quote of 0.62 bps spread on $2.24 million of open interest - it has since grown to ~$72M within the first week. As of writing this article at launch, funding is asymmetric: 0.0157% for longs and 0.0068% for shorts on the standard daily interval.
The protocol encodes a key tradition from TradFi FX and money markets: on most markets, the funding rate triples every Wednesday to account for the weekend carry, since most trades settle two business days later and the Friday-to-Monday stretch involves three calendar days. Rather than leave that gap unpriced, the market charges it all at once in a triple-sized instalment.
Note: not every market times it the same way; US100, for instance, charges triple on Fridays instead, but the underlying logic, pricing the extra weekend days upfront, is the same.
“Flat carry” refers to funding, the ongoing cost of holding a position, being predictable instead of variable. It says nothing about whether the entry spread itself stays flat as an order gets bigger. In fact, the data says the opposite: Swaps spreads widen as size increases — they just start from a much lower base than the competition. The largest single trade recorded so far illustrates this: a buy of 87 US100S contracts for $2.5 million in notional executed at just 1.08 bps of slippage.
Like-venue comparisons
Spread data from Swaps' launch markets, US100 (Nasdaq-100) and XAUS (gold), shows how the new model diverges from traditional perps.
Perp DEXs, including Variational's own Omni, publish flat fees. For Omni, that's 1.54 bps on Gold Perp. Swaps, in contrast, offer dealer-priced spreads that widen with order size, reflecting genuine market depth and size-based risk instead of applying a single rate to every trade.
For an outside benchmark, we're also comparing against Ostium, a dedicated onchain RWA perps venue built on Arbitrum. It's the deliberate choice here because it lists Gold and Nasdaq-100 directly, the same two underlyings Swaps is launching with, giving a same-asset comparison rather than a proxy. Ostium's live taker fee, the more complete figure to cite over its docs-quoted opening fee alone, runs flat at 5.73 bps on Gold and 5.29 bps on Nasdaq-100.
On US100, Swaps undercut Omni's own live perp fee by 19x at $1,000 and 9.5x at $1 million, wider margins against Ostium's live fee (52.9x and 26.4x). On XAUS the advantage is there, but narrower: 4.2x at $1,000 against Omni and 14.8x against Ostium, compressing to 1.5x and 5.5x respectively by $1 million. The advantage is largest where most trading actually happens, and on the one asset with the deepest existing perp market, it hasn't been shown to hold at the very top end the way it does lower down. 
…vs TradeXYZ
Over the trailing 13 weeks, the TradeXYZ picture has been mixed. TradeXYZ is still roughly 3x bigger by outright volume ($255 billion vs. Variational's $80 billion), but its own 30-day volume just fell 44.4%, its worst stretch in the tracked window, while Variational's share of the two venues' combined volume climbed from a low near 14% in late July to 36.4% most recently. The leader here is shrinking faster than the challenger is growing, and the gap, while still wide, has compressed rather than held steady.
Worth noting: TradeXYZ isn't an independently built venue, it's the first HIP-3 market on Hyperliquid's own infrastructure. Thus, its scale reflects someone else's liquidity stack, not a vertically integrated one like Variational's, which is why it's the most relevant benchmark in the category without being a true like-for-like competitor.
Down the road
Variational launched swaps on September 1, which timed nicely with an X Space co-hosted with DefiLlama Research. The initial rollout, US100 and XAUS only, was Phase 2 on Variational's own public roadmap. Phase 3, enabling 24/7 swap trading, is already the named next milestone. In the week since launch, Swaps have processed $2.8 billion in lifetime volume, with total open interest across the three live markets peaking at $245 million on September 8. US100S has emerged as the largest individual swap market, with open interest reaching $83 million on September 8. It also carries the highest trading activity of the three, with 24-hour volume hitting $416 million on the same day.
Further out, the roadmap includes the public launch for Omni (currently still in private beta), a planned token, and a separate institutional line called Variational Pro, due sometime in 2027. Lucas Schuermann has said most of the platform's listings over the next year will be swaps, with room to eventually list thousands of RWA markets. The nearer-term, more checkable version of that goal: cross 100 RWA markets with real TradFi liquidity and genuine trading activity behind them. Variational has hit a target like this before: In May, the team posted a public goal of 100 TradFi listings in 90 days, and hit it on the dot.
Some of that ambition is easier to take at face value once you know who's behind it. Schuermann and co-founder Edward Yu built and sold a quant trading firm, ran OTC crypto trading at Genesis, and helped raise $61.8M in funding for Variational (for more background history, check out Lucas’ and Edward’s personal websites).
This is not a group learning on the fly, and the sum of all that ambition has so far been an onchain growth masterclass.

Swaps, a new instrument type that routes trader exposure directly to TradFi dealer liquidity instead of aggregated crypto-native depth. Funding on Swaps is largely flat, benchmarked to USD borrow rates rather than the long/short imbalance that drives a typical perp funding rate.

Variational Swaps

We partnered with Forgd to build Universal Token Ratings, a tool that measures token disclosure and tracks token performance in real-time. Last week, we dove into the UTR data, and came to several unexpected conclusions. Chief among them: token issuers are opaque. 
We found that issuers’ limited disclosure is weighing on the scores of tokens that are otherwise performing quite well (as measured by liquidity depth, bid-ask spreads, volume, exchange coverage, derivatives conditions, market maker adherence, and on-chain tokenomics delivery). This was a surprise — disclosure might be tedious, but it’s easy. After all, it’s entirely within the issuer’s control. 
While less common — this was true in just one quarter of rated tokens — some issuers who opted for transparency failed to earn high scores due to relatively middling or less-than-stellar token performance. 
This week, we’ll take a look at the disclosure-performance gap. Before we get to the numbers, a quick reminder: the Disclosure and Performance axes are each scored 0 to 10. The overall score is their product, 0 to 100, not their average and not a percentage. A token scoring 7 out of 10 on both axes lands at 49 out of 100, not 70. That means strength on a single axis isn’t enough to lift a token to a top grade.
Transparent issuers, middling performance
The mean Disclosure score among rated tokens is 4.98 out of 10. Their mean Performance score sits at 6.07. For three out of every four rated tokens, the Disclosure axis is weighing down their total score.
Still, there were some standouts when it came to transparency. As of Tuesday, there were more than 30 tokens with a Disclosure score above 6. For the most part, they rated highly on the Performance axis — only four had a Performance score below 5: Decred, Saffron, CoW Protocol, and Turtle. 
Those four suffer from some combination of the following issues: meager trade volume relative to their market capitalization, lack of access to a top-tier exchange, and modest fully-diluted valuation (a figure that includes the value of all tokens, even those that have yet to enter circulation). All lack any data concerning market maker performance, but in that regard, they were in good company — market maker data weren’t available for 88% of rated tokens. 
These tokens’ ratings might improve along with data collection. In the UTR, an unmeasured criterion defaults toward zero rather than being excluded from the average. These category scores read low mainly because the underlying data isn't there yet, not because projects are failing the test.
The UTR difference 
Monero’s XMR isn’t a standout on the Disclosure axis — it earned a rating of 5.19 — but it still outperformed in that regard, given the low mean Disclosure score across rated tokens. And its overall score was limited by a middling Performance rating of 3.6. 
Among other things, the Performance axis flagged relatively weak trade volume and order-book depth for a token of that size. 
This gets at the key difference between the UTR and several disclosure-only token rating frameworks that came before. While any disclosure-based framework can tell you whether an issuer is transparent, few can tell you in real-time whether you can, say, purchase a token in size. 
Monero meticulously detailed its tokenomics, earning a AAA rating across three of the four sub-sub-categories. A would-be buyer might think they have all the information they need to make an informed investment. But middling order-book depth might give them pause. 
High performance, poor disclosure 
There are roughly a dozen tokens that perform exceptionally well despite a troubling lack of transparency. That group includes Hyperliquid, Pump.fun, and PancakeSwap. 
Hyperliquid, for example, enjoys remarkable trade volume, 30-day price performance, order-book depth, and open interest. Incredibly, this is all relative to its sky-high $18 billion market capitalization. Those figures offset its high rate of annualized inflation — 53% — and a substantial number of locked tokens that could weigh on the price when they enter circulation. 
This crop of tokens earns mixed scores regarding disclosure of tokenomics and service provider engagements. But they all suffer when it comes to detailing legal and financial arrangements and the makeup of the team. 
This issue is understandable. The cypherpunk ethos that birthed Bitcoin and virtually every subsequent cryptocurrency doesn’t merely give teams an excuse to operate in the dark — it creates a moral imperative to do so. A fully-doxxed founder with a company registered in the continental US, for example, doesn’t scream “overthrow the financial system with censorship-resistant, stateless money.” 
And yet there are few teams that seem to have that as their north star. More practically, crypto’s culture of anonymity has made it a favored industry for scammers and hackers. Investors are right to think twice before investing in a token run by “Chef Kids” (PancakeSwap).  
This is why the UTR demands high scores across the board. Our hope is that it creates an incentive for token issuers to do better. Blockchains are transparent. Their creators should be, too.

We partnered with Forgd to build Universal Token Ratings, a tool that measures token disclosure and tracks token performance in real-time. This week, we’ll take a look at the disclosure-performance gap.

Universal Token Ratings: The Disclosure-Performance Gap

Shunyet Jan

Q&A format focused on the financial super app vision, inclusivity and access, 24/7 trading, and price discovery.
Shunyet Jan spent the years before crypto on Wall Street, going back to the 1990s. He joined Binance in December 2025, and now oversees everything trading-related on the exchange, from spot and perpetuals to bStocks, Binance's tokenized equities product.
This conversation, pulled from a July 22nd fireside chat between our Head of Research, Ryan Calej, and Binance’s Head of Exchange & Trading, Shunyet Jan on X, followed the release of our latest Binance report, From Exchange to Infrastructure: How CEXs Are Reshaping Global Market Access. It covers four threads from that report: Binance’s financial super app vision, and Shunyet’s thoughts on access and inclusivity, 24/7 markets, and price discovery.
Binance’s financial super app vision
Binance started as a place to trade crypto. What's changed most fundamentally about the role exchanges play for users today?
Users first came to Binance to swap one crypto for another. Stablecoins were the next big shift, turning "stable to volatile" into a routine conversion. Perpetuals came after that: a 24/7 asset class that aggregates liquidity into a single instrument tracking spot price, instead of the month-by-month contracts TradFi futures markets are built around.
From there, the pattern repeated. Users didn't just want to only trade anymore. They wanted to make payments and get access to other asset classes, and exchanges had already built the infrastructure, 24/7 trading, instant settlement, cross-margining, to support it. Competitors that started in TradFi are now moving into crypto, and Binance is moving the other way, into TradFi assets and payments. Everyone is converging on the same goal: users who used to need three or four different apps now want one.
Is growth going to come more from the sheer number of assets you list, or from integrations on the collateral and settlement side, letting people use these assets as collateral for trading?
Both. Binance has started testing letting TradFi assets be used as collateral to trade other TradFi assets or crypto derivatives, and is working with regulators to expand that carefully. But the growth driver nobody's really talking about is a different one: wealth management. US equities are a massive asset class partly because there's an entire industry of mutual funds and hedge funds managing that money for people. Crypto doesn't have that. Strip out Bitcoin, Ethereum, Solana, and BNB, and there isn't much diversification left, and there isn't much of an industry helping users manage what they hold.
Vaults exist today, but only for crypto assets. Tokenizing an ETF is close to tokenizing a fund, but most ETFs are index products, meaning passive management. Active management, and asset managers who actually understand both crypto and TradFi, are rare. Most global managers stick to one jurisdiction or one asset class. Whether the answer ends up being vaults, or opening access to third-party asset managers, isn't settled yet. But asset management is the piece that's still missing.
Do you see the line between "crypto exchange" and "TradFi exchange" blurring, to the point where CEXs eventually do as much TradFi volume as crypto volume?
 The broader asset market, equities and fixed income, is much bigger than crypto. Nvidia alone is bigger than the entire crypto market combined. CEXs took the lead on 24/7 trading and cross-margining, and that advantage is going to pull volume toward normalization based on market cap. It's plausible that half of Binance's volume eventually comes from TradFi assets.
Right now, most of that TradFi exposure is US-centric. But demand isn't only for US assets: users in the US may want access to Korea, users in Africa may want access to Hong Kong. Getting a brokerage account outside your home jurisdiction is hard even if you live in a developed country. CEXs are positioned to be the first to open that up broadly.
Inclusivity and access, 24/7 trading, and price discovery
Tokenized stocks have been tried before. Why is this the year it's actually taking off?
Regulatory clarity is the biggest factor. Earlier attempts at tokenizing equities were a good first effort, but they missed pieces of what users actually wanted. The newer generation of products, bStocks included, fills those gaps. That coincided with a market backdrop where the hype and growth in TradFi, especially around AI and the hyperscalers, gave people an additional reason to want in.
Is that demand retail-driven, and if so, why?
 Very much so, and very inclusive. A lot of bStocks volume is in small, fractional positions starting at $5. Opening a brokerage account isn't hard if you live in a developed country. Most of the world doesn't have that option. Stablecoins already proved the model: they made the US dollar available globally, and that same rail now makes other assets easy to access too. Historically, buying Nvidia or Microsoft meant needing a US bank account, US dollars, and a brokerage account, three things that are simply not available to most of the world's population. Tokenized stocks, through CEXs and DEXs alike, open that door.
How important is round-the-clock access, versus the five-day trading week TradFi users are used to?
It shows up on both the derivatives side and the equities side. Hyperliquid's volume is now roughly 40% TradFi perps. Binance has followed the same curve: TradFi volume on its perp market was at zero in January and is around 30% today. TradFi itself hasn't caught up, because it hasn't built the rails for 24/7 trading. CEXs and DeFi both have, and the volume is the proof that the demand was always there. A meaningful share of bStocks volume happens outside US trading hours, when TradFi markets are simply closed.
Do the same users trade both bStocks and crypto, or do they stay in separate lanes?
 There's a lot of overlap. Binance's users tend to track the news closely, and this year that's meant AI, which has driven both the TradFi equity rally and a lot of crypto activity. Users interested in one tend to be interested in the other.
As Binance expands into commodities, ETFs, and equities, what needs to be true for those markets to have deep, reliable liquidity?
Binance is already the price setter for crypto spot and crypto perpetuals: in leader-laggard terms, when the market moves, it tends to start on Binance, even with roughly 50% market share. That isn't true yet for TradFi assets during normal trading hours, but it flips on weekends. When TradFi markets are closed, and something moves markets, a geopolitical event, a company announcement, Binance and a handful of competitors become the price setters, simply because they're the only venues open.
The bigger structural piece is Direct Equities, a product that sources real US liquidity directly and lets users tokenize it for free. That solves the liquidity question during regular trading hours: if size isn't available on Binance, it can be sourced through TradFi channels and tokenized on demand. On weekends, when TradFi liquidity isn't accessible at all, Binance's own order book is the liquidity, and it holds that position by default.

Shunyet Jan spent the years before crypto on Wall Street, going back to the 1990s. He joined Binance in December 2025, and now oversees everything trading-related on the exchange, from spot and perpetuals to bStocks, Binance's tokenized equities product.

This conversation, pulled from a July 22nd fireside chat between our Head of Research, Ryan Calej, and Binance’s Head of Exchange & Trading, Shunyet Jan on X, followed the release of our latest Binance report, From Exchange to Infrastructure: How CEXs Are Reshaping Global Market Access.

When markets never close: A Q&A with Binance’s Shunyet Jan

A leaderboard invites a single question: “Who's on top?”
But dig a little deeper into DefiLlama’s Universal Token Ratings. “Who’s on top?” turns out to be the least interesting needle in our data haystack.
Our attempt to define and recognize quality tokens generated surprising insights, to wit: few tokens clear the Disclosure and Performance axes simultaneously; one axis is a far more common bottleneck than the other; exchange access moves the score more than token size; and several of the lowest scores reflect missing data rather than failing grades.
Before we get to the numbers, a quick reminder: Disclosure and Performance are each scored 0 to 10. The overall score is their product, 0 to 100, not their average and not a percentage. A token scoring 7 out of 10 on both axes lands at 49 out of 100, not 70.
It’s lonely at the summit 
Only one token in the rated universe earned AAA, and its score is 60.6 out of 100. That's the multiplicative structure of UTR at work: reaching AAA requires a token to score highly on both axes. Strength on a single axis isn’t enough to lift a token to the top grade.
Across the full set:
→ 1 token at AAA → 24 at AA → 17 at A
→ 26 at BBB
→ 25 at BB
→ 12 at B
→ and 23 at CCC.
Rated tokens earned a mean score of 30.6 out of 100. The lowest-rated token earned a score of just 5.6 while the highest, as mentioned above, earned 60.6. 
Disclosure and Performance scores are built from several weighted sub-criteria, so closing the distance from, say, 7.75 to a perfect 10 would mean eliminating weaknesses across the entire framework at once. That’s why a 100 is expected to stay rare no matter how large the rated universe grows. (For more information on how UTR scoring works, visit the documentation or check out this article.)
Disclosure is the true bottleneck
Token performance takes care of itself. Once a token trades, its liquidity, spreads, and volume sit on public venues. What happens next is largely outside the team’s control. 
Disclosure works the other way. A project has to actively publish a vesting schedule, name its multisig signers, name its market maker, and then keep all of that information current as facts change. Points in the Performance axis accrue automatically. Earning points in the Disclosure axis requires ongoing paperwork somebody has to actually do.
The data reflect that. For three out of every four rated tokens, the Disclosure axis weighs on their total score. Rated tokens’ mean Disclosure score sits at 4.98 out of 10. Their mean Performance score sits at 6.07. And the correlation between the two axes is weak (0.24), meaning a project doing well on one tells you little about how it does on the other.
One caveat to the Disclosure-as-albatross finding: the two lowest-scoring sub-categories in the system are on the Performance side. In the Market Makers sub-category, rated tokens earned a mean score of 0.95 out of 10. In the Performance-side Tokenomics sub-category, they earned a mean score of 4.22.
But this has little impact on total scores. That’s because this data was relatively hard to gather across the board. Market maker data weren’t available for 88% of rated tokens. Buyback, burn, and inflation data weren’t available for more than 70% of rated tokens. In the UTR, an unmeasured criterion defaults toward zero rather than being excluded from the average. These category scores read low mainly because the underlying data isn't there yet, not because projects are failing the test.
If anything, that downward bias weighing on the Performance side makes the Disclosure finding more robust — even though most tokens’ Performance scores suffer an artificial penalty, their Disclosure scores still emerge as the primary burden.
Big, or broad? 
Exchange listings have an outsize impact on the Performance axis. Liquidity depth, spreads, and trading volume are, of course, measured on the venues where a token trades, and getting accepted onto trading venues takes ongoing work from a project. In other words, Binance won’t alway onboard a new token just because it has a large valuation.
In fact, valuations barely impact the UTR score, appearing only once, as a single input inside the Tokenomics sub-criteria. That means a token can carry a billion-dollar valuation and still register as thin on the exchanges we use to measure token performance.
That shows up clearly in the averages. Moving from the smallest valuation band to the largest lifts the average score by only 9.2 points out of 100. Moving from zero top-tier exchange listings to three or four lifts it by 22 points, more than twice as large a swing, and that difference sits entirely in how widely a token is listed rather than its market capitalization. Of the 128 rated tokens, 9 carry no top-tier listings at all, while 57, nearly half the universe, carry all four tracked.
What that means in practice: a token's UTR score currently says more about how widely it has reached recognized trading venues than about how much it’s worth. A large, well-funded token with narrow exchange coverage will score closer to a small, thinly-traded one than to a token that has landed broad distribution, because the score puts a premium on reach, not size.
Of course, it can be difficult to disentangle the two criteria — larger tokens tend to also carry more exchange listings, so this shows two real patterns side by side, rather than a controlled test isolating one from the other.
Sector patterns, with a big caveat 
Here’s the average overall score by sector: lending, real-world assets, and DeFi-spot tokens lead with a mean score of 36.0 (32 tokens), followed by perpetuals-DEX at 34.7 (7 tokens), and privacy at 34.7 (6 tokens). Infrastructure, the largest sector with 66 tokens, sits at 28.1. Artificial intelligence follows at 27.2 across 16 tokens. The single rated launchpad token earned a score of 26.9.
Lending, RWA, and DeFi-spot tokens earned their lead with robust Disclosure practices — they had a mean Disclosure score of 5.71 out of 10, the best of any sector.
Infrastructure and AI tokens were competitive in the Performance axis but suffered due to their lackluster Disclosure scores. 
We shouldn’t draw too many conclusions regarding privacy, perpetuals-DEX, and launchpad tokens given their small sample sizes. The launchpad "average" is one token's score, not a sector pattern. Infrastructure and lending-RWA-DeFi-spot are the only two sectors with enough tokens to support a sector-level claim with any confidence.
Low score ≠ bad token (yet)
Some data remains difficult to gather, and that weighed on tokens that might have otherwise earned decent scores. Market maker uptime and SLA adherence was unmeasured for 89% of rated tokens. Market maker volume share and depth share were both around 88% unmeasured. Buyback and burn events was 79% unmeasured. Annualized inflation, 70% unmeasured. Disclosed versus actual vesting was the easiest thing to measure, yet we were unable to gather that data for 34% of rated tokens. 
An unmeasured criterion defaults toward zero rather than being dropped from the average. That means a low score in these specific categories today reads closer to "not yet disclosed or trackable" than "measured and found lacking." 
That gap should close as more projects claim their profiles and formally disclose market maker relationships get formally disclosed. 
In a way, it’s a reminder of the reason we built the UTR. Yes, it’s meant to help people navigate the treacherous world of token investing. But it’s also meant to force teams to do better, to disclose more. It’s meant to help the industry clean up its own act. One more FTX and crypto entrepreneurs might pine for the days of Gary Gensler and the Biden Administration. 
Built and powered by DefiLlama and Forgd, Universal Token Ratings is a free, open leaderboard — any project can claim its profile, submit disclosures, and get scored on the same Disclosure x Performance framework covered in this piece.
Claim your token’s profile here.

Built and powered by DefiLlama and Forgd, Universal Token Ratings is a free, open leaderboard — any project can claim its profile, submit disclosures, and get scored on the same Disclosure x Performance framework covered in this piece.

Universal Token Ratings: What the First 100 Scores Reveal

FIP.16 passed on April 24 with 98.06% in favor, one of the strongest mandates in Flare's governance history. It restructured FLR issuance, burn fees, transaction fees, staking weight and protocol revenue in a single proposal. 
Four months on, most of the roadmap is live, and the effects are now visible onchain. Inflation dropped in May, the same month FIRE began collecting FAssets minting fees and FXRP tag fees. The hard fork followed in July, while FDC attestation fees started accruing and were first collected into FIRE in August. FAssets redemption fees began routing in September.
Other components, notably MEV capture and Confidential Compute fees, are still staged for later by design. With the core pieces now in place, this is a good time to assess what has shipped and what the early data shows.
Flare - brief refresher
Flare is a layer-1 EVM chain, live since July 2022, built around one design choice: its oracles are part of the protocol rather than a service bought from outside. Two enshrined protocols do that work, the FTSO (Flare Time Series Oracle) for price feeds and the FDC (Flare Data Connector) for attesting to events on other blockchains and Web2 APIs, both secured by the same set of data providers weighted by staked FLR. That infrastructure is what makes FAssets possible, bringing assets from chains without smart contracts onto Flare, currently XRP minted as FXRP, and giving their holders somewhere to put them to work in DeFi.
What FIP.16 changed
Inflation came down, and so did the base it applies to
Annual issuance fell from 5% to 3% on May 14, with the yearly hard cap dropping from 5 billion to 3 billion FLR. Applied to an inflatable supply of roughly 87 billion, that puts gross annual issuance at about 2.6 billion tokens.
The rate cut is the visible half. The less visible half is that FIP.16 also shrank the base the rate applies to. Permanently burned tokens, unearned rewards sitting in penalty pools, and FLR held by FIRE are all excluded from the inflation calculation. That creates a compounding effect: as the burn address grows and FIRE accumulates, the pool that 3% is calculated against keeps getting smaller, producing fewer tokens each period.
Every transaction burns more
The July 14 hard fork raised the base transaction fee from 25 gwei to 500 gwei, a 20x increase that significantly lifted the burn rate (see chart in the data section). For users the cost stays negligible, at roughly 0.064 FLR for a simple transfer, which is invisible at any realistic transaction volume.
The mechanism is encoded and runs by itself with no governance decision, no treasury discretion and no allocation vote, making it efficient. Every transaction on Flare destroys FLR automatically, the tokens are removed permanently, and the effect scales linearly with usage. 
Staking carries more weight
Influence on Flare is measured in signing weight. Nothing the network produces is final until operators holding more than half the total weight sign off on it, and those same operators do three jobs at once: producing blocks, running the FTSO price oracle, and running the Flare Data Connector. Weight comes from FLR, either staked and locked to a validator on the P-chain (Flare's platform chain) or as wrapped FLR delegated on the C-chain (the EVM chain), which stays liquid.
FIP.16 now counts P-chain stake at 5x the weight of C-chain delegation. It also raised the maximum validator size from 200 million to 300 million FLR and set a 20% minimum delegation fee network-wide, giving operators room to grow and a floor under their economics. The core effect is to move influence toward committed staked capital, since delegated FLR can exit at any moment and locked stake can't.
Additionally, FIP.16 unified how that weight is counted. Before the change, FTSO anchor feeds counted C-chain delegation only, while FDC and block-latency feeds used a combined weight that treated staked and delegated FLR as equal. Now all three share one calculation, and it's the one weighting locked stake at 5x. In other words no part of the network is cheaper to attack than another.
FIRE gives protocol revenue somewhere to go
The Flare Income Reinvestment Entity (FIRE) is the structurally new piece. It's a governed pool with a primary mandate to reduce FLR supply through burns and open-market buybacks, and secondary mandates covering asset issuer rewards, dApp yield and liquidity, and Foundation sustainability. Its first operational use is replacing inflation-funded validator and staker rewards with revenue the network generates itself, which serves the same mandate from the other side: FLR that never gets minted is supply that never gets created.
The proposal names several revenue sources, and four fee streams are routing today. FAssets minting fees go to FIRE in full. 90% of FDC request fees are redirected into it, with the remaining 10% continuing through existing reward channels. 10% of FAssets redemption fees route in, with the majority staying with the agent and the collateral pool. And FXRP destination-tag registrations, a permanent piece of XRPL namespace that lets a minter route straight to their own Flare address, are paid to FIRE in FLR. Flare Smart Account fees, Flare Confidential Compute fees, and captured MEV will arrive in the upcoming phases.
FIRE is administered by the Flare Foundation initially. After the first year, the community can vote to move it into joint governance, a decision that requires 50% of total inflatable FLR supply to vote in favor. If it passes, four representatives are elected from providers running infrastructure on both Flare and Songbird.
The data so far
Staking responded fastest
Flare now has 21.5 billion FLR staked, up from roughly 16 billion in July. Staking's share of all staked-or-delegated FLR climbed from about 32% in April to roughly 46% by late August. Most of that move landed within weeks of the hard fork.
The burn scaled by an order of magnitude
Flare has burned 15.6 million FLR year to date through transaction fees, with more than 40% of that total burned since the July 14 fork and the subsequent increase in gas fees. That puts the current burn rate at more than ten times its pre-fork baseline.
Protocol revenue is live and early
FIRE has been collecting since May and has taken in $31,438 to date. FAssets minting fees are the largest contributor at $18,248 across 7,708 mints, followed by $12,676 in FDC request fees, which began claiming into the pool on August 18. Tag registrations add $505 and redemption fees $9. Two of the four streams pay in FLR and two in FXRP, so the dollar figure moves with prices as well as with volume. The trend is worth watching as FAssets and FDC volumes scale.
Conclusion
Four months in, the structural work is done. Inflation is lower and applies to a shrinking base, the burn runs an order of magnitude above where it started, staking has shifted toward capital committed to securing the network, and protocol revenue has a governed destination for the first time.
What's live today is the bootstrap phase, and the proposal sequenced it that way. The largest parts of FIRE's mandate, Smart Account fees, Confidential Compute, and protocol-level MEV capture, arrive as the underlying infrastructure ships. That means today's figures reflect a system still being switched on, with mechanisms designed to compound as usage grows. The machinery is built and running, governance approved it near-unanimously, and the first onchain evidence points where the proposal said it would. What comes next is volume.

FIP.16 passed on April 24 with 98.06% in favor, one of the strongest mandates in Flare's governance history. It restructured FLR issuance, burn fees, transaction fees, staking weight and protocol revenue in a single proposal.

Flare's tokenomics revamp: the transition to activity-linked economics

It is clear that the observable universe of prediction markets on DefiLlama converges to these two companies taking most of the pie. However, while they were once neck to neck, Kalshi has blown far past Polymarket and overtaken them as volume leader.

Are Prediction Markets Still Hot?

The past twelve months have been rough on centralized venues, with major exchanges such as BitMart and BitMEX announcing shutdowns of their operations. Additionally, regulation has added further pressure, with MiCA leaving many exchanges, such as Binance, unable to serve certain EU residents.
None of these events is catastrophic in isolation, and exchanges that wind down or can no longer serve their user base usually process withdrawals. But some don’t, and the ones that do often leave users spending weeks in a queue with no certainty about the outcome, and even worse, sometimes asking them to provide additional documents before accessing their funds. And that’s when “not your keys, not your coins” takes on its full meaning,  the moment users realize that having a non-custodial solution shouldn’t be an option but a baseline requirement.
This context makes it a reasonable time to revisit decentralized alternatives that cover the same daily operations a CEX offers. And for those who were there in 2021 and judged the experience as poor across most DeFi protocols, a complete revisit is definitely needed. In five years, for the ones that survived, significant progress has been made, and judging them based on 2021 memories no longer tells you much.
THORChain is one of the clearest case studies. In 2021, it was barely usable in practice because only a few chains were connected, support was limited to their main tokens, and slippage was high. But the protocol has matured, and in Q1 2026 it processed $2.82 billion in swap volume, generating $3.3 million in fees. What follows is a refresher on what the protocol actually is, and a walkthrough of the changes that got it there, demonstrating that some of the “OG” DeFi projects have actually found their market.
What THORChain actually is
THORChain launched in 2021 as its own blockchain with the primary purpose of being a crosschain decentralized exchange (DEX). For the first time in crypto history, users could swap native assets across chains in a fully permissionless manner. For example, a user can swap Bitcoin directly from their Bitcoin wallet and receive Ethereum in their EVM wallet, all without relying on a centralized entity or wrapped assets.
That design choice matters for three reasons. First, there’s no account to freeze. Users keep custody in their own wallets at all times, except during the swap itself, when funds briefly transit through vaults collectively secured by the node set, whose nodes don’t censor transactions. Second, no wrapped asset can trade away from the value of its underlying collateral, as can happen with IOUs, because nothing is wrapped. Third, there is no smart contract bridge securing a pool of locked assets, a structure that is usually vulnerable to hacks.
Mechanically, THORChain can be considered an AMM. Every pool pairs an asset against RUNE, making RUNE the settlement asset for both legs of any cross-chain swap: BTC into RUNE, then RUNE into ETH, all within a single transaction. Fees are paid in RUNE too.  However, this all happens in the background, and users don’t need to own RUNE to swap on THORChain or ever know RUNE exists. Beyond the liquidity pools, RUNE also plays a central role in network security, with nodes bonding RUNE to secure the network.
And because THORChain operates as its own blockchain rather than as a set of smart contracts deployed on someone else’s chain, every transaction settles transparently onchain and smart contracts/applications can also be built on top of the network (with the nodes’ permission).
THORChain Features
Chain expansion: covering what the rest of DeFi can't reach
In late 2021, THORChain natively supported six chains. Today, that number has grown to fourteen: Bitcoin, Ethereum, XRP Ledger, Tron, Solana, Monero, Zcash, Base, BNB Smart Chain, Dogecoin, Bitcoin Cash, Litecoin, Avalanche, and Cosmos Hub.
Beyond simply adding new chains, the strategy is clear and deliberate: provide decentralized interoperability to asset holders who currently have very limited options. THORChain is not chasing the easier EVM integrations commonly seen across bridges. Instead, it has prioritized ecosystems where native cross-chain liquidity is scarce.
That is why Monero and Zcash have been key priorities in recent months. As Monero continues to be delisted from centralized exchanges and remains inaccessible through DEXs, THORChain will become the first DEX to support native XMR trading.
As for the next integrations, Dash is expected to follow soon after, further cementing THORChain’s role as the main privacy DEX. Bittensor (TAO) should come next, a token with no native DEX liquidity today, making it another asset where THORChain can fill a clear market gap.
Swap execution: moving size without paying for it
Back in 2021, slippage was one of the main criticisms of THORChain, with price impact sometimes reaching 10% to 20%, especially when moving larger amounts. This is certainly one of the areas where THORChain has improved the most over time.
The main innovation, Streaming Swaps, launched in summer 2023 and did much of the work. Rather than hitting the pool once, a large order is split into a series of smaller sub-swaps executed across multiple blocks.
As the market matured and new designs such as solvers, intents, and CLOBs emerged, THORChain had to improve execution efficiency to compete. Limit Swaps and Rapid Swaps are two major innovations in that direction. The first lets users set a target price and execute the trade once the pool reaches the desired price. The second allows flows moving in opposite directions to be matched, helping users save on fees and settlement time.
On top of that, Memoless Transactions significantly improved the swap interaction model while also enhancing security. By abstracting away the memo previously required in a transaction, THORChain lets users specify their intent, receive transaction instructions, send funds to a vault, and receive the desired asset in another wallet. The key benefit is that users can swap from any wallet without connecting it directly, which is particularly convenient when using cold wallets.
The combined result is a protocol that routinely settles swaps in the millions with slippage around 0.2%, competitive with any major CEX. This has made THORChain the world's largest Bitcoin DEX by volume.
Rujira: the native DeFi Layer
In 2021, native DeFi infrastructure was still immature, and THORChain experimented with THORFi, bringing lending and yield products directly into the base protocol. Unfortunately, the model ultimately created too much balance sheet exposure, leading THORFi to be sunset in early 2025 and affected users to receive TCY, a claim on 10% of protocol revenue. The approach echoes Bitfinex’s response to its 2016 hack, when creditor tokens were later redeemed in full using subsequent cash flows. THORChain has since moved toward a more modular model.
In late 2024, the Kujira team, best known for products such as FIN and ORCA, brought its DeFi expertise and product stack into the THORChain ecosystem through a merger that gave birth to Rujira. The Rujira team developed the ability to create smart contracts on THORChain, which is colloquially called the app-layer. The first products launched in May 2025, with spot and perpetuals trading. Since then, the platform has continued to evolve with lending, concentrated liquidity, and automated strategies such as DCA. Capitalizing on past lessons, the rollout has been deliberately progressive, particularly when it comes to the size and exposure of each feature, with caps being increased gradually over time.
The key difference versus THORFi is architectural. Rujira sits above THORChain’s consensus layer rather than being embedded directly into the base protocol, meaning the risks associated with individual applications are isolated from THORChain itself. Just like the risk with smart contracts on Ethereum is isolated from the Ethereum blockchain.
Distribution: the protocol you've probably already used
One of THORChain’s key drivers of success is how easily it can be integrated. Any wallet or exchange can integrate it with minimal code and almost no ongoing maintenance, either through THORChain’s own API or directly. Integrating directly doesn’t require permission, a partnership, or a counterparty relationship.
That is how, in just a few years, Trust Wallet, Ledger, Tangem, Bitget Wallet, SafePal, and OKX Wallet have all gained the ability to route swaps through THORChain. If you have swapped BTC for ETH inside a wallet interface over the past few years, there is a good chance you may have used THORChain without even knowing it. The added benefit for integrators is that they can monetize these swaps by charging affiliate fees. The numbers have been substantial: Trust Wallet generated around $17 million in just under three years, while Ledger generated roughly $7 million in under two. 
Of course, with wallets charging an additional fee, users may also end up paying a markup on top of the underlying swap cost. To address more cost-sensitive and DeFi-native users, THORChain launched its own frontend swap.thorchain.org in December 2025, where swaps incur only the base protocol fee (and potential slippage) with no additional frontend fee.
Tokenomics: fully sustainable
Like most protocols at launch, THORChain initially relied on block rewards to incentivize nodes and liquidity providers and bootstrap the network. As protocol revenue grew, however, the need for these subsidies steadily declined.
In February 2025, THORChain removed block rewards entirely, meaning the network now relies 100% on revenue generated by real activity: primarily the base fee charged on swaps, alongside a share of activity generated through Rujira. Of those fees, 5% is burned, 5% goes to the dev team, 5% to the marketing fund, 10% to TCY holders, and the remainder is shared between nodes and liquidity providers.
At the same time, several governance decisions reduced RUNE’s total supply to 354 million. Today, all the tokens are circulating, and only around 5% of that supply remains in the reserve, used mainly as a rainy day fund.
The end result is a tokenomics model with limited dilution risk for investors, with real protocol activity driving revenue instead of relying on subsidies.
Worth a revisit
A handful of protocols from the 2021 cycle survived, kept shipping, and quietly turned into functioning businesses while the market stopped paying attention. THORChain is a prime example, deserving the OG moniker.
Five years produced a protocol supporting fourteen chains and counting, trading billions competitively, funding itself entirely through fees with no emissions, offering DeFi features with a CEX-like experience, and reaching users through the wallets they already use, often without them even realizing it.
For those who still dismiss older protocols and blockchains based on their 2021 experience, it’s time to reassess that judgment and take another look at THORChain.

The past twelve months have been rough on centralized venues, with major exchanges such as BitMart and BitMEX announcing shutdowns of their operations.

THORChain is not chasing the easier EVM integrations commonly seen across bridges. Instead, it has prioritized ecosystems where native cross-chain liquidity is scarce.

As CEXs shut down, DeFi deserves a revisit: the case for THORChain

Bitcoin spent the first eighteen days of August going nowhere, trading between $62,000 and $66,000. On August 19 it broke out, and by the 25th it had touched roughly $81,500. That's close to 27% from the breakout base, one of the sharpest weeks in the past five years.
The reflex reading of a move that fast is leverage. However, our data points to something else. Spot volume grew 50% faster than perpetuals, ETF liquidity was already flowing in two sessions before the breakout, and open interest fell in Bitcoin terms even as it rose in dollars. In simpler terms, this rally was bought, not levered.
Which raises the second question. If the move ran on cash rather than leverage, which venues did it run through?
It was bought, not levered
One way to assess whether a rally is spot-driven or leverage-driven is to compare spot and perpetual volumes. 
Through the first eighteen days of August, the market traded $6.02 in perpetual volume for every dollar of spot volume. During the breakout week, that ratio fell to $4.97, a 17.4% compression. The reason is straightforward: spot activity accelerated faster than derivatives with the average daily spot volume rising 153% over the period to $50.79 billion, while perpetual volume increased 109% to $252.59 billion. The gap suggests that the breakout was supported by a meaningful increase in spot participation rather than being driven primarily by leverage.
The ETF flows reinforce this observation. US spot Bitcoin ETFs recorded net outflows throughout the August 10-14 week, before reversing with two positive sessions just ahead of the breakout. The shift then persisted: since August 14, there has not been a single session of net outflows. Month-to-date net inflows now stand at $3.05 billion, making ETF demand an important trend to watch over the coming weeks.
As for perpetuals, we can’t ignore that volume nearly doubled over the period, while open interest in dollar terms climbed from about $21.9 billion just before the breakout to a peak of nearly $26 billion on August 24. But measured in BTC rather than dollars, the picture looks different. The market was carrying roughly 341,000 BTC in open positions before the breakout, compared with around 312,000 BTC today, an 8.5% decline.
In the end, perpetual trading remained elevated throughout the period, but open interest contracted in BTC terms as prices climbed and spot inflows returned. That combination demonstrates that the rally was supported by fresh capital rather than being driven primarily by leveraged bets. The next question is: who captured those flows?
Who captured those flows?
Tracking Bitcoin balances on centralized exchanges gives us a good view of where that fresh capital landed. Here, two venues stand out: Binance and OKX, which captured the bulk of the increase. Binance added $2.63 billion in BTC balances over the same period, double of OKX’s $1.27 billion. To put that into perspective, Binance alone absorbed nearly as much as the entire US spot Bitcoin ETF complex, which recorded $3.05 billion in net inflows over the same period.
Unsurprisingly, those inflows translated into trading activity. Both platforms saw spot volumes surge during the breakout, but Binance stood in a league of its own. Average daily spot volume climbed from $4.75 billion to $13.69 billion, a 188% increase, making it the only tracked exchange to exceed $10 billion a day. OKX operated at a more modest absolute scale, but its 234% growth was the strongest among major CEXs and well above the market average of 153%.
Together, Binance (46%) and OKX (11%) accounted for nearly 57% of the spot volume traded across the exchanges tracked, reinforcing the same picture seen in balances: fresh capital did not spread evenly across the market. Instead, it concentrated around a handful of venues, with Binance and OKX capturing most of both the inflows and the resulting trading activity.
What comes next?
With open interest peaking and then falling, much of the short squeeze seemingly behind us, and liquidity flowing back into the market, what happens next will largely depend on how spot and ETF demand evolves, a conclusion also shared by Binance Research.
That gives traders a clear signal to watch. If open interest starts rising faster than spot volume and the perpetual-to-spot ratio climbs again, the market will have shifted from one driven by fresh capital and more sustainable price discovery to one fueled by leverage again.

Spot volume grew 50% faster than perpetuals, ETF liquidity was already flowing in two sessions before the breakout, and open interest fell in Bitcoin terms even as it rose in dollars. In simpler terms, this rally was bought, not levered.

Bitcoin spent the first eighteen days of August going nowhere, trading between $62,000 and $66,000. On August 19 it broke out, and by the 25th it had touched roughly $81,500. That's close to 27% from the breakout base, one of the sharpest weeks in the past five years.

Bought, Not Levered: What Actually Drove Bitcoin's August Rally

RedStone

Hyperliquid has become the de facto venue for onchain perpetual futures, clearing roughly $220B in perp volume a month with over $13B in open interest. Thanks to the HIP-3 framework, anyone can launch a perpetual market for almost any asset, from crypto to equities to commodities, with the deployer taking on the oracle. This setup has drawn several teams building real-world-asset and equity markets to the network.
The appetite for these markets skyrocketed earlier this year, during the US-Iran conflict. With traditional oil markets closed over the weekends and news breaking around the clock, traders turned to Hyperliquid's crude perpetual to price the moves in real time, and oil briefly became one of the platform's most-traded markets. 
It was an early demonstration that onchain venues can carry price discovery when the traditional market is dark. That dynamic is not limited to commodity markets, with tokenized stocks and pre-IPO markets now trading on Hyperliquid around the clock.
The Rise of Pre-IPO Onchain markets
On August 24, a new market for tokenized Anthropic shares went live on Hyperliquid under the ANTH ticker. Since then, the market has generated $24.01M in trading volume. Open interest currently sits at approximately $3.3M.
The ANTH market on Hyperliquid is not the only venue pricing Anthropic ahead of its IPO. Binance listed its own pre-IPO perpetual on June 2, hours after Anthropic filed its confidential S-1, with up to 20x leverage. Coinbase followed in late June. The A.I. giant was valued at around $965B in its May 2026 Series H, the largest private funding round on record, and has since filed confidentially for an IPO with a listing reportedly targeted for later this year.
DefiLlama's own Pre-IPO dashboard tracks what the market prices Anthropic at right now: across the 18 venues it follows quoting an Anthropic perpetual, the average implied valuation sits near $1.96 trillion, more than double the Series H mark and a 42% premium over Anthropic's current secondary valuation of $1.38 trillion, itself up 717% over the past year.
Previous pre-IPO markets on Hyperliquid have already made the round trip, with varying degrees of accuracy. SpaceX traded as a pre-IPO perpetual for weeks before its June 2026 listing, and in May, the perpetual for chipmaker Cerebras came within a couple of percent of its Nasdaq open in the hours before listing. 
When things go smoothly, the perpetual settles once the company goes public, and the market moves on. But that isn't always the case, and what separates the clean close from the messy one is the data feeding the market, which is why choosing a reliable oracle is fundamental for any team launching HIP-3 markets.
RedStone has been supporting HIP-3 markets from day one with a purpose-built oracle stack for Hyperliquid's permissionless markets. For assets like ANTH that have no continuous public price, RedStone Live stays on when TradFi closes, switching to derivatives markets that trade around the clock on their own order flow.
How a pre-IPO market gets priced
On a pre-IPO market there is no public exchange listing the asset, so there is no authoritative price to read. The oracle has to assemble one from weaker signals: the market's own order book, the live bids and offers from traders taking positions, and data from other venues where the same exposure trades, including private-share marketplaces and onchain pre-IPO markets. 
Neither input is enough on its own. A reliable model combines both inputs into a single price through a liquidity-weighted mechanism. The more real volume backing the orders on the book, the more the price leans on it, and the thinner that volume is, the more it leans on the external aggregate. 
On the Anthropic perpetuals DefiLlama tracks, Binance and Bitget alone carry roughly three-quarters of total open interest. Coinbase's book is thin enough that its price would lean far more on the external aggregate than the order book.
External events like a new funding round or a stock split can also reset the price. If these inputs are not properly managed, they can lead to cascading liquidations.
On May 28, 2026, a SpaceX pre-IPO market on Hyperliquid fell from around $2,277 to $1,254 in under half an hour. The 45% drop came from the offchain feed that mishandled a 5-for-1 stock split, reading a routine corporate action as a price collapse. The market was too thin to absorb the bad data, so it cascaded into forced liquidations across roughly 400 traders before the price recovered. Nothing about the underlying company had changed.
RedStone handles corporate actions through a primary/fallback price state system with independent offsets, designed to distinguish a price move from a mechanical reset. On top of that, a 4-of-6 multisig quorum ensures no single node can move the price unilaterally, preventing oracle manipulation attacks.
The rise of Hyperliquid: growing fast, priced carefully
Before perp markets, trading that exposure was reserved for accredited investors. Even then, these markets trade on secondary platforms with steep minimums or OTC desks quoting in the millions. That changed with the arrival of synthetic perps on exchanges like Binance, Coinbase, and Hyperliquid, the latter of which runs without a centralized intermediary.
On August 19, at a gathering of blockchain and technology executives, President Trump stated that the CFTC is working to bring Hyperliquid into the US "in a fully compliant and legal fashion." Hyperliquid currently blocks US users, so an onshore path would widen its addressable market considerably.
If the CFTC does bring Hyperliquid onshore, a larger and less forgiving crowd shows up, trading at size on markets where the right price is not easily available. In this scenario, getting the data layer right is probably the most important task for any team building HIP-3 markets.

Hyperliquid has become the de facto venue for onchain perpetual futures, clearing roughly $220B in perp volume a month with over $13B in open interest.

Thanks to the HIP-3 framework, anyone can launch a perpetual market for almost any asset, from crypto to equities to commodities, with the deployer taking on the oracle.

The rise of HIP-3 pre-IPO markets, and how to price them

Real-world assets (RWAs) have proven to be one of crypto industry’s most bustling segments over the recent years, growing from $10.5 billion to $36 billion in total value across all chains year on year. While issuance has been growing across all chains in 2026, its uptick on Stellar is especially significant, rising from $525M to $2.6B+, an increase of more than x5 within just 8 months, making its RWA segment one of the fastest-growing ones across the industry. Behind this expansion is a multi-layered infrastructure streamlining RWA issuance across several key verticals, including compliance and DeFi utilization.
The backbone: Stellar’s infrastructure
Stellar offers institutions several key architectural features enabling compliant RWA launches. These include a set of native KYC standards, such as SEP-9, which standardizes the KYC information across protocols on Stellar and is integrated with the deposits and withdrawals API (SEP-24) as well as the SEP-12 KYC API. 
Besides the native KYC framework, Stellar also features a dedicated standard for issuing regulated assets (SEP-8), which enables the issuer to sign off any transfers for the asset, an important feature for institutional-grade assets from the AML perspective. Assets leveraging this standard include WisdomTree’s WTGXX (Treasury Money Market Digital Fund), CRDT (Private Credit and Alternative Income Digital Fund), and its physical-gold-backed Gold Token, all distributed through WisdomTree Prime. Issuers are also free to explore bringing their RWAs onchain via the general fungible token standard SEP-41.
The execution: Stellar’s validator network
Stellar’s native standards enable the assets to be issued and managed with compliance, but execution makes them usable. The Stellar Consensus Protocol is a federated Byzantine agreement model where validators pick their quorum slices, or validator subsets tasked with confirming a transaction; in other words, finality, does not require a signoff from the entire validator set, which enables transactions to settle within 5 seconds. Running the validators is a select group of Tier-1 validator organizations (Franklin Templeton, Blockdaemon, MoneyGram, Figure, Range, etc), all running multiple nodes that are spread between different physical locations.
 
Stellar’s RPC infrastructure works as the gateway for bringing data onchain to the network, and it also has a few peculiarities. It features two main components: a REST API for routine operations such as payments and setting onchain permissions, and Soroban RPC, a JSON-RPC interface specifically for smart contract interactions. Multiple RPC operators, including Alchemy, Validation Cloud, QuickNode, or Ankr, enable builders on Stellar to leverage their RPCs instead of running their own nodes.
The DeFi unlock: The pricing infrastructure on Stellar
Issuance brings an RWA onchain, but in itself, that doesn't grant it the same composability as a crypto-native asset. To trade on DEXes, work as collateral, or otherwise function in DeFi, an RWA token requires a price feed delivered in a standard onchain shape. On Stellar, pricing is handled under the SEP-40 standard, which sets the unified interface for consuming price data, sparing builders the need to set up custom adapters for every price feed and integration.
While a crypto token is priced based on deep-liquidity markets trading 24/7, pricing most RWAs represents a much larger challenge. Most of them, such as stocks, either trade within designated windows on business days, or, like gold and the largest indices, 23/5; additionally, a variety of RWAs, such as money market funds, aren’t priced based on trading at all, but rather based on their underlying NAV, which is reported by fund manager. That complexity is a large part of why RWAs remain broadly underutilized in DeFi, and Stellar illustrates the gap clearly: its RWA active market cap has grown to $2.64 billion, yet DeFiLlama’s RWA dashboard shows $0 in tracked DeFi Active TVL for those assets on Stellar, against overall Stellar DeFi TVL of roughly $238 million.
 
The challenge has birthed a variety of oracles specializing in RWAs. Among them, several have already adopted the standard and are providing SEP-40 price feeds for RWAs on Stellar. As an example, RedStone is pricing 10 RWAs on Stellar, varying from sovereign debt, both US and non-US, to tokenized gold and collateral loan obligations. RedStone price feeds have unlocked these assets for Templar Protocol, which is running lending markets against Centrifuge-tokenized deJTRSY and deJAAA by Janus Henderson Investors, USTRY (tokenized T-bills), and CETES (tokenized sovereign debt of Brazil), 
The outlook: DTCC is the next milestone
Stellar’s next RWA milestone may already have a name. The Depository Trust and Clearing Corporation (DTCC), the financial company that automates, clears, and settles trades for global financial markets, has announced plans to tokenize the assets in custody of its subsidiary DTC on Stellar, aiming to start next year. DTCC is a clearing house powering the global stock trade and processing quadrillions of dollars in securities; DTC, or the Depository Trust Company, custodies more than $114 trillion worth of assets, per DTCC’s own figures.  
The momentum behind Stellar’s RWA sector has already clearly manifested itself, and its partnership ecosystem sets it up for further growth. The main challenge ahead is not bringing more RWAs onchain, but rather unlocking what’s already out there for use in DeFi, a challenge resolved first and foremost via robust pricing infrastructure.

RWAs have proven to be one of crypto industry’s most bustling segments over the recent years, growing from $10.5 billion to $36 billion in total value across all chains year on year. Behind this expansion is a multi-layered infrastructure streamlining RWA issuance across several key verticals.

Real-world assets (RWAs) have proven to be one of crypto industry’s most bustling segments over the recent years, growing from $10.5 billion to $36 billion in total value across all chains year on year.

The Infrastructure Behind Stellar’s RWA Boom

Shane Molidor

FORGD

Crypto has never had a shortage of badges. There are audit stamps, transparency checklists, self-reported scorecards, and way too many superlatives. The industry has produced plenty of ways for projects to signal credibility. But many of these signals are deeply flawed. The industry needs a trusted, go-to metric to measure credibility. One that's verifiable and hard to game.
That’s what we’re launching today.
DefiLlama has teamed up with Forgd to launch the industry’s first continuously updated metric for scoring token disclosure and performance: Universal Token Ratings (UTR). We’ve intentionally designed the system so that a high rating is difficult to earn, and even harder to keep. 
Every rated project gets a standardized 0–100 score and a familiar AAA–CCC letter grade. Both recalculate as the market moves. If liquidity dries up, a new batch of tokens unlock, or a market maker falters under stress, the score will update automatically, without any human oversight required. 
Why UTR was built
DefiLlama exists to surface honest, unbiased data. For years, that has meant tracking what's verifiable: TVL, volumes, fees, unlocks, and other onchain analytics.
There is often a gap between what projects say and the realities of the market. A project can publish a beautiful tokenomics page and still trade with paper-thin liquidity. It can disclose a market maker engagement, but quickly let it lapse. Static transparency initiatives can't catch this, because they don’t continuously monitor the tokens they have rated. 
Universal Token Ratings close that gap by design. The methodology is built on two independently scored axes, each graded 0–10:
The Disclosure Axis (D) measures what a project reports about itself — team, tokenomics and vesting, legal and treasury structure, and service provider engagements including market makers, auditors, and exchanges.
The Performance Axis (P) measures whether live market behavior actually supports these claims by tracking liquidity depth, bid-ask spreads, volume, exchange coverage, derivatives conditions, market maker adherence, and on-chain tokenomics delivery.
The final score is deliberately multiplicative: UTR = D × P.
A project with perfect disclosures and a broken market can't score well. Neither can a project with deep liquidity, but missing disclosure statements. Only projects strong on both dimensions can earn top grades. Verified on-chain events like token unlocks, new listings, and changed market maker mandates flow straight into the score with no refiling and no human in the loop.
The incentive shift
Here's the part we care about most: how this rating will influence behavior across the industry.
When a rating is a one-time exercise, the rational strategy for token projects is to optimize for that audit — pass the review, frame the badge, and never worry about it again. Nothing about that structure rewards staying transparent or keeping markets healthy.
A continuous rating flips the incentive. When a token’s grade can deteriorate in near-real-time, transparency stops being a launch-week deliverable and becomes an operating principle. Projects are rewarded for maintaining liquidity, honoring vesting schedules on-chain, keeping market maker engagements active, and refreshing disclosures as facts change. Otherwise their stellar AAA-rating will drift toward CCC, and investors will quickly abandon the project.
As Shane Molidor, Founder and CEO of Forgd, puts it: “Crypto has had transparency checklists, but never a continuously updated rating system. A UTR score is something you have to keep earning.”
That's the kind of accountability institutional capital and regulators have been asking digital assets to develop. The industry needs something that functions more like a bond rating than a badge. 
Where the data comes from
The Performance Axis is powered by our data partnerships with Forgd, alongside Forgd's proprietary market maker monitoring infrastructure, which tracks live engagements across more than 500 token projects and 35 market-making firms. This builds on a partnership that started earlier this year with the integration of Forgd's market maker leaderboard on DefiLlama.
The full UTR methodology — every category weight, scoring rubric, and grading threshold — is published openly. If you want to know exactly why a project scored what it scored, you can trace it.
What you'll find at launch
Starting today, you can explore ratings for 100+ tokens directly on DefiLlama, with the ability to:
Track newly launched tokens from day one
Filter by valuation, sector, and exchange listing tier
And drill into full category-level breakdowns showing exactly where a project is strongest, and where the gaps are
The bigger picture
Ratings calculated from live market data are a natural extension of what DefiLlama has always done: give millions of users, researchers, and journalists a neutral, free, open source of truth. Universal Token Ratings take that a step further, drawing a line from a project’s disclosures and public statements to their measurable market performance.
We think the projects doing things right will welcome this. A market where credibility is continuously verified is a market where transparency confers a durable competitive advantage.
Explore Universal Token Ratings on DeFiLlama and learn how the scoring works.

DefiLlama has teamed up with Forgd to launch the industry’s first continuously updated metric for scoring token disclosure and performance: Universal Token Ratings (UTR). We’ve intentionally designed the system so that a high rating is difficult to earn, and even harder to keep.

A Rating You Have to Keep Earning: Forgd and DefiLlama Introduce Universal Token Ratings

On 4 May 2026, Tydro suspended every market it operated. Chaos Labs, then its sole price data provider, had disclosed an attack on its infrastructure and recommended a halt. No incorrect price reached the markets, and no user positions were impacted. About 48 hours later, Chaos Labs approved a restart, but the markets remained closed as the operator migrated the feeds.
The incident exposed something structural. Tydro runs Aave v3, and a white-label Aave deployment does not inherit the system that would normally have handled a moment like that. Three functions go missing. Tydro bought two of them back within a week, and the third is why its v2 release will carry an independent Credora rating.
What a white-label Aave deployment inherits
Tydro is a non-custodial lending protocol on Ink, Kraken's Ethereum L2, carrying $56.15M in TVL and $48.67M in active loans across 11 pools per DefiLlama, roughly half of the chain's DeFi TVL.
The ARFC responsible for approving the deployment establishes the terms: Aave DAO will license the code and will take 'a share of all the revenue generated by the platform'. This share should be equal to or greater than the equivalent of a Reserve Factor of 5% calculated on the borrow volume in all pools.
That is where the relationship ends. Since the same ARFC states "the instance will initially be centrally governed by the Ink Foundation without a governance token", Aave DAO has no voting power over listings or parameters. BGD Labs, Aave's core development team, put it plainly: "the decision to include/not to assess in the listings set is of the friendly fork manager, not ours.” Bad debt is not covered by the backstop since "this instance will not be protected by the DAO-run Umbrella". The provision allowing trusted parties to modify the guardrails without a governance vote does not apply here.
This is the general form of the fork risk in Aave v3. The contracts are the same, and there is no oversight, leaving the operator with three functions they can replace: parameter stewardship, a loss backstop, and an external viewpoint on whether the markets are correctly priced from the start.
Of the 17 recorded security incidents across Aave forks tracked by DefiLlama, not including the Tydro incident, nine (just over half) were oracle-manipulation attacks, accounting for $52.0M of the $119.8M lost.
Aave's own governed instances have lost $918K in two incidents within the same dataset. One notable incident was a March 2026 oracle-parameter misconfiguration, which caused about $27M in liquidations. Aave DAO then voted to reimburse affected users, recovering $862K of the loss. No fork in the dataset has had an equivalent safety measure.
What an operator can buy
Two of the three are procurable. Feed redundancy is a vendor decision, and Tydro made it in May: Chainlink Data Feeds as primary, RedStone as the redundancy layer, migrated under a 48-hour timelock, with markets reopening around 10 May behind a grace period that halted liquidations while borrowers repaid or topped up.
That decision isn't the norm: across all DefiLlama-tracked protocols with an oracle on record, only 190 of 1,116 (17%) run two or more concurrent price feeds; most operators, forked or not, rely on a single source.
Transaction screening is software, and Hypernative's Transaction Guard now checks multisig transactions across Ethereum, Optimism, and Ink against policy before execution.
Both closures were the result of budget constraints and the use of engineering time; any operator possessing either of them could carry out the same action.
The part that has to come from outside
The third function does not work that way.
The reason that Aave's listing votes and Risk Stewards were useful had nothing to do with the mechanism itself; it was because the mechanism was situated outside the entity that held the position. A deployment that evaluates its own markets and then releases the grade is engaging in marketing. The value of a risk assessment lies in the fact that the assessor has no interest in the outcome, a property which no operator can create internally regardless of how much money they spend.
The gap remained open into May, and the v2 release is what closes it.
Adding a DeFi risk rating layer
Assets and markets with exposure to kBTC and USDC will carry a Credora rating on an A+ to D scale, shown in the Tydro interface, under the same methodology applied to every protocol on that scale.
The two metrics behind the letters differ by object. Markets are scored on a Probability of Significant Loss: the annualized probability that a market will lose more than 1% of its principal over a one-year horizon. That threshold suits a lending market, where the realistic failure mode is a partial loss which lenders absorb rather than a binary default. The estimate comes from 100,000 Monte Carlo simulations of price movements, liquidity conditions, and correlated shocks affecting the market's collateral, the probability being determined by the proportion of simulation paths that exceed the threshold. The five factors involving smart contract risk, liquidity risk, oracle risk, counterparty risk, and collateral quality risk are each modeled separately and thus affect the final rating.
Assets are scored on a Probability of Default, built by asset type: direct ratings where reserves are themselves rated tokens, published cumulative default rates for instruments such as T-Bills, and simulation for crypto-native reserves.
Both resolve onto one letter scale. For example, an ‘A’ corresponds to a PSL of roughly 0.1% to 0.4% annually, in the range of an investment-grade BBB/BBB- in traditional credit. Because the same scale covers markets on other protocols, a position here can be set against one on a protocol with entirely different mechanics, on a single axis, against a methodology published in full.
The scope is limited. A rating measures the downside associated with a particular market or asset that has been rated. It does not constitute an audit of the underlying protocol or chain, it does not claim that any position is free of risk, and it is not a replacement for a backstop. It is an assessment, and assessments do not cover losses.
The takeaway
A white-label Aave deployment inherits the code but not the oversight, and the three functions it leaves behind are not all equally replaceable. The feed and screening functions are acquired as external services. An independent assessment of the risk involves a party that does not report to the operator. It is the one function of the three that cannot, in principle, be set up in-house.
On Tydro v2, that party is Credora, rating every market and asset in scope.

On 4 May 2026, Tydro suspended every market it operated. Chaos Labs, then its sole price data provider, had disclosed an attack on its infrastructure.

On 4 May 2026, Tydro suspended every market it operated. Chaos Labs, then its sole price data provider, had disclosed an attack on its infrastructure and recommended a halt.

White-label Aave deployment: the risk function an operator cannot rebuild alone

The Landscape Of Stocks On The Blockchain
In the past year the industry has seen the growth of stocks on blockchain rails, both on spot and on perpetual DEXes. Tokenized equities' on-chain active market cap has gone from roughly $429M a year ago to $3.76B today — a ~9x increase — while a parallel and much larger market has emerged in leveraged synthetic exposure: perp venues referencing individual stocks, ETFs, and indices, which have moved $105.7B in the last 30 days alone.
These two verticals are new and growing fast — but as seen many times in crypto, growth in market capitalization and trading volume don’t necessarily correspond to long-term anti-fragile adoption. That’s the question this piece sets out to answer: do stocks on crypto rails actually have product-market fit so far? 
A Brief Explanation Of Stocks On-Chain And Stock Perps
Broadly speaking, there have been two categories of adoption for stocks on crypto rails and they operate under fundamentally different mechanics. 
On-chain tokenized equities are blockchain-native representations of real-world stocks. A regulated issuer holds the underlying shares with a qualified custodian and mints tokens that track those shares 1:1. The token lives on the blockchain and can be traded on decentralized venues, and in some cases, be redeemed for the underlying stock. The market currently stands at roughly $2.42 billion in active market capitalization.
Stock perpetuals are the other side of the coin. These perpetual futures contracts reference individual stocks, ETFs, or equity indices, and are traded on derivatives exchanges. But they carry no claim on the underlying share. What they offer instead is leverage — typically 2x to 20x — and 24/7 access to price exposure. 
In the last 30 days alone, perp venues referencing equities moved approximately $105.7 billion in notional volume. For context, the entire DeFi derivatives market across all assets (crypto pairs, forex, commodities, and equities) processed $489.5 billion over the same period. Stock perps alone represent roughly 21% of that.
Breaking Down The Landscapes 
On-Chain
Three major entities dominate the on-chain tokenized-equities landscape: Backed/Kraken's xStocks, Ondo Global Markets, and bTech Holdings.
Backed Finance, the issuer of xStocks, was acquired by Kraken in a deal announced December 2, 2025. Since its debut, xStocks has passed $10B in combined exchange and on-chain trading volume. On-chain, xStocks currently carry $866M in tracked market cap across 191 distinct assets.
Ondo Global Markets launched September 3, 2025 with 100+ tokenized U.S. stocks and ETFs on Ethereum, backed by securities held at U.S.-registered broker-dealers. It has since scaled to 415 total tracked assets, with $917.2M in tracked on-chain market cap, making it the largest on-chain issuer of tokenized equities.
bTech Holdings, a Binance Group affiliate registered in the Abu Dhabi Global Market, issues bStocks — the tokenized-equity product available on Binance and self-custodied on BNB Chain. bTech is the third-largest issuer, with $564.5m in on-chain market cap — 18% of the tracked market.
Issuer
Onchain Mcap
Share
Assets
Ondo Global Markets
$917.2M
30.26%
415
Backed (xStocks)
$866.0M
28.57%
191
bTech (bStocks)
$564.5M
18.62%
35
Perpetuals
The tokenized-equity perp landscape looks structurally different from the on-chain spot market — one venue has already mostly won. Across the 19 venues DefiLlama tracks for stock/index/ETF-referencing perpetuals, a single venue, trade.xyz on Hyperliquid, accounts for $93.5B — an 88.47% share.
Venue
30d Volume
Open Interest
30D Fees
Notes
trade.xyz
$93.55B
$2.90B
$23.39M
Hyperliquid HIP-3 deployer
Variational
$2.75B
$216.2M
$0
Standalone
Ondo (perp desk)
$2.37B
$35.7M
$415K
Permissioned, separate from Ondo Global Markets spot
Extended
$2.01B
$13.2M
$503K
Standalone
Trade.xyz isn't a standalone matching engine; it's a HIP-3 deployer running on Hyperliquid’s HyperCore infrastructure — builders can deploy their own perpetual markets on top of Hyperliquid under its HIP-3 framework. Trade.xyz's RWA flow alone equals 61.5% of all volume moving through Hyperliquid's derivatives engine over the same window ($118.4B against Hyperliquid's $192.5B 30-day total). RWA perps aren't a side product for Hyperliquid anymore, they're a majority of its printed volume.
This is very clearly a winner-take-most market. Beyond trade.xyz, the top five venues capture 96.86% of all 30 day volume, leaving no space for any other competitors.
Analyzing Product-Market Fit Across Both
To measure product-market fit, we will be examining both on-chain and perpetuals across the same set of metrics:
Name
What it measures
Methodology
Size
Growth over time and capital commitment
On-chain: active/on-chain market cap. 
Perps: open interest. 
Turnover
Is that capital moving, or just sitting there
On-chain: 30d volume ÷ market cap. 
Perps: 30d volume ÷ OI
Concentration
Is it broad-based, or a couple of players
On-chain: issuer/asset share 
 Perps: venue share 
Consistency
Is activity steady, or a few spike days doing all the work
Coefficient of variation on daily volume. 
Size - Growth Of The Asset Class 

On-chain tokenized equity market cap grew from $22.6 million in April 2025 to $2.42 billion by August 2026 — a 99x increase over 16 months. This incredible rate of change is only outmatched by the growth on the side of perpetual open interest, which grew from roughly $162.6 million in December 2025 to $3.25 billion by August 2026, a 20x increase in under eight months. However, when comparing the two during a common period — from the launch of perpetual open interest markets in December 2025 until now — perps grew by 20x, while on-chain equities only grew by 4.5x. 
Two things stand out from this apples-to-apples comparison:
Perps are growing faster than on-chain issuance, and have overtaken it in absolute size. Perp open interest started with less than a third of on-chain market cap and finished ~44% larger, in a shorter span of time.
On-chain growth is real but slower than the headline multi-year figure suggests. The ~99x figure spans the whole series from a very small April 2025 base (~$22.6M) — a base low enough that even modest dollar inflows produce large multiples. The 4.5x common-window figure is the fairer read of recent on-chain momentum next to perps.
While perps are growing faster than on-chain, one thing is clear: both are growing at astonishing rates that only products with PMF can exhibit.
Turnover - Capital Measurement Of Both Asset Classes

While size tells us how much capital each market has attracted, turnover tells us what that capital is actually doing — whether it's being actively traded or simply parked. We measure it on a consistent 30-day window for both sides (30-day volume ÷ active market cap on-chain, 30-day volume ÷ open interest for perps), scoped to the same universe (public equities, equity indices, and equity ETFs):
Metric
On-Chain
Perps
Total size (mcap / OI)
$2.36B
$3.26B
Total volume (30d)
$11.25B
$105.7B
Turnover ratio (native window)
4.66x/30d
32.39x/30d
Turnover ratio (daily-equivalent)
0.155x/day
1.08x/day
On a properly matched 30-day basis, perps turn over capital roughly 7x faster than on-chain tokenized equities per day. Despite holding a broadly comparable capital base (perp OI is only ~38% larger than on-chain active mcap - $3.26B vs $2.42B), perp venues moved $105.7B over 30 days, more than 9x the on-chain figure over the identical window.
However, the lack of turnover on-chain doesn’t necessarily mean the market is fake. The more likely answer is that the two markets are built for, and attracting demand for, genuinely different trading profiles.
Perps exist to be traded. A leveraged derivative position accrues funding costs and requires active management — margin rebalancing, rolled positions, funding paid. One can say that high turnover is the mechanical requirement of the product. The $111.4B moved in 30 days is unambiguous evidence that people are actively using these venues as trading instruments, at real scale.
On the other hand, on-chain tokenized equities exist to be held. A wrapped share is functionally closer to a brokerage position than a trading chip. And at 15.5%/day, while modest next to perps, is still decent. 
Concentration - Measuring The Playing Field
Metric
On-Chain Issuers
Perp Venues 
Entities in scope
21 issuers
19 venues
Top-1 share
30.26%
88.5%
Top-3 share
77.45%
96.9%
Both markets are concentrated, but perps are in an entirely different tier. On-chain issuance sits at the highly concentrated threshold, with the Top-3 gaining 77% market share. Yet the top perp venue has a market share that represents almost 90% of all volume traded via equity perps. This is near-total dominance. 
On-chain: Ondo Global Markets and Backed Assets hold 30.26% and 28.57% of scoped mcap respectively ($917.2M and $866M), with bTech (Binance) at 18.62% - three players covering 77.45% of the market.
Perps: trade-xyz alone holds 88% of open interest ($2.9B of the $3.26B total) - essentially one venue is the perp stock market by this measure.
Volume tells a similar but slightly different story. By 30-day volume (rather than size), perp concentration is nearly identical — Trade.xyz still does close to 90% of volume — so its dominance isn't just in open interest. Trade.xyz is where the trading happens. 
On-chain volume concentration is heavily two-sided. Of the issuer groups with any tracked 30-day volume, bTech Holdings and Ondo Global Markets alone account for 98.6% of the market (64.8% and 33.8% respectively), with everyone else combined barely registering. Notably, bTech is the smaller issuer by market cap but drives most of the trading activity, probably because it is on Binance, the world’s largest crypto exchange, and is featured as Binance’s on-chain stock solution. 
Consistency - Whether activity is steady, or spiked across a few key days
On-chain tokenized equity volumes are choppy from a day-to-day perspective, but the 7-day average line traces a clear staircase upward. The perp volume has the same overall shape, but much bigger swings — the bars are spikier and less even throughout. 
Overall, perps are lumpier day-to-day, but on-chain concentrates into fewer big days. Both markets tell the same underlying story: volume is trending up. The 7-day average on-chain volume roughly tripled from $90-120 million a day in March and April to $330-400 million by early August. Perps show an identical shape at roughly ten times the scale, with the 7-day average rising from $1-1.2 billion a day in March to a peak of $4.86 billion in July. This shows that it’s not a few spike days doing all the work but expansion with some volatility layered on top. 
Looking deeper, we can see that half of all perp volume happens in just 31 days out of 150. On-chain takes 38 days to accumulate half of all current on-chain volume. This draws similar conclusions to the concentration section: a few big pairs on both these platforms do outsize volume on certain days.
Concluding Thoughts 
For both on-chain and perpetual markets, all the data points to a few conclusions:
Perps are ahead of on-chain tokenized stocks on every measurable dimension of product-market fit. Perps are larger, faster, and spikier. On-chain is smaller, slower, and steadier. 
Activity is concentrated in a small number of specific issuers/products for both.
But ultimately, they both seem to have product-market fit, but at different scales. Both rails are real and growing; one faster than the other.
Across four metrics - size, turnover, concentration, and consistency, the data points the same way. Stock perps have more capital at work, that capital turns over faster, and the market is more active day to day. On-chain tokenized stocks are smaller, slower, and steadier. In both cases, you have users showing up and using something actively at scale - but the signal is much stronger for perps than for on-chain tokenized stocks. 
However, the difference can be boiled down to the fact that the two markets are not the same thing. A perpetual swap is a trading swap, while a tokenized stock is ownership infrastructure. The perp attracts a user who wants exposure and wants it now — leverage, 24/7 access, no custody of the underlying asset. The on-chain user, however, wants to hold an asset that can sit in a wallet, be used as collateral, or be redeemed. The former user exists at large in the financial markets, while the latter is still growing. 
The data supports this. The turnover numbers tell you that on-chain capital turns over ~4.6x every 30 days. Thus, it’s a user base that doesn’t want to trade, but rather is consistent with money that mints, sits, and gets used as exposure or collateral. Perp turnover numbers of ~32x every 30 days is consistent with behavior at a typical trading/speculation venue. So they both have different economic functions — low turnover for a holding instrument, high turnover from a leverage/speculation instrument. 
Both are growing,. But for perps, it is quite clear they’ve already reached product market fit. On-chain tokenized stocks, meanwhile, are seeing real, growing capital, but there’s some ways to go before we can conclude that they’ve enjoyed long-term anti-fragile adoption. , they have both shown incredible signs of success from where they began.

In the past year the industry has seen the growth of stocks on blockchain rails, both on spot and on perpetual DEXes. Tokenized equities' on-chain active market cap has gone from roughly $429M a year ago to $3.76B today — a ~9x increase.

These two verticals are new and growing fast, but as seen many times in crypto, growth in market capitalization and trading volume don’t necessarily correspond to long-term anti-fragile adoption. Do stocks on crypto rails have product-market fit so far?

Stocks On The Blockchain - Do They Have Product-Market Fit?

Gate's derivatives-heavy trading book sits on top of a reserve base that's both verifiably substantial onchain and, by the exchange's own broader accounting, comfortably over-collateralized. Here's how the two pictures fit together.
DefiLlama tracks 87 centralized exchanges, with a combined $31 billion in 24-hour spot volume and $128.5 billion in 24-hour derivatives volume — a total market of roughly $159.5 billion per day. On the reserve side, the same 87 exchanges collectively hold $212.9 billion in tracked "clean" reserves (balances excluding an exchange's own native token).
Trading activity is heavily concentrated at the top. Binance alone processes $50 billion in daily volume, about 30% of the entire tracked market. OKX does $17.7 billion, and Gate follows with a combined 24-hour volume at $9.5 billion.

Gate As a CEX — How It Stacks Up
As shown above, Gate ranks third among the top 15 tracked exchanges by 24-hour spot volume, sitting just behind OKX and ahead of Bybit:
Spot volume: 3.85 billion per day, 113% larger than Bitget(1.80B), and about 8% ahead of Bybit ($3.57B)
Spot share of the top-15 cohort: ~6.7% of the combined 24h spot volume across these 15 exchanges
Gate's position among the top three exchanges by spot volume reflects strong spot-market activity in its own right.
Gate's spot ranking above tells only part of the story. That same day, Gate's derivatives volume was $21.7 billion, roughly 5.6x its spot volume, bringing its combined spot + derivatives volume to 25.6 billion.On a combined basis,Gate trails Bybit (34.6B combined, ~26% larger) but still runs about 20% ahead of Bitget's combined volume ($21.3B). This confirms Gate's overall trading activity leans derivatives-heavy.
Exchange
30d Net Inflow
OKX
+$1.35B
Gate
+$308.1M
MEXC
+$24.5M
Bybit
-$91.0M
Bitget
-$116.0M
Binance
-$142.4M
On top of that, Gate's on-chain tracked reserves have seen sustained net accumulation over the past month: DefiLlama data shows Gate pulled in $308.1M in cumulative net inflows over the trailing 30 days, with positive inflow days outnumbering outflow days across that window. That places Gate 2nd of 6 major exchanges we tracked side-by-side (behind only OKX's $1.35B) — and notably ahead of Binance (-$142.4M), Bitget (-$116.0M), and Bybit (-$91.0M), all of which saw net outflows over the same period. 
Gate’s Reserve & Coverage Ratio 
From Defillama data, Gate currently holds $4.61 billion in onchain tracked reserves (this excludes their GT token), placing them seventh of the 87 exchanges DefiLlama tracks. This is substantial and puts them just behind Bitfinex. 
Asset
Amount
USD Value
BTC
19,942.19
$1.28B
GT (own token)
108.8M
$727.9M
WETH
357,212.65
$709.0M
USD1
370.7M
$370.5M
(stablecoin, unlabeled)
312.8M
$312.2M
DOGE
2.30B
$161.3M
stETH
61,599.82
$115.7M
USDT (spUSDT)
105.6M
$108.4M

From the table above, we can see that Gate’s on-chain reserves are well diversified and aren’t concentrated in 1 or 2 assets. In fact, they are comfortably collateralized and have their holdings in many different assets, which is good to see for an exchange. 
We would also like to highlight that Gate has reserves beyond what we can observe onchain. On July 27, 2026, Gate published a report that announced an overall reserve coverage ratio of 117%, spanning ~500 user asset types, including reserves of 26,775 BTC (24.2% over-collateralized) and 456,798 Ether (22.02% over-collateralized). 
This follows reports in June and January that detailed the following figures: 
On June 22, 2026, Gate claimed a coverage ratio of 115%, with 25,292 Bitcoin and $782 million in USD1
Asset
User Holdings
Platform Reserves
Coverage Ratio
Overall (all ~500 assets)
—
—
117%
BTC
21,557 BTC
26,775 BTC
124.2%
ETH
374,348 ETH
456,798 ETH
122.02%
GT (native token)
—
—
131.13%
XRP
—
—
116.5%
Stablecoins (USDT+USDC+USD1+GUSD)
1.336B tokens
1.59B tokens
118.97%
And so — although there seems to be a discrepancy between DefiLlama's centralized exchange data and Gate's own reserve reports, that's to be expected —there are subtle but important differences in how the organizations track reserves. Here's how to read them together:
The DefiLlama table measures only onchain wallet balances for addresses Gate has disclosed to us. Those balances are independently verifiable — we can see the Bitcoin, WETH, USD1, and other assets sitting in Gate's tracked wallets in real time. But it's necessarily a floor, not a ceiling: it only counts what's onchain and disclosed, so it structurally cannot capture off-chain custody, undisclosed wallets, or the full ~500-asset scope Gate reports internally.
Gate's Proof of Reserve reports measure a self-attested comparison of platform reserves against user liabilities, across roughly 500 asset types, something no blockchain explorer or onchain tracker (including DefiLlama) can independently verify, because liabilities (what Gate owes users) simply aren't an onchain-observable fact. Only the exchange itself knows that number.
This is why Gate's self-reported total reserve figures run meaningfully ahead of DefiLlama's onchain read. The January report claims $9.4 billion in total reserves, while DefiLlama reports just $4.6 billion in clean reserves. The two numbers are measuring different parameters — DefiLlama for the verifiable, onchain floor, and Gate’s reports for a broader total that includes offchain assets and custody arrangements. 
Taken together, the two data sets are complementary rather than competing. DefiLlama independently confirms that Gate's disclosed onchain reserves are real, substantial, and well-diversified. No single asset dominates the tracked balance sheet, and Bitcoin (the largest holding) makes up just 23.1% of onchain tracked value. 
Gate then supplements that with its own report, which includes its full, undisclosed reserve base and its liabilities. This fuller picture shows that coverage runs comfortably above 100% across every major asset category, according to Gate. The company’s latest report also shows a particularly comfortable reserve base for the two largest cryptocurrencies, Bitcoin (124.2%) and Ether (122.02%).
Conclusion
Gate's position in the CEX landscape is best understood as two distinct but complementary pictures. On the trading side, Gate is a genuine top-tier player. That's a derivatives-heavy franchise punching above its overall size, and one of the strongest, most defensible data points in Gate's favor.
On the reserves side, the picture is solid. Gate’s reserves are genuinely diversified, spread across many distinct tracked assets with no single holding dominating the balance sheet. Bitcoin, the largest position, is only 23.1% of the total. On top of that, with a self-reported reserve coverage ratio that has stayed persistently above the 100% safety benchmark across its last three published reports, and with every major asset category—BTC, ETH, GT, stablecoins—individually over-collateralized, they have sound reserves. 
Thus, Gate is a top exchange backed by a substantial, diversified reserve base that extends into comfortable over-collateralization across its full ~500-asset reserve program. They are substantial and well-run, fully covered with room to spare.

Gate's derivatives-heavy trading book sits on top of a reserve base that's both verifiably substantial onchain and, by the exchange's own broader accounting, comfortably over-collateralized. Here's how the two pictures fit together.

Inside Gate’s Reserve Position: Diversified & Comfortably Over-Collateralized

Crecimiento

Simón Puebla

Santiago Cristobal

Argentina had roughly 8.6 million crypto users as of 2024, representing around 20% of the population. Of those, 12.5% are active on a monthly basis, significantly higher than any other country in Latin America. Volumes speak for themselves: more than $91 billion in value moved through Argentina over the twelve months to mid-2024, while the broader LATAM region generated $730 billion over 2025.

Argentina had roughly 8.6 million crypto users as of 2024, representing around 20% of the population. Of those, 12.5% are active on a monthly basis, significantly higher than any other country in Latam.

Argentina had roughly 8.6 million crypto users as of 2024, representing around 20% of the population. Of those, 12.5% are active on a monthly basis, significantly higher than any other country in Latin America. Volumes speak for themselves: more than $91 billion in value moved.

Report: Argentines’ financial reality and the impossible dollar access

Hyperliquid

In October 2025, Hyperliquid activated HIP-3, opening the platform to permissionless RWA Markets. Our first report found these markets pulled in real new users, 169,514 wallets, 31.7% of all new wallets between January and June 2026, but not significant revenue: RWA-first wallets generated just 8.3% of new-user fees, since they stayed confined to RWA Markets while Other-first wallets showed strong volume performance on both markets.
That left an open question: do these newly onboarded users actually stay on the platform and trade frequently? Asked simply, is retention good on both markets and for both cohorts? This follow-up uses two new datasets that track, at the wallet level, the number of active days per market and per cohort, and the time it takes wallets to cross over. The goal is to assess how each cohort behaves in its own market versus the other.
The key insight: engagement stays low for both cohorts, with RWA-first less active than Other-first, and the two cohorts remaining largely segregated by entry market, with one exception, a newly derived cohort of frequent, market-agnostic traders.
Quick definitions: RWA Markets cover the HIP-3 builder-deployed markets (tokenized equities, indices, commodities). Other Markets are everything else, mainly crypto perpetuals. RWA-first and Other-first classify a wallet by the market type of its first trade on Hyperliquid, same methodology as the first report.
Engagement plateaus, and RWA-first lags Other-first
If RWA-first and Other-first wallets kept trading at the pace they arrived at, weekly active wallets should be climbing alongside the cumulative base. That's not what the data shows. Cumulative wallets grew 82.6% for RWA-first (92,830 to 169,514) and 93.3% for Other-first (188,732 to 364,781) between April and June, while weekly active wallets barely moved: RWA-first stayed between 7,000 and 12,000, Other-first between 20,000 and 33,000. 
Looking at the two cohorts, if we measure daily active wallets over cumulative wallets on their own market, RWA-first averages 5.0% through the April-June window against 8.1% for Other-first. RWA traders are roughly 38% less active than Other-first traders on their respective home markets, something we hinted at in the first report.
This is confirmed by the trading frequency: nearly half of RWA-first wallets (47.4%) traded on the RWA Markets exactly once and never came back, and 68.0% traded three days or fewer over the full six-month window. Other-first wallets show more staying power: around one third (34.7%) traded once, 59.0% traded three days or fewer. And looking at high-frequency traders leads to a similar conclusion: of the roughly 13,000 wallets that traded nearly twice a week (51+ active days on their market), 10,978 were Other-first wallets (3.0% of the cohort), and 2,023 were RWA-first wallets (1.2% of the cohort).
RWA-first wallets trade less often, stick around for fewer days, and make up a smaller share of the platform's most active users. 
Crossing is rare but is immediate when it happens
If Other-first is the more active cohort, it should also be the more adventurous one, crossing into RWA Markets more often than RWA-first crosses into Other Markets. That's not what the data shows. 80.9% of RWA-first wallets and 82.0% of Other-first wallets have never touched the Other Markets at all, an almost identical split. 
When we take a look at the wallets that do cross, we can note that the move happens fast: 31.4% of RWA-first and 32.6% of Other-first crossers do it the same day as their first trade, and 59.3% (RWA-first) and 53.5% (Other-first) cross within three days. This means that crossing is not a barrier and confirms what is said above: the crossers are global traders that are more asset-agnostic.
The overall takeaway is that most traders are asset specific: TradFi traders trade TradFi, crypto traders trade crypto. The ones who cross show a different pattern: they are more asset agnostic, and the fact they cross within days rather than weeks suggests they don't need specific events or incentives to get on the Other Markets.
Other first crossers carry outsized weight on the opposite market 
With a near-identical crossover rate, we might assume both cohorts weigh the same on the market they cross into but that assumption ignores absolute size. Other-first's 65,834 crossers alone equal 38.8% of RWA-first's entire cohort (169,514), while RWA-first's 32,319 crossers amount to just 8.9% of Other-first's cohort (364,781). In other words, the Other-first wallets crossing into RWA Markets represent a meaningful chunk of that market's user base. But are these wallets active enough to matter?
Looking at how crossers distribute across activity levels confirms they are. Other-first makes up between 30% and 40% of RWA Markets's wallets in every active-day bucket (except the first). Conversely, RWA-first's presence on Other Markets is low, staying between 5% and 11% across every bucket. The tie to the first report is direct: we found that Other-first generated roughly 40% of trading volume on RWA Markets, and that volume comes from broad participation across the cohort, not a handful of big wallets.
The busiest traders go fully agnostic
Staying focused on the crossers, frequency of trading has something interesting to tell us: as trading activity rises, so does the agnostic behavior of participants, and this holds in both cohorts.
As displayed clearly in the charts, the more a cohort trades on its market, the more it trades on the other one. Among RWA-first crossers who trade 51+ days on RWA market, 58.2% also trade 20+ days on Other Markets, and among Other-first crossers who trade 51+ days on Other Markets, 53.8% also trade 11+ days on RWA Markets. That pattern reverses almost completely at low frequency.
The takeaway is that the most frequent traders are asset agnostic and trade across both markets, leading to the definition of a third, smaller cohort in our dataset: the Frequent Traders who treat RWA and crypto as just assets, without real preferences.

Conclusion
Retention stays low across the board, and RWA-first lags Other-first on every measure, number of daily active wallets, engagement rate and trading frequency. Most wallets in both cohorts, over 80%, never touch the market they didn't enter through, and the ones that do decide within days, without the need of incentives or market events.
But those crossers matter more than their share suggests. Because Other-first is more than twice the size of RWA-first, its 65,834 crossers alone equal 38.8% of RWA-first's entire cohort, a broad enough footprint to explain the volume weight the first report found on RWA Markets. And as we move into the higher end of the trading-activity spectrum, the concept of RWA-first and Other-first makes less and less sense, leading to the finding of a third, smaller cohort composed of high-frequency traders who treat RWA and crypto as one and the same.
Three cohorts, not two, is the headline here: RWA-first and Other-first traders who stay largely market-specific, and a frequent-trading core that's genuinely asset agnostic. The next step is looking at what that core actually trades, position size, duration, and leverage, to see whether it's also the most valuable one.
Methodology: This piece uses the same RWA-First and Other-First cohorts and the same January 1 to June 30, 2026 window as the first report, built from two new datasets: daily active wallet counts by cohort and market, and a wallet-level table tracking active days on each market and the date of first crossover, where applicable. Data provided by Llama Eden.

If RWA-first and Other-first wallets kept trading at the pace they arrived at, weekly active wallets should be climbing alongside the cumulative base. That's not what the data shows.

Hyperliquid and the RWA retention question: what the data shows

Wei Zhou

Coins.ph

Wei Zhou is the CEO of Coins.ph, the Philippines' largest regulated cryptocurrency exchange with 18M customers, and is spearheading Coins’ global expansion into Australia, Europe, Latin America, and Africa. Before Coins, he was CFO at Binance, guiding it to become the world’s largest exchange, and served as CFO at Zhaopin.com and Charm Communications. He also led the landmark acquisition of Grindr, where he served as Vice Chairman.
Coins.ph started as a crypto exchange, but your latest initiatives increasingly position the company as a key infrastructure player in the global payments stack. How do you describe Coins.ph’s identity and core mission today?
 We still run an exchange, but I think of Coins.ph more broadly as financial infrastructure. The exchange gives us the liquidity layer, but we are also a regulated wallet, a payments platform, and a bridge between local currencies and global digital assets.
My co-founder and I describe the goal as getting financial transactions as close as possible to T+0 plus zero. In the payments world, T+0 refers to settlement time, and the second zero is the cost. You want both as close to zero as possible.
In the Philippines, we are already quite close in certain corridors. You can move from USDC into pesos at very tight pricing, receive the pesos directly in your wallet and use them for everyday payments. The broader mission is to build that infrastructure across more markets and corridors.
Why are stablecoins rapidly becoming an essential tool for cross-border money movement right now? From a macroeconomic and technological standpoint, what has changed over the past few years to drive this institutional and retail adoption?
 Stablecoins have reached a point where they are no longer used only by crypto traders. They are becoming real financial infrastructure for consumers, businesses, fintechs and financial institutions.
If you look at USDT specifically, in emerging markets it has given anyone outside the US access to a US dollar checking account, essentially. I don’t get any interest on it, but I can use it. All I need is a wallet, and I can convert my local currency through whatever exchange I use locally and get USDT pretty easily. That gives me access to basically everything dollar-denominated.
In Asia, USDT is still the dominant stablecoin, whether it is Hong Kong, Singapore, Thailand, Indonesia, or other regions. From the US perspective, it is mainly USDC.
The macro side is also important. In markets where local currencies are depreciating, you typically see people willing to pay a premium to buy dollars, USDT or USDC. But for stablecoin usage to really take off, you need more than demand. You need regulated infrastructure, bidirectional flow, and good pricing.
On our exchange, our price is comparable every day, every hour, to the bank mid-market rate for FX, and you can do it at size. That is a key piece of infrastructure. Once that exists, you don’t just have crypto traders using it. You have real companies, financial institutions, banks, payment companies and newer fintech startups.
The Philippines is globally recognized as one of the largest and most dynamic remittance markets. What foundational lessons from building and scaling payment solutions in the Philippines can be applied to solving payment friction in other diverse regions, such as Europe and Latin America?
 The common issue is that you are going to need a license to operate. You can’t run something like LocalBitcoins with stablecoins. No regulated financial institution is going to run on a P2P model. If you are a business, you want to do things you can scale.
The model we built in the Philippines is quite unique compared to other regions because you need flow. We have a license in Thailand and have been operating there, but Thailand isn’t a major cross-border payment or remittance country. The flow just isn’t there, even if you have the infrastructure.
You need to sit in major flows in both directions. If people only want to move out of the local currency because it is depreciating, that is one-directional demand. You need bidirectional flow for it to really work.
You also need a mature banking system. If your banks don’t have APIs that you can connect to for automatic deposits, you are still stuck at the fiat leg. A lot of times, people still have to call a desk or physically go inside a branch.
The lesson is that blockchain solves part of the problem, but you still need licenses, banking access, liquidity, and real payment flows.
Many local stablecoins are being introduced to the market. Even Coins.ph mentioned plans to introduce its own stablecoin. What unique opportunities does a regulated, fiat-pegged peso stablecoin unlock for domestic everyday commerce, and how does it bridge the gap for global businesses interacting with the Philippine market?
We run an exchange, so people naturally come in to exchange pesos for peso stablecoins. As long as we provide that channel, and you can do it anytime and at any size, you can essentially get out of using banks even for pesos.
It is about providing services without friction, and the friction most people face is at the banking level, not the crypto level. You have that ease of use with US dollar stablecoins already, but not really with pesos right now. Once we add a peso stablecoin, a lot of that friction goes away.
What it unlocks is that anyone holding USDC can buy peso stablecoins anytime, on a DEX or on our platform. If you want to invest in peso-denominated financial products, you then have that option.
Once we build that liquidity layer between the peso and the dollar, it will unlock opportunities for peso holders to access the dollar world and for dollar holders to access the peso world.
That is something we have been educating regulators about. I tell them, ‘Don’t think of it as stablecoins. Think of it as dollars on a blockchain. It moves at the speed of light. Your currency still moves at the speed of water. You need your currency to move at the same speed, or everyone holding it is going to move into the faster one.’
A lot of the more progressive regulators are starting to understand that. They have given us a sandbox approach for the peso-backed stablecoin, and we are hoping to get out of the sandbox by the end of the year.
Regions like Latin America are seeing explosive growth in digital transactions. What makes these fast-evolving payment landscapes so attractive for stablecoin integration, and what role do you see them playing in the future of global money movement?
 One market we are really excited about is Latin America, mainly because you have major flows with the US in both directions.
Whether it is buyers of agricultural products, people living in the US sending money home or Chinese sellers in Latin America wanting to take money out, you see a flow that is similar to what we see in the Philippines.
As each country rolls out licensing regimes for crypto and payments, I think there is a pretty big opportunity to copy and paste what we have built in the Philippines into Latin America, namely Brazil, Argentina and Colombia. Those are the major markets down there.
The key is that the flows have to exist in both directions. Once you combine those flows with licensing, local banking access and stablecoin liquidity, you can significantly reduce the time and cost involved in moving money internationally.
You’ve spoken about building a repeatable framework for entering new markets and connecting different payment networks. What are the core pillars of this infrastructure playbook, especially regarding compliance and interoperability, and where do you see this model being applied next?
You need a license, and you need to sit in major flows both ways. You also need a corresponding payment partner or payment license holder.
In the payment world, you can do third-party collection and third-party disbursement, but as a licensed crypto exchange, you can’t necessarily do that. Under the travel rule, crypto companies can generally only send to the same name. If you try to go the other way, people will ask what the purpose of the money is.
You can’t really act as a third-party intermediary processing payment transactions without a payment license. So we would want a partner in the payments world that uses our services to do that.
From an expansion standpoint, that is the infrastructure we have realized is necessary. Once it is in place, stablecoins basically become synonymous with the US dollar. Products built on top of stablecoins, whether that is stocks, lending or cards, make much more sense because you can swap your money in and out without thinking about the cost.
But you have to build it corridor by corridor. Maybe it is the Europe-to-Philippines corridor or the Hong Kong-to-Philippines corridor. Once we have built out the Philippines corridor with many of the sending countries, we can add Brazil, Colombia or Nigeria on the receiving end, either ourselves or with local partners.
At the end of the day, it is about building a network. One of the best ways to improve pricing and liquidity is to drive as much volume through that network as possible. Initially, it is not retail. It is mostly businesses, because they move real, significant volume. A lot of the early adopters are cross-border payment and cross-border e-commerce players.
Looking ahead, as the lines between traditional banking infrastructure and digital assets continue to blur, what does success look like for Coins.ph over the next five years?
 My co-founder and I have a goal of getting financial transactions down to T+0 plus zero, as close as possible. In the payment world, T+0 refers to how long things take to settle. The second zero is the cost, and you want that as close to zero as possible.
In certain markets and countries, that is technically impossible because there are intentional barriers set up to keep money from moving as fast as you would want. But in certain corridors, we are pretty close.
Between the US and the Philippines today, if you are a normal person with a Coinbase wallet and a Coins.ph wallet, you can get close to T+0 plus zero. You deposit money, buy USDC, send it to Coins, sell the USDC into pesos at maybe five basis points, and then you have pesos in your wallet.
Because we are an e-wallet in the Philippines, you can scan and pay as you would with normal money. So that is T+0 plus five or ten basis points.
That is what we hope to achieve more broadly, but we have realized that you have to build it corridor by corridor. Once we build out the Philippines corridor with more sending countries, we can add markets such as Brazil, Colombia or Nigeria on the receiving end, either ourselves or with local partners.
Ultimately, success is building a network where money can move across borders almost instantly and at close to zero cost.

Wei Zhou is the CEO of Coins.ph, the Philippines' largest regulated cryptocurrency exchange with 18M customers, and is spearheading Coins’ global expansion into

Wei Zhou is the CEO of Coins.ph, the Philippines' largest regulated cryptocurrency exchange with 18M customers, and is spearheading Coins’ global expansion into Australia, Europe, Latin America, and Africa.

T+0 Plus Zero: Coins.ph CEO Wei Zhou on the Speed of Money

For years, financial executives touted the revolutionary potential of tokenized equities. While many remained cautious about cryptocurrencies, they recognized that blockchain infrastructure could transform how stocks and bonds are traded. For much of that time, widespread adoption always seemed just over the horizon. Suddenly, it’s at our doorstep. 
The value of all tokenized equities has grown 140% in 2026, to more than $1.9 billion, according to DefiLlama data. Bitget, Binance, Kraken, Ondo, and Bybit have launched tokenized equity products. The momentum extends beyond crypto-native platforms. Robinhood recently launched its own blockchain for tokenized real-world assets, highlighting growing interest from traditional finance firms.
In our latest report, we take a look at Bitget’s Stocks 2.0, a product suite that offers users two ways of gaining exposure to both private and publicly listed companies, including companies such as SpaceX and Nvidia. 
In the report, our analysis found that Bitget’s tokenized equities products feature some of the tightest spreads and deepest liquidity in the market, which aligns with the strong trading activity observed since launch.
The traditional option 
Bitget Stocks 2.0 introduced two products: Bitget Stock+ and Reality rTokens. 
The former offers conventional ownership in company equity. Users remain in the Bitget app, but execution, clearing, settlement, and custody of the securities are handled by traditional financial institutions. 
Securities purchased through Stock+ are held in their own account and are not issued as onchain tokens. As such, they cannot be withdrawn to a crypto wallet or used in the broader DeFi ecosystem. They do, however, confer direct ownership of company shares, as well as the right to any dividends or participation in shareholder votes. 
Stock+ serves users looking for a traditional brokerage experience, while Reality rTokens target crypto-native investors seeking onchain exposure. 
rTokens’ retail rush 
Bitget’s Reality rTokens are, as their name implies, tokenized equities. Here’s how they work. 
Bitget’s Reality Protocol handles the minting and burning of rTokens. It also maintains the relationship between the supply of rTokens and the securities they represent, which are held in segregated reserve accounts. Finally, the protocol distributes dividends and adjusts user balances based on actions taken at the corporate level, such as stock splits or reverse splits. 
To be sure, rTokens have their drawbacks. Holders are not registered owners of the underlying securities. That means holders have no voting rights, among other things. They are, however, able to withdraw their tokens to crypto wallets and use them across DeFi applications such as lending markets, liquidity pools, and more. It also means they can be minted or redeemed 24 hours a day, five days a week. Secondary trading on Bitget is available 24/7. 
Bitget’s rTokens have generated significant momentum since their launch. Between June 2 and July 19, rToken spot markets saw $1.16 billion in cumulative trading volume. The increase wasn’t limited to the initial launch period. The average daily trading volume in June was $17.9 million. Through the first 19 days of July, it was $33.8 million. 
The overwhelming majority of that volume has come from retail traders. Individual users accounted for $1.1 billion, or 95%, of cumulative spot volume. Their average trade size was just $422. These figures reflect the promise of tokenization, which allows for fractionalization of stocks that might otherwise be too expensive for a small-dollar investor to purchase. 
Thus far, users have shown interest in just a handful of stocks, with 10 markets accounting for 83% of spot volume. Semiconductor tokens accounted for just under half of spot volume, reflecting strong demand for AI-related equity exposure. Other tech companies, such as SpaceX and Tesla, accounted for another 34%. 
An apples-to-apples comparison of five markets across five separate issuers of tokenized equities, Bitget, Bybit, Binance, Kraken, and Ondo, shows that Bitget recorded some of the tightest bid-ask spreads and deepest order-book liquidity among the platforms analyzed.
To read the rest of our analysis, including how Bitget’s real-world asset perpetuals fare against the competition, read our report here.

Bitget’s tokenized equities products feature some of the tightest spreads and deepest liquidity in the market, which aligns with the strong trading activity observed since launch.

Bitget's Tokenized Stocks Gain Traction as Retail Demand Accelerates

DefiLlama

NEW YORK, August 5, 2026 — Crypto traders and analysts can now evaluate market-maker performance the same way they evaluate volume, TVL, and fees. Forgd, a Web3 investment bank and advisory platform, today announced that its Market Maker Leaderboard is live on DefiLlama, bringing standardized market-maker performance data to one of crypto's most widely used analytics platforms.
The leaderboard ranks leading crypto market makers, the firms whose constant buying and selling keeps tokens tradable, on standardized measures of pricing, market depth, reliability, and execution quality. How well these firms do their job affects every trade, yet until now there's been no way to check their track record.
"Crypto traders already look at volume, liquidity, TVL, unlocks, fees, and other market indicators," said Shane Molidor, Founder and CEO of Forgd. "Market-maker quality is another indicator that should be closely analyzed. You can't accurately judge a token's liquidity without understanding who's providing it, and how well they perform.”
What traders and analysts can do with the data
Because the leaderboard lives inside DefiLlama and other widely used analytics platforms, users can fold market-maker quality directly into their existing workflows, including:
Token research — check whether a token's liquidity is backed by firms with a strong execution record before taking a position
Liquidity and risk analysis — distinguish durable liquidity from headline volume, and flag markets likely to break down under stress
Venue comparison — see how the same firms perform across different exchanges and market environments
“Volume alone does not tell you whether a market has healthy, reliable liquidity,” said Ryan Celaj, Head of Research at DefiLlama. “Bringing Forgd’s market-maker performance data into DefiLlama gives traders another practical signal for evaluating execution quality and market resilience, directly within the research workflow they already use.”
Behind the data
The leaderboard is powered by data from more than 500 token projects and 35 market-making firms that use Forgd's tools to monitor liquidity across live engagements — a standardized record of how firms have performed across different tokens, venues, and market conditions. Forgd first introduced the leaderboard in May 2026 as a tool for token projects evaluating market makers; the DefiLlama integration extends that same data to the broader market.
The integration is part of Forgd's ongoing effort to make crypto market structure more transparent.
About Forgd
Forgd is a Web3 investment bank and advisory platform equipping blockchain projects with tools for tokenomics design, market-maker engagement, liquidity monitoring, and exchange listing support. With options for bespoke advisory services and free self-service products, Forgd helps token projects design, launch, and manage healthier markets. New projects can join for free at Forgd.com.
About DefiLlama
DefiLlama is the largest open-source data platform for decentralized finance (DeFi) and the broader onchain economy, tracking key metrics across thousands of protocols and chains. Its coverage spans stablecoins, real-world assets, blockchains, yields, and traditional finance, built on an open-source, publicly auditable methodology that gives researchers, traders, institutions, and journalists a neutral source of truth for the crypto industry. Beyond its free core dashboards, DefiLlama offers a growing suite of products, including LlamaAI, its AI-powered data assistant, the DefiLlama Mobile App, and DefiLlama Pro, a subscription tier with advanced dashboards and premium tools. Explore the data for free at defillama.com.

Forgd, a Web3 investment bank and advisory platform, today announced that its Market Maker Leaderboard is live on DefiLlama, bringing standardized market-maker performance data to one of crypto's most widely used analytics platforms.

Traders and analysts can now see how the firms behind a token's liquidity actually perform, from inside the platform they already use every day

Forgd's Market Maker Leaderboard Is Now Live on DefiLlama, Offering Traders a New Source of Alpha

Bitget

With the launch of Bitget Stocks 2.0, Bitget doubled down on its investment in this growing market. To understand how it compares, we look at how Bitget’s tokenized equities function, how they differ from competing products, and what their spreads and order-book depth reveal about its standing in the tokenized-equity race.

With the launch of Bitget Stocks 2.0, Bitget introduced two products that provide users with access to equities: Bitget Stock+ and Reality rTokens, each targeting a different audience.

Bitget Stocks: Bridging Traditional Equities and Onchain Markets

Tokenization makes stocks more accessible, more liquid, and in principle tradable 24/7, from anywhere. Traditional exchanges still close on schedule, but the venues distributing tokenized versions of the same stocks don't have to. The question: who actually shows up when Wall Street doesn't?
As of August 2, roughly six weeks after launch, Binance's bStocks have already crossed $500M in AUM. On top, bStocks also recorded $2B in trading volume over a single weekend, a signal worth stress-testing against real usage data. In this article, we compare Binance's bStocks, Kraken's xStocks, and Ondo's Global Markets as they represent three distinct tokenized-stock models while offering consistent hourly volume data. Using six weeks of DeFiLlama data, from June 14 to August 2 , 2026, we examine how trading volume shifts between weekday market hours, weekday off-hours, and weekends, whether one platform takes the lead once traditional markets close, and what these patterns reveal about its users.
Weekdays - A Market That Never Fully Closes
The weekday hourly profile confirms the basic premise: trading doesn't stop when the traditional market does, and tokenized venues hold their promise of round-the-clock access. However, the profile isn't uniform. Activity clusters tightly around the moment TradFi opens, with volume across all three venues surging between 13:00 and 15:00 UTC and slowly decreasing until the market closes.
Visually, it's tempting to conclude that most of the volume concentrates during the opening hours. However, that picture needs nuance. If we count 13:00-20:00 UTC as TradFi's opening hours, 49% of weekday volume falls in those 7 hours. But the remaining 51% still happens off hours. That's significant, and it says something specific: traders are here round the clock, and TradFi's trading hours function as a constraint. In other words, traders have to adapt around the opening hours.
As 24/7 access becomes standard across more venues and more tickers, that concentration should fade. The more normal it becomes to trade a tokenized stock outside market hours, the less trading should cluster around the one moment TradFi happens to be open.
By venue, Binance's lead holds through the entire day, ranging from 64% to 86% of combined volume in any given hour. But the pattern underneath is clear: Binance's dominance actually increases during off-market hours and is strongest over the weekend. 
Off-market hours, Binance averages 82% of volume, against 15.2% for Ondo and 2.8% for Kraken. During market hours, Binance’s share moderates to 69%, while Ondo rises to 27% and Kraken to 4.0%.
That gap points to Ondo and Kraken being relatively more active while the underlying market is live. Two main reasons can explain this. First, user typology: Binance leans more retail while Ondo is more institutional. Second, the structural layer or how each product is actually built and what it relies on to function.
The products may look similar on paper, but their liquidity models are fundamentally different. bStocks are converted 1:1 from Binance Stocks positions and trade directly on Binance’s order book. That liquidity also extends to BNB Chain, creating a bridge between CeFi and DeFi positions, particularly interesting for market makers. xStocks, by contrast, rely on pre-funded liquidity pools. These pools can become thinner once the underlying market closes. Ondo follows another model: tokens are minted and redeemed against live trades on NASDAQ and the NYSE. As a result, its liquidity is closely tied to US market hours. These structural differences help explain why some venues perform better outside traditional trading hours and especially on weekends.
Weekends - How Token Design Sets The Dominance
Weekend trading is still a small slice of the total, about 4.3% of combined volume across the sample. That's likely a floor rather than a ceiling: as more venues and more tickers get built for continuous trading, weekend share could grow from here.
Interestingly, the pattern within the weekend is relatively flat. Volume stays mostly in the $380K-$2.04M range across the sample, with a notable dip in activity on Saturday evening and again on Sunday evening, both in UTC. It only picks back up as Sunday evening turns into the Asian start of the week, ahead of Monday. 
Across the roughly 350 weekend hours in the sample, Binance holds somewhere between 85% and 97% of combined volume in nearly every one of them, averaging 92.8%. Kraken averages just 3.24%, spiking as high as 50.4% in a handful of isolated hours. Ondo trails further still, averaging 3.93% and dropping to zero in some hours entirely.
The main factor explaining this is the same structural element mentioned earlier, and in the end it comes down to where the liquidity is actually sourced from. Ondo's tokens are minted and burned against an actual Nasdaq trade happening in real time, so once Nasdaq shuts for the weekend, there's no live transaction to mint or burn against, though that's starting to change: Ondo enabled 24/7 minting and redemption for six of its assets in late June. Kraken's xStocks trade against pre-funded liquidity pools rather than a live link to the exchange, and only 10 of its 141 tickers are enabled for weekend trading. As for Binance, once a bStock is converted, it can be traded on Binance’s own centralized order book, the same infrastructure that already runs every other asset on the exchange at any hour, and on-chain. In other words, it doesn't rely on a pool that can run dry, or a mechanism that needs traditional markets to be awake to function.
On the surface, these figures could be read as just one venue accumulating more liquidity. But they point to something more important: price discovery increasingly happening on Binance rather than elsewhere. Indeed, the venue holding the deepest, most continuous order book during the hours nobody else is trading isn't just capturing volume, it's becoming the place where the market actually finds its price.
Conclusion
Two things stand out from the full-period view. Raw volume splits almost evenly between opening hours and weekday off-hours, 46.2% against 48.0%, with weekend trading a small 5.8% slice on top. But those weekday periods aren't the same size, opening hours are a fraction of the week's hours compared to nights, and normalized for that, they account for 66.0% of volume-per-hour against 28.2% for weekday off-hours. In other words, TradFi's operating hours still set the tempo today, though that grip should loosen as adoption spreads across time zones and 24/7 access becomes the norm rather than the exception.
By venue, Binance leads across every segment, and its lead grows in off-hours: 69.0% during market hours, 82.0% overnight, 94.4% on weekends. Ondo moves in the opposite direction, from 27.0% down to 15.2% and then to just 3.7%, a trajectory that tracks almost exactly with how much live Nasdaq activity its mechanism has to work with. Kraken barely moves at all, hovering between 1.9% and 4.0% regardless of the segment, which suggests its volume is small enough in absolute terms that session timing isn't really the constraint, distribution and catalog coverage are.
As of now, Binance bStocks end up looking like the product that best matches two things at once: how crypto-native users already trade, i.e., continuously, and what tokenization was supposed to deliver in the first place, i.e., deep liquidity and access anytime, anywhere.
Methodology: hourly and daily USD trading volume for Binance bStocks, Kraken xStocks, and Ondo Global Markets, June 14 to August 2, 2026 (UTC). bStocks volume is measured on Binance (56 assets), Kraken’s xStocks volume on Kraken (141 assets), and Ondo volume on the Ondo Global Markets platform (400+ assets). Market hours are defined as NYSE regular sessions, 13:30-20:00 UTC (considered 13:00-20:00 UTC in the analysis), weekdays. Data extracted by the DeFiLlama team.
Raw data: https://gist.github.com/FelixBruguera/7d10c40fc0cd8a885e7c0d82dd2a09a2

Tokenization makes stocks more accessible, more liquid, and in principle tradable 24/7, from anywhere. Traditional exchanges still close on schedule, but the venues distributing tokenized versions of the same stocks don't have to. The question: who actually shows up when Wall Street doesn't?

bStocks After Hours: Who is Capturing Liquidity When Markets Close?

Hylo Leveraged SOL

This research provides a clear, balanced breakdown of Hylo’s xSOL leveraged token for the general public. It examines the genuine advantages (no liquidation risk and no continuous funding fees) while transparently showing the actual costs and trade-offs, including dynamic mint/redeem fees, volatility decay, variable leverage, and other risks. The goal is to help readers understand whether xSOL truly delivers “3x SOL without hidden costs” or if there are important realities every investor should know before using it.
Hylo markets its xSOL token on two guarantees: no liquidation and no funding rate. And so far this holds true, as nobody holding xSOL has ever been forced out of a position, and nobody pays an ongoing fee to hold leveraged SOL exposure the way they would on a perpetual exchange.
But no liquidation doesn't mean no loss. Between October 2025 and June 2026, under Hylo's first version, xSOL lost nearly all of its value, even though SOL itself, the asset it's meant to track with leverage, fell by a much smaller amount. Meanwhile, hyUSD, the stablecoin this whole system exists to protect, never lost its dollar peg through any of it, another outcome promised by the protocol.
That gap is the actual subject of this piece. If nobody was liquidated despite the leverage, where did the extra loss go, and who ended up paying for it? The short answer is dilution, and we will dig into why by revisiting the stress event that happened in the period mentioned above. Then we'll assess whether V2 actually changes things and limits or reduces that risk. But before any of that, let's briefly revisit how Hylo works under the hood.
Under the hood: the promise, and what it takes to keep it
No leverage, no funding rate…
Hylo's basic principle is as follows: take one pool of staked SOL, and split it into two claims on that pool, ranked. One claim, hyUSD, sits first in line and is always worth exactly $1, as long as there's enough value in the pool to cover it. The other claim, xSOL, gets whatever is left over, no ceiling, no floor, all of the pool's ups and downs land here. That's the entire mechanism.
To make things more tangible, here is a concrete example: say the pool holds $150 of staked SOL. $100 of that backs hyUSD, fixed at $1 a token, while the remaining $50 belongs to xSOL. If SOL rises 33%, the pool grows to $200. hyUSD still only needs its $100, so the remaining $100 becomes xSOL's new total value. That means on a 33% SOL move, xSOL's value doubled, a 100% price appreciation, three times the size of SOL's own move. On the opposite side, if SOL falls and the pool drops to $100, hyUSD still needs its $100, and xSOL is left with nothing. The ratio itself, three times the swing either way, is where "3x leverage" comes from.
Because leverage comes from the pool's composition, not borrowed money, there's nothing to force-close. xSOL holders don't owe anyone anything, they just hold a claim that can shrink toward zero in a bad enough crash, or grow substantially if the market appreciates.
The ‘no funding rate’ part comes from the underlying collateral itself. Staked SOL earns real staking rewards just by existing, roughly 7 to 8% a year. That yield flows through the system and effectively pays for the arrangement, so nobody needs to charge xSOL holders an ongoing fee.
…but proper risk management is still required
None of this holds together on its own, as the pool needs active management to stay near that target, and Hylo handles that in two layers.
The first layer is pricing. To mint hyUSD, users deposit collateral (SOL, BTC, or USDC) into the pool of their choice and receive hyUSD in return. Redemption reverses this process: hyUSD is burned, allowing users to withdraw collateral from a selected pool. Rather than applying fixed fees, Hylo adjusts pricing dynamically according to each pool’s condition. Minting is cheaper in healthier pools, while redemption is cheaper from pools experiencing greater stress. This fee structure creates economic incentives that direct user activity toward restoring balance across the system.
If the Collateral Ratio (CR) for a pool, defined as the total value of the pool divided by what hyUSD needs to stay whole, drifts outside the healthy range, roughly 136% (or 3.8x leverage) to 167% (or 2.5x leverage), a second layer kicks in. The protocol trades directly against its USDC reserve, buying collateral back when CR runs high to re-lever the pool, selling collateral into USDC when CR runs low to de-lever it. Pricing on these trades starts in the protocol's favor and shifts toward the trader as CR nears the edges of that range, which is what pulls arbitrageurs in to help push things back to center. 
This is where xSOL holders incur a real cost: selling collateral for USDC during periods of stress permanently reduces the SOL exposure backing each token and a full recovery in SOL’s price may not restore the position to its previous state.
Last but not least, in the true worst case, if a pool is still short on backing after all of that, the Earn Pool's own hyUSD gets burned to make up the difference and restore full backing. That's a direct cost to Earn Pool depositors, not to xSOL holders, and falls outside the scope of this report.
Put to the test: what happened to Hylo's first version
An important note before diving into the case study: everything below describes Hylo's first version, not the current one. V1 supported a single collateral asset, SOL, with no isolation between markets, no USDC pool, and no Earn Pool. Instead of the USDC-based rebalancing described above, the protocol relied on the Stability Pool and defended the peg through conversion of hyUSD to xSOL rather than by selling collateral. That difference matters as this mechanism created the dilution xSOL holders faced, which is the indirect cost of having no funding fee and no liquidation.
The peg held, exactly as promised…
Hylo's first version ran for about eight months on a single SOL pool before it hit a real crash. Between October 2025 and June 2026, SOL fell around 69% from its highs. Through all of that, hyUSD never lost its dollar peg and the stablecoin side of this system got exactly what was promised, at every point.
However, xSOL holders got a very different outcome. Over that same window, xSOL lost close to all of its value, down about 99% from its October high. This isn't really a surprise given the example from the last section: a 33% SOL drop is exactly the point where xSOL's claim hits zero and the pool's collateral ratio falls to exactly 100%.
But with a 69% drop in SOL price, the pool alone can no longer fully back hyUSD without help. And as hyUSD held its peg the whole time, even while the pool's own collateral was worth far less than what was needed to back it fully, the natural question that comes is “who paid for it?”.  And the answer is xSOL holders through a dilution mechanism that only experienced traders might have identified.
…but someone still had to pay for that
As Hylo’s collateral ratio declined, V1’s defense mechanism was activated to protect the hyUSD peg. This mechanism differed structurally from the USDC-based rebalancing system used in V2 and expanded xSOL’s supply through two distinct channels. The first was minting at net asset value (NAV): users could deposit fresh SOL and receive newly issued xSOL priced according to the pool’s actual per-token value at the time of minting. The second was the Stability Pool conversion mechanism, which automatically burned staked hyUSD and issued new xSOL in its place once the collateral ratio fell below a specified threshold. Together, these mechanisms caused xSOL’s supply to increase approximately 28x over the period, from around 8.4 million tokens to 235 million.
Roughly 126 million of the new tokens came from ordinary buying: someone deposits fresh SOL, the pool genuinely grows, and new xSOL mints against that deposit at the prevailing NAV. However, the depressed NAV meant that each dollar of new capital received significantly more xSOL than it would have before the crash. Although the issuance price was fair relative to the pool’s value at the time, it substantially expanded the token supply and for existing holders, this created dilution: any subsequent recovery in the pool’s value would be distributed across a much larger number of xSOL tokens.
The remaining 43% came from Stability Pool conversions. Unlike the standard minting described above, this process introduced no fresh capital into the system. Instead, the protocol burned hyUSD deposited in the Stability Pool and issued xSOL in exchange, reallocating collateral already held within the protocol.
In both cases, xSOL issuance served the same broader objective: absorbing pressure from hyUSD and helping preserve its peg. However, this protection came at a significant cost to existing xSOL holders. As the token supply grew, each existing xSOL represented a smaller share of the collateral pool and any recovery would therefore be spread across many more tokens.
What it would take to get back to even
With that much new supply in circulation, one question follows naturally: what would SOL actually need to do for xSOL to see $1 again?
The answer, based on the pool's state as of the July 2026 snapshot, is a SOL price somewhere between roughly $630 and $719 assuming xSOL supply stays in a similar range. That's a move of 8.3x to 9.5x from where SOL was actually trading at the time, and still way above SOL's previous all-time high.
That range exists because of the re-conversion: above a healthy 150% collateral ratio, tokens sitting in the Stability Pool get gradually converted back into hyUSD, shrinking xSOL's supply as things improve. So far, about 79 million xSOL, roughly a third of June's supply, has already been reconverted back into hyUSD and that's what pulled the frozen-state breakeven down from initially $1,024 in June, right after the event, to $719 today.
Either way, the conclusion is the same: SOL simply recovering isn't enough anymore. xSOL needs SOL to blow past its old all-time high before an early holder sees their money back.
Is V2 a real change in the risk?
V2 touched almost every part of the system, but does that mean the risk has changed for xSOL holders?
Multi-asset collateral
Splitting the system into independent pools — SOL, BTC, and more planned — means a crash in one no longer drags the others down with it. Look closely at what that actually protects, though, and it's hyUSD's aggregate backing, not xSOL's risk. Each pool's own health is still calculated locally, so a SOL crash puts exactly the same pressure on the SOL pool, and on xSOL holders, regardless of how healthy a BTC pool sits next to it. 
The USDC pool
The new USDC pool allows users to mint and redeem hyUSD directly against USDC. As a result, hyUSD growth no longer depends as heavily on demand for SOL, as it did under V1. It also gives arbitrageurs a permanent, liquid venue through which they can rebalance the system, rather than forcing them to wait for occasional trading opportunities. This should make the hyUSD peg more resilient.
For xSOL holders, however, the main risk appears when the protocol sells SOL for USDC during periods of stress to keep the collateral ratio within an acceptable range. Once conditions improve, the protocol may use that USDC to buy SOL back. But because the original sale took place after SOL had already declined, the proceeds may not be enough to repurchase the same amount of SOL if the market subsequently recovers. The protocol therefore rebuilds only part of the original SOL exposure, meaning that each xSOL may remain backed by less SOL than before the stress event.
The Earn Pool
hyUSD holders can deposit into the Earn Pool to earn yield, staking rewards, BTC borrow-rate revenue, and fee income, receiving eHYUSD as a receipt.
Under V1, this pool, then called the Stability Pool, doubled as Hylo's backstop: when collateral ran short, deposits got converted into freshly minted xSOL to defend the peg. That's where roughly 43% of V1's xSOL dilution came from.
V2's Earn Pool is now delta-neutral instead. In the true worst case, the protocol burns some of its hyUSD outright rather than converting it into xSOL, a bounded loss for depositors instead of a directional bet.
For xSOL holders, that closes off a real source of dilution: no new tokens get minted this way anymore.
hyUSD is the protocol's first-class citizen, xSOL pays the difference
No liquidation and no funding rate are both real and structurally true. However, they come with a hidden cost that not all participants might have understood. This one is dilution.
The case study showed it very well. In Hylo's first version, new xSOL were minted to maintain the collateral ratio and protect the peg for stablecoin holders. The minting was so aggressive that supply grew 28x, and with roughly 43% of that new supply coming from Stability Pool conversions, the recovery price now sits around $630 to $719.
V2 improves the situation, but it doesn’t fully eliminate the risk for xSOL holders. The Earn Pool (Stability Pool redesign) removes one major source of dilution: during periods of stress, stablecoin deposits are no longer converted into fresh xSOL. Instead, the protocol can sell collateral for USDC to support a stressed pool. This doesn’t increase the xSOL supply, but it reduces the amount of SOL backing each token with no guarantee of recovery due to price volatility. As for the second dilution channel, it remains unchanged: during stress events, users can still mint xSOL at NAV by depositing fresh capital. When NAV is depressed, the same amount of capital receives more xSOL, expanding the supply and reducing existing holders’ share of any future recovery, just as it did under V1.
In the end, the system protects stablecoin holders by transferring the cost of that stability to xSOL holders. No liquidation is the promise and dilution is how xSOL holders pay for it.

Hylo markets its xSOL token on two guarantees: no liquidation and no funding rate. And so far this holds true, as nobody holding xSOL has ever been forced out of a position, and nobody pays an ongoing fee to hold leveraged SOL exposure the way they would on a perpetual exchange.

Hylo xSOL: No Liquidation, No Funding - But What Are the Real Costs?

Disclaimer: This report is provided for informational and educational purposes only and should not be construed as investment advice. Nothing contained herein constitutes a recommendation or solicitation to purchase, sell, or hold any token or other digital asset, or to use any protocol or service.
It is time to examine how the risk-curation and vault economy is taking shape. To understand its development, it is useful to begin where vaults first emerged: DeFi lending. The data presented in this report was collected from DeFiLlama in July 2026
DeFi lending has been one of the largest categories in crypto since the first major lending boom in 2021. After a quieter period, the sector returned strongly in 2025, with more than $35 billion now sitting across lending protocols
What is less discussed is how the structure of lending has changed. The market is no longer defined only by large pooled lending protocols, where users deposit into shared pools and rely on protocol-level risk parameters.
A growing share of lending activity now happens through curated vaults. In this model, users deposit capital into a vault, and that capital is allocated across selected lending markets according to predefined risk rules set by risk curators.
Across 55 tracked risk curators, this market has grown to $7.18 billion, up from $4.75 billion the previous year. This implies year-over-year growth of roughly 51%. What is notable is that, over the same period, lending TVL declined by 36%. This shows that while lending as a whole has lost capital, the demand for curated vaults has continued to grow, allowing this construct to gain market share over other lending products.
How we ended up with Vaultification 
The vaultification of lending came from a series of innovations.
Vaults first became widely understood through yield aggregators such as Yearn. In that model, users deposited assets into a vault, and the vault automated strategy execution across DeFi.
The next step was standardisation. ERC-4626 created a common interface for tokenised vaults, making deposits, withdrawals, share accounting, and integrations easier to standardise across DeFi. 
The final innovation was risk curation. This became more explicit with Morpho Blue and MetaMorpho, which separated the lending base layer from the vault layer. Instead of only automating yield, the vault became a capital allocation layer. 
The curators could set parameters by which the vaults would operate including eligible markets, capital allocation split, risk limits that would risk adjust exposure overtime.
The model that Morpho helped pioneer was then adopted across multiple lending protocols, including Euler, Kamino, Gearbox, and other lending markets. 
Market overview: Key players, assets, revenues, and chains 
Risk curation is already a concentrated market. The category now includes 55 tracked curators, but only 10 curate more than $100 million in TVL. The top three curators, Steakhouse Financial, Sentora, and Gauntlet, account for 75.9% of TVL. The top five account for 80.9%, while the top ten account for 89.1%.
That concentration reflects how allocator trust develops in lending markets. Capital tends to follow curators with a track record, recognised risk frameworks, strong integrations, and experience curating large pools of stablecoin and collateral exposure. 
The long tail is expanding, but the core market is still controlled by a small group of professionalised operators.
Rank
Curator
TVL
Market Share
1
Steakhouse Financial
$2.03B
28.20%
2
Sentora
$1.97B
27.40%
3
Gauntlet
$1.46B
20.30%
4
K3 Capital
$210.2M
2.90%
5
KPK
$147.8M
2.10%
6
Telos Consilium
$131.4M
1.80%
7
Clearstar
$124.8M
1.70%
8
Sky Money
$124.6M
1.70%
9
Hyperithm
$111.0M
1.50%
10
RockawayX
$101.7M
1.40%
Competition at the top of the risk-curation market is tight. Steakhouse Financial leads with $2.03 billion in TVL, followed closely by Sentora at $1.97 billion and Gauntlet at $1.46 billion. The gap between Steakhouse and Sentora is only $60.7 million, separating the two by less than one percentage point in market share. Gauntlet is also a major player, giving the market a clear top-three structure where the leading curators are competing at a much larger scale than the rest of the field.
Beyond the top three, the market drops off quickly. K3 Capital is the fourth-largest curator with $210.2 million in TVL, meaning the gap between third and fourth place is more than $1.2 billion. This makes the current market structure clear: a competitive race between three large curators at the top, followed by a much smaller long tail.`
The market’s concentration likely reflects the fact that leading vault curators have demonstrated stronger risk frameworks and conducted more thorough due diligence in asset selection. This becomes especially important amid the increase in DeFi exploits. TRM Labs recorded 207 incidents in the first half of 2026, more than double the 83 incidents recorded during the same period in 2025.
Despite these incidents, Sentora has reported zero losses across the vaults it curates by combining technical and economic due diligence before adding asset or protocol to its strategies,  adheres  to predefined risk limits, and follows up with continuous monitoring which helps explain its strong position as the second-largest risk curator.
Revenues across the top 5 risk curators 
Revenue distribution looks different from TVL distribution. The top five risk curators generated approximately $13.07 million in 30-day fees and $975.6K in 30-day revenue. Sentora leads on revenue, generating $559.2K over the past 30 days, or roughly 57.3% of the top-five total. However, there are stark differences in how effectively each risk curator converts fees into revenue. 
Curator
Fees 30d
Revenue 30d
Revenue 7d
Revenue / Fees 30d
Sentora
$4.77M
$559.2K
$137.5K
11.71%
Steakhouse Financial
$4.22M
$290.9K
$60.5K
6.90%
K3 Capital
$658.4K
$64.7K
$15.5K
9.82%
Gauntlet
$3.31M
$50.5K
$10.7K
1.53%
Hyperithm
$115.8K
$10.5K
$2.8K
9.03%
The main explanation for the difference between fees and retained revenue is the performance-fee structure used by each curator. In curated lending vaults, the underlying lending markets generate yield for depositors. The curator only captures revenue when part of that yield is retained through a performance fee, management fee, or separate revenue-sharing agreement. As a result, two curators can generate similar levels of fees or TVL  but report very different revenue if their vaults apply different take rates.
Sentora generated the most revenue among the top five, retaining $559.2K from $4.77 million in fees, equivalent to an 11.71% revenue-to-fee conversion rate. Its vaults charge performance fees ranging from 0% to 15%, while its PayPal USD Main vault also applies a 0.99% management fee which contributed substantially to its revenue. Combined with its high fee generation, this stronger effective take rate allowed Sentora to generate almost twice as much revenue as Steakhouse Financial.
Steakhouse Financial generated the second-highest revenue at $290.9K from $4.22 million in fees. Its Morpho vaults charge performance fees ranging from 0% to 25%, but its effective conversion rate was 6.90%. This suggests that a significant share of its fee generation comes from vaults with lower performance fees. 
K3 Capital ranked third, generating $64.7K in revenue from $658.4K in fees. Its 9.82% conversion rate broadly aligns with the 10% performance fee applied across its vaults. This consistent fee structure allowed K3 to generate more revenue than Gauntlet despite producing significantly fewer fees.
Gauntlet generated $50.5K in revenue from $3.31 million in fees, giving it a conversion rate of 1.53%. Although its vault-level performance fees range from 0% to 20%, the low effective conversion rate means that most of its fee generation comes from vaults with low or zero performance fees.
Hyperithm generated the least revenue among the top five, retaining $10.5K from $115.8K in fees. Its vaults charge performance fees ranging from 0% to 10%, and its effective conversion rate of 9.03% suggests that a meaningful share of its fee generation comes from vaults near the upper end of this range. However, its smaller fee base limits its absolute revenue.
Overall, the data shows that performance-fee and management-fee design plays a major role in determining whether a curator can convert usage into revenue. 
Curators such as Sentora and K3 Capital retain a significant share of the fees generated by their vaults, while others, such as Gauntlet, generate substantial activity but capture only a small portion of it as revenue. This highlights the clear differences in strategy across the market: some curators prioritise direct monetisation, while others appear more focused on distribution.
The stablecoin Anchors of the Vault Economy 
The asset mix shows what curated vaults are mainly being used for: stablecoin yield. Across the top 10 risk curators, USDC is the dominant asset, with $2.81 billion in TVL. That represents 43.8% of the analysed top 10 curator asset base, making USDC the primary asset of the vault economy.
After USDC, the market becomes more fragmented. Paypal’s stablecoin, PYUSD is the second-largest asset at $921.6 million, followed by USDT at $420.5 million, RLUSD at $298.6 million, and WBNB at $292.5 million. No other asset comes close to USDC’s level of dominance.
Asset
TVL
Share
USDC
$2.81B
43.80%
PYUSD
$921.6M
14.40%
USDT
$420.5M
6.60%
RLUSD
$298.6M
4.70%
WBNB
$292.5M
4.60%
WETH
$248.5M
3.90%
KBTC
$201.3M
3.10%
EURCV
$170.9M
2.70%
USD1
$137.6M
2.10%
USDT0
$133.2M
2.10%
This matters because the main use case is not directional exposure to volatile assets. For many allocators, curated vaults are a way to access stablecoin lending with clearer risk limits, curator oversight, and more structured withdrawal assumptions.
The market is not made up exclusively of USDC vaults. PYUSD, RLUSD, USDT, EURCV, USD1, and USDT0 point to broader stablecoin adoption, with an increasingly institutional and issuer-led component. This is visible in the competition between PayPal’s PYUSD, Ripple’s RLUSD, and World Liberty Financial’s USD1, showing that newer stablecoins issued by established names such as PayPal and Ripple have a chance to capture market share.
Sentora has implemented various vaults with PYUSD and RLUSD as the main deposit assets, with its Sentora curated RLUSD Main Vault holding $234.32M in total deposits and its PayPalUSD Main Vault holding $317.5M in total deposits, supporting the growth of these stablecoins.
Meanwhile WBNB, WETH, KBTC show that curators are also managing exposure to other types of assets which are not stablecoins.
Still, the centre of gravity is clear. Curated vaults are becoming one of DeFi’s main risk-managed stablecoin allocation layers, and while USDC remains the largest stablecoin in DeFi other players such as PYUSD are starting to gain traction. 
Chain Distribution: Ethereum and Base Dominate Curated Vault TVL
Curated vault capital remains concentrated on the most liquid DeFi venues. Ethereum is the largest by a wide margin, with approximately $3.46 billion in risk curator TVL, representing 48.2% of the category. Base follows with $1.62 billion, or 22.5% of the market. Together, Ethereum and Base account for roughly 70.7% of total risk curator TVL. 
Chain
TVL
Share
Ethereum
$3.46B
48.20%
Base
$1.62B
22.50%
Solana
$548.4M
7.60%
BSC
$509.2M
7.10%
Plasma
$144.3M
2.00%
Monad
$119.3M
1.70%
Hyperliquid L1
$62.8M
0.90%
Katana
$60.4M
0.80%
Avalanche
$43.2M
0.60%
Arbitrum
$28.9M
0.40%
Other chains
$567.6M
7.90%
Ethereum’s lead is expected. It remains the deepest venue for stablecoin liquidity, blue-chip collateral, and institutional DeFi activity. Base’s position is more notable. With more than $1.6 billion in curator TVL, Base has become the largest L2 venue for curated vault capital and now holds more than three times as much risk curator TVL as Solana.
Beyond Ethereum and Base, the market becomes more fragmented. Solana and BSC form the second tier, with $548.4 million and $509.2 million in TVL, respectively. After that, no individual chain exceeds $150 million.
What is notable is that chains that have historically been large DeFi venues, such as Arbitrum and Avalanche, are losing market share to newer chains such as Katana, Hyperliquid L1, Monad, and Plasma. This suggests that, when it comes to risk curation, curators are increasingly operating across newer DeFi ecosystems rather than sticking to the old guard.
K3 Capital’s growth on Monad provides a clear example of this trend, with its vaults on Morpho and Euler reaching $60 million in TVL across different vault strategies.
The takeaway is that vault infrastructure is becoming multi-chain, but capital is not spreading evenly. Newer chains are taking advantage of this infrastructure by working with risk curators to create competitive vaults. At the same time, well-known chains like Ethereum, Base, Solana, and BSC continue to thrive. 
A concentrated market with emerging pockets of competition
The risk-curated vault market is growing quickly, but it is not yet a broadly distributed market. From almost every angle, the sector remains relatively concentrated. Steakhouse Financial, Sentora, and Gauntlet dominate curator TVL, Ethereum and Base account for the majority of chain-level capital, and USDC remains the clear anchor asset for risk-managed vault strategies.
This concentration shows that allocator trust is still clustering around a small number of familiar curators, blockchains, and widely adopted stablecoins
At the same time, the data underneath that dominance is where the market becomes more interesting. Smaller curators can still be highly profitable relative to their size when they use higher or more consistent performance fees, as seen with K3 Capital. Newer chains such as Monad are also proving that curated vault infrastructure can help emerging ecosystems compete with, and in some cases outgrow, older DeFi venues like Arbitrum and Avalanche. 
Meanwhile, the asset mix shows that USDC dominance does not mean stablecoin uniformity. PYUSD and RLUSD are already among the largest vault assets, pointing to growing institutional and issuer-led competition inside the vault economy.
This market creates an opportunity for risk curators like Sentora to capture market share through differentiated strategies. This can come from introducing vaults for assets that compete with USDC, as well as deploying across a growing set of chains. While the market remains highly concentrated, Sentora is one of the clear winners as the second-largest risk curator, and it can further differentiate itself through unique vault strategies and new chain deployments. 
Another important trend is that major financial platforms are starting to use curated DeFi infrastructure behind familiar products. Coinbase allows users to lend USDC through Morpho vaults directly in its app, while Deel lets contractors who hold DLUSD opt into a rewards programme where funds are deployed through Morpho and curated by Sentora. Robinhood has also launched an Earn product that allows users to lend USDG through a Morpho vault curated by Steakhouse Financial.
On top of that, Kraken’s DeFi Earn product is powered by Sentora’s strategies behind its vaults. Kraken has already surpassed $500 million in deposits, showing how powerful this model can be for risk curators. 
Instead of requiring users to interact with DeFi directly, these onchain earn solutions enable vault infrastructure to be embedded into familiar financial products. This may become one of the main ways the next wave of capital enters DeFi, as neobanks and other financial applications can use the expertise of risk curators to offer yield to their customers.

A growing share of lending activity now happens through curated vaults. In this model, users deposit capital into a vault, and that capital is allocated across selected lending markets according to predefined risk rules set by risk curators.

It is time to examine how the risk-curation and vault economy is taking shape. To understand its development, it is useful to begin where vaults first emerged: DeFi lending. The data presented in this report was collected from DeFiLlama in July 2026.

The Vault Economy: Scale and Composition

When Hyperliquid activated HIP-3 in October 2025, it made a structural bet. By allowing any qualified builder to deploy a permissionless perpetual market on HyperCore (staking 500,000 HYPE worth $3.5M at the time of writing) the protocol effectively handed listing decisions to the market itself. The context matters: centralized exchanges have spent the past two years racing to absorb RWA flow, listing tokenized equities, indices, and commodities as fast as regulation allows. Hyperliquid's answer was different, make the listing layer permissionless and let builders do it onchain, 24/7, without approval cycles.
The question that follows isn't whether new asset classes would appear but whether they'd pull in genuinely new users and, if so, whether those users would matter to the platform's revenue. 
To shed light on this topic, we used in-house data to answer two questions directly. First, does opening Hyperliquid to RWA Markets attract net new users, or does it mostly give existing traders something else to trade? Second, do those RWA-First users generate significant volume and fees, or does the revenue continue to come from the crypto-native base regardless?
RWA Markets attracted 169K new wallets
To assess whether RWA Markets are attracting new users to Hyperliquid, we identified every wallet whose first-ever trade on the platform occurred on a HIP-3 RWA Market on or after January 1, 2026, referred to as RWA-First wallets. The remaining new wallets, those who entered through Other Markets (defined as crypto perpetuals or any non-RWA Market) are referred to as Other-First wallets. Wallets with any prior trading activity on Hyperliquid, or any USDC deposit into the Arbitrum bridge before that date, were excluded. 
This leaves a cohort of 534,362 new wallets active between January and June 2026. Among them, 169,514 wallets, representing 31.7% of all new wallets, executed their first trade on an RWA Market.
RWA-First onboarding remained relatively modest at the beginning of the year before accelerating from February onwards. Daily inflows generally ranged between 300 and 700 new wallets, punctuated by several distinct onboarding waves linked to major product launches and macro-driven trading opportunities.
The largest event occurred in mid-March following TradeXYZ's launch of the S&P 500 perpetual under an official S&P Dow Jones Indices license, bringing more than 38,000 new wallets to Hyperliquid in just eight days. Earlier spikes in late February and early April reflected a different catalyst. Heightened geopolitical tensions surrounding the Iran conflict increased demand for oil and silver exposure, with Hyperliquid emerging as one of the few venues offering continuous 24/7 trading on these markets. The final major wave, in mid-June, coincided with the listing of the SpaceX pre-IPO market.
The onboarding profile of Other-First wallets follows a different pattern. Daily inflows remained consistently higher, typically between 1,500 and 2,500 new wallets, with peaks largely tracking broader crypto market activity rather than individual RWA listings. This contrast suggests that the two cohorts are responding to different demand drivers throughout the period.
All of this leads us to the first conclusion: RWA Markets are bringing new users onto the platform.
...but they don't move the revenue needle
If RWA Markets account for 31.7% of all new wallets joining Hyperliquid, a broadly similar contribution to trading activity would be expected.
On trading volume, this expectation largely holds. Between January and June 2026, the 169,514 RWA-First wallets generated $111.6 billion in trading volume, representing 31.5% of the $354.2 billion traded by all new users during the period. Their share of trading activity is therefore closely aligned with their share of new user acquisition.
However, the picture changes when looking at fee generation. Over the same period, the RWA-First cohort generated $34.1 million in fees out of the $412.6 million paid by all new users, accounting for just 8.3% of the total.
The explanation, as discussed in the following sections, lies in two factors: the trading behavior of the cohorts and the fee structure of the assets they trade.
RWA-First wallets trade RWA products…
The trading activity of the RWA-First cohort is highly concentrated in the products through which they entered the platform. Of the $111.6 billion in trading volume generated between January and June 2026, $93.2 billion (83.6%) was executed on RWA Markets, while just $18.3 billion was traded across crypto perpetuals and Other Markets.
The evidence suggests that RWA-First users remain largely focused on RWA products rather than expanding into the broader Hyperliquid ecosystem. RWA Markets therefore appear to attract users who might not otherwise have joined the platform, although cross-market participation remains limited.
…while the crypto base is trading it all
The behavior of the Other-First cohort differs significantly. Of the $242.6 billion in trading volume they generated over the period, $187.3 billion (77.2%) was traded on Other Markets, their primary entry point. The remaining $55.3 billion (22.8%) was executed on RWA Markets, with this share increasing steadily as the RWA offering expands. As of today, RWA Markets account for approximately 30-35% of the cohort's trading volume.
This pattern translates into a dominant position for the Other-First cohort across both market segments. On Other Markets, primarily crypto perpetuals, Other-First wallets generate around 90% of total trading volume. At the same time, they also account for roughly 40% of trading volume on RWA Markets, making them a major source of trading activity in that segment as well.
The contrast between the two cohorts is clear: RWA-First users remain largely specialized in RWA products, whereas Other-First users are considerably more asset agnostic, readily expanding into newly listed markets. For Hyperliquid, this means RWA listings serve two distinct purposes: they attract a dedicated group of new users while simultaneously creating additional trading opportunities for the platform's crypto user base, increasing both trading volume and fee generation.
And Fees are pretty clear on this
The fee breakdown removes any ambiguity about where Hyperliquid's revenue comes from between the two cohorts.
On RWA Markets, fee generation closely mirrors trading volume. Throughout the period, RWA-First wallets generated $11.7 million in fees (60.6%), while Other-First wallets generated $7.6 million (39.4%), broadly reflecting their respective shares of RWA trading activity.
The picture changes on Other Markets. Other-First wallets generated $370.9 million (94.3%) in fees on crypto perpetual markets, compared with just $22.4 million (5.7%) for the RWA-First cohort. This gap is substantially larger than the corresponding volume split, with RWA-First wallets still accounting for 8.3% of trading volume on Other Markets.
Looking at the effective fee rate adds another data point, though it should be interpreted with caution. By dividing total fees paid by each cohort on Other Markets by their total trading volume on those same markets, we obtain 12.2 bps for RWA-First wallets versus 19.8 bps for Other-First wallets, a roughly 1.6x difference despite both cohorts trading on the same infrastructure at the same nominal fee rates. This points to a potential behavioral difference between the two cohorts on Other Markets. Without per-trade granularity, the exact cause cannot be confirmed, but several explanations are plausible. For example, RWA-First wallets may trade more passively as makers (0.015%) rather than takers (0.045%), or the two cohorts may fall into different fee tiers due to staking or differences in their 14-day rolling trading volumes. However, investigating this further falls outside the scope of this report.
Overall, the answer to the second question is clear: RWA markets are not yet a meaningful driver of fee generation for Hyperliquid, regardless of the cohort considered. More than 80% of all fees generated by new users originate from Other-First wallets trading on Other Markets. What is equally striking is that despite a rapidly expanding RWA product offering, new equity listings, the S&P 500 license, pre-IPO markets, fee generation from RWA markets hasn't grown proportionally. Both cohorts trade more on RWA Markets over time, but the fees those markets produce remain a marginal share of the total. 
Conclusion: new users, different economics
Between January and June 2026, 169,514 verified wallets entered Hyperliquid through an RWA Market, representing 31.7% of all new users. Their trading volume followed proportionally, accounting for 31.5% of all new-user activity. The evidence is therefore clear: RWA Markets are attracting genuinely new users at meaningful scale.
Those new users, however, behave differently from Hyperliquid's crypto-native user base. RWA-First users remain concentrated in the markets through which they entered, with 83.6% of their trading volume staying within the RWA Markets. By contrast, crypto-native users are considerably more asset agnostic, already directing around 30-35% of their trading volume toward RWA Markets as the product offering expands.
As for fees, several key takeaways emerge. First, RWA Markets account for less than 10% of the platform's fee revenue among new users despite generating meaningful trading volume. Second, RWA-First users generate relatively few fees because they largely remain within RWA Markets. Third, when they do venture into crypto perpetuals, they tend to adopt a more conservative trading profile, resulting in lower fee generation. Most importantly, more than 80% of all fees generated by new users come from wallets that entered Hyperliquid through Other Markets and keep trading on these markets.
The next question is therefore no longer whether RWA Markets can attract new users but whether those users become more economically valuable over time by expanding into crypto perpetuals, increasing their activity on RWA Markets or as the RWA Markets keep evolving.
Methodology
This analysis covers the period January 1 to June 30, 2026. A wallet is classified as an RWA-First user when its first ever trade on Hyperliquid occurs on a HIP-3 RWA Market, specifically one of the following builder-deployed DEXs: vntl, xyz, flx, km, cash, para, abcd. To ensure the cohort contains only genuinely new participants, each candidate wallet was cross-checked against on-chain Arbitrum bridge deposit history. Any wallet that had deposited USDC into the Hyperliquid Arbitrum bridge before January 1, 2026 was removed. The two groups, 169,514 RWA-First wallets and 364,848 Other-First wallets, together account for 534,362 wallets active in the window.
Volume and fees are tracked across all days each wallet traded, not just their first. Both are split across a 2x2 matrix of wallet group (RWA-First vs Other-First) and market type (RWA Markets vs Other Markets). Volume is reported as single-side USD notional. Fees reflect only the main Hyperliquid trading fee, builder fees and HIP-3 deployer fees are excluded. The analysis processed 1.44 billion fill events across 582 GB of decompressed data from the public Hyperliquid raw fill archive.
Full methodology and raw data : 
https://gist.github.com/noateden/6ee849a5348c99636681164878c40694

Does opening Hyperliquid to RWA Markets attract net new users, or does it mostly give existing traders something else to trade?

Hyperliquid and the RWA onboarding question: what the data shows

As traditional assets move onchain and the boundary between crypto and traditional finance becomes less distinct, CEXs are positioned to take another step: becoming integrated access and infrastructure layers through which users and institutions can move between asset classes, settlement systems, and onchain applications.

Centralized crypto exchanges have evolved from simple marketplaces for digital assets into integrated financial platforms combining trading, payments, lending, yield products, custody, and institutional services.

CEXs are therefore evolving beyond trading venues into integrated access and infrastructure layers connecting multiple asset classes, financial products, settlement systems, and onchain applications for both retail and institutional participants.

From Exchange to Infrastructure: How CEXs Are Reshaping Global Market Access

Alan Orwick

Quai Network

Alan Orwick is the co-founder of Dominant Strategies, the company behind Quai Network. His work focuses on energy-based money, scalable Proof-of-Work infrastructure, and the machine economy taking shape around AI agents, distributed compute, and autonomous payments. Before Dominant Strategies, Alan worked at Apple on supply-chain analytics and studied distributed-ledger scalability at UT Austin. He is based in Austin, Texas.
Alan is building toward a future where money is globally accessible, programmable, and anchored to measurable energy expenditure rather than institutional promises. Under his leadership, Dominant Strategies launched Quai Network mainnet, introduced Qi as a “hashrate-linked decentralized energy dollar,” and developed SOAP, a merge-mining protocol that lets miners from other Proof-of-Work ecosystems contribute work to Quai while external rewards fund QUAI buybacks and burns.
Much of the discussion around new blockchain launches today revolves around TVL, users, or transaction volume, whereas Quai focuses more on revenue. Why do you think the industry has been optimizing for the wrong scoreboard?
 The simple answer is that TVL, users, and volume are easy to observe and easy to grow with incentives. They aren’t useless metrics. They can tell you whether people are trying a product, whether capital is present, and whether the system is active. But the problem is they’re easy to game, and they don’t always tell you whether the network is creating durable value.
Revenue asks a more direct question. Did someone pay to use the system because it produced something valuable? That’s why we think it deserves more weight. Revenue isn’t perfect, but it’s harder to manufacture than deposits or volume, which can be inflated with subsidies or the promise of an airdrop. That kind of mercenary capital is liable to vanish once the subsidy ends.
The point is not that TVL, or users, or volume should be ignored. The point is they should support the revenue story, not replace it. If activity only exists while incentives are high, that is customer acquisition. If usage produces revenue after the incentives fade, that’s a much stronger signal of real demand.
At first glance, SOAP looks like another buyback-and-burn mechanism, but it seems to be after something bigger. How does it actually work, and what problem were you trying to solve when it was first conceptualized?
 The problem SOAP is tackling is that Proof-of-Work mining creates sell pressure, since miners have real operating costs in the form of hardware and electricity. That is normal, and the point is not that miners should hoard. The issue is that the network's token has to absorb recurring sell pressure while the protocol has no matching source of recurring buy pressure. For example, Bitcoin does roughly $30 million a day in mining rewards with greater than 80% of that typically sold into the market. 
SOAP was designed around a different question: can the network capture the benefits of Proof-of-Work while mitigating miner sell pressure? With SOAP, Quai offsets these emissions through subsidies from Bitcoin Cash, Litecoin, DOGE, Ravencoin, and eventually Bitcoin itself.
These mining subsidies from other networks become protocol-funded demand for QUAI. The miner still gets paid for useful work, but part of the value created by mining is converted into buy-and-burn for QUAI holders and the long-term security budget. That was the core insight behind SOAP.
The playbook today is still very much “spend money on incentives or promise an airdrop to attract users,” but SOAP flips that by trying to make the network earn money instead. Can you walk us through that strategy, philosophically and economically?
 The key point is that a Proof-of-Work blockchain is already going to pay emissions for security. Those rewards are the cost of getting miners to do work and produce blocks. In the normal model, the chain pays that cost, miners receive the reward, and if miners sell, the network's asset absorbs the sell pressure without a matching revenue source.
Merge-mining changes the accounting. It lets the same mining infrastructure produce useful work for more than one Proof-of-Work network. SOAP uses that by routing the external merge-mining reward through Quai. The outside-chain reward is sold at market, and the proceeds are used to buy and burn QUAI. In plain terms, SOAP tries to attach revenue to emissions the chain was going to pay anyway.
Philosophically, that is different from paying users with incentives or promising an airdrop. Those programs spend the network's own asset to create activity. SOAP starts from the security budget a Proof-of-Work network already needs, then asks how much external revenue can be captured around that mining activity. Economically, the loop is external mining rewards fund buy-and-burn, buy-and-burn offsets some of the emission pressure, stronger asset economics can attract more hashrate, and more hashrate strengthens security.
With SOAP, the Quai Network captures a large chunk of the value it creates. . Why not let that value flow entirely to miners, validators, or traders?
 Miners, validators, and traders all provide useful functions. They should be paid for those functions. But if all of the value created by the network immediately flows out to service providers or short-term participants, the protocol itself does not build long-term strength.
Retained value is how the network aligns itself with long-term holders and users. In Quai's case, SOAP buy-and-burn accrues value to the asset rather than directing all of the economic benefit to miners at the moment of block production. That does not mean miners are less important. It means the network and its long-term participants also share in the value being created.
This matters for security as well. In Proof-of-Work, the value of the asset is connected to the security budget. A stronger asset can support a stronger mining ecosystem. So value retention is not only a financial design choice. It is also part of how the network supports long-term security.
Crypto has largely moved toward Proof-of-Stake over the past few years. You chose to build a scalable Proof-of-Work network instead. What assumption do you think the rest of the industry made that you simply don't agree with?
 We disagree with the assumption that Proof-of-Work and scale are fundamentally incompatible. That assumption pushed much of the industry toward Proof-of-Stake. Our view is that the right architecture can preserve the advantages of Proof-of-Work while still scaling throughput.
Proof of Entropy Minima (PoEM) and merged-mined execution shards are our answer to that. They let Quai add capacity while sharing one security pool. The goal is to avoid the tradeoff where each new chain or shard fragments security.
We also disagree with the idea that energy is only a cost to minimize. In Proof-of-Work, energy is the external input that ties the system to the physical world. That matters for Quai because the monetary design depends on a real-world cost of production. Proof-of-Stake is a valid design path for many systems, but it does not give you the same energy anchor. For energy-linked money, that anchor is essential.
The Quai team seems rooted in the belief that every form of money ultimately traces back to energy, whether that's gold, fiat, or Bitcoin. How did you arrive at that conclusion, and why does it matter enough to build a blockchain around it?
 The starting point is that durable money usually has a cost behind it. Gold has value partly because it is hard and expensive to extract. Fiat has value because states spend real resources enforcing tax systems, legal systems, borders, and monetary policy. Bitcoin made the production cost explicit through mining.
Those systems are very different, but they all connect money to work in some way. Quai takes that observation seriously and makes the link direct. Qi is designed so issuance tracks difficulty, which is the protocol's measure of the work required to secure the network.
That is why the blockchain has to be built around Proof-of-Work. If the goal is energy-linked money, the system needs a measurable energy input. Qi uses that input as the basis for a medium of exchange whose monetary policy is tied to production cost rather than a central authority.
Qi is an attempt to tie monetary policy to the cost of energy rather than to the decisions of a central authority. If that idea succeeds, what changes about how people think about money?
 It changes money from something people understand mainly as a policy decision into something closer to a production-cost measurement. Today, most currencies depend on central bank decisions, government balance sheets, and policy targets. That means people's savings are exposed to decisions they cannot control.
Qi is designed around a different reference point. Its issuance is tied to the work required to produce blocks. That does not remove every source of volatility or complexity, but it gives the currency a rule based on energy rather than discretion.
If that works, people can start thinking about money in more physical terms. A unit of money is not just a promise from an institution. It is connected to the cost of producing and securing it. That is especially important globally, because energy is a universal input. No country owns the concept of a joule.
The team intentionally spent more time building the network's infrastructure before introducing much of a DeFi ecosystem. Looking at the industry today, do you think we've been too focused on building financial products before building the economic foundations they depend on?
 Yes, in many cases. DeFi has built very sophisticated financial products, but a lot of those products depend on assets that do not have external cash flows. When the main source of yield is another incentive program, the system can become very reflexive. It grows when prices and rewards rise, and it contracts when they fade.
Our view is that sustainable DeFi needs a stronger economic base underneath it. That means money with credible monetary policy, security with a real cost behind it, protocol revenue, and assets that produce yield from activity outside the financial loop.
That is why Quai's order of operations matters. First build the monetary system. Then build a scalable Proof-of-Work system. Then build value capture through SOAP. Then build financial products around assets like liquid mining tokens and mining cash flows. DeFi is more useful when it sits on top of productive economic activity.
That said, we just introduced our first foray into DeFi and launched Quainance on July 8th. It marks an important milestone, as it has been heavily requested by our community and we’ve been putting all of the building blocks in place to ensure the perfect environment was ready for it. 
To reduce risk, the incentives we’ve provided with it are ownerless, as are the deployments to the LP factory. With SOAP revenue now in place, our long-term goal is to drive TVL of USDT for the QUAI/WQI pairs for more overall liquidity in the ecosystem.
Looking ahead five or ten years, if there's one idea behind Quai that you believe the rest of the industry will eventually adopt, what do you think it will be?
 I think the idea is that crypto's highest-value use case is still money. Not points, not governance, not abstract blockspace, but a private, digital monetary asset that can stand on its own. The industry has spent a lot of time building financial applications around tokens, but the foundation is the money itself.
For that to work, the asset needs two things. It needs privacy, because money that exposes your entire financial life is not really cash. And it needs a real cost of production, because neutral money should be anchored to something outside politics and outside its own token economy. That is why Proof-of-Work matters. It connects the monetary asset to energy, and energy is the most universal cost in the world.
So the broader idea I think the industry adopts is energy money. A monetary system where the base asset is produced through real work, secured by real energy, and usable privately as money. For Quai, QUAI is the store-of-value asset, Qi is the energy-linked medium of exchange, and Proof-of-Work is what makes both of those claims credible. SOAP fits into that by making the network economically productive, but the deeper thesis is that the next major crypto asset is money with privacy, energy, and Proof-of-Work at the core.

For Quai, QUAI is the store-of-value asset, Qi is the energy-linked medium of exchange, and Proof-of-Work is what makes both of those claims credible.

Quai Network co-founder on fixing the Proof-of-Work economy and the launch of Quainance

Patrick Scott

0xngmi

LlamaSwap

There are hundreds of decentralized exchanges. Traders looking to swap tokens face an overwhelming number of choices.
DEX aggregators make traders’ lives a little easier, taking orders and scanning select exchanges for the best quotes. But they have their own fees, and their own methods of identifying the best trades.
Enter LlamaSwap, the self-described “aggregator of aggregators.” 
Launched by DefiLlama in 2023, LlamaSwap queries the price of trades on 1inch, CoW Swap, and several other aggregators. If those aggregators offer the cream of the crop, LlamaSwap is an attempt at offering la crème de la crème. 
Traders have taken notice. In 2025, LlamaSwap processed $50 billion in trades. In another sense, however, it has flown under the radar. Uniswap, the world’s largest decentralized exchange, has processed more than $36 billion in trades in the past 30 days alone.  
So let’s shine a spotlight on LlamaSwap, one of the premier venues for trading crypto. 
Free!
First and foremost, LlamaSwap is free to use. 
It doesn’t charge its own fee, so the price offered is the same price a trader would pay if she went directly to the chosen aggregator. 
But LlamaSwap does generate revenue. Some aggregators profit off “positive slippage” — if a trade executes at a superior price to the one quoted, the aggregator pockets the difference. And some of those aggregators share a portion of that revenue with LlamaSwap.
Lest anyone worry that LlamaSwap only routes trades to aggregators with a revenue share agreement, rest easy — in its attempt to find traders the best price, LlamaSwap also scans aggregators that don’t share their revenue. 
LlamaSwap has grown since its 2023 debut, and it’s no longer just an exchange aggregator. Its “earn” function displays yield opportunities for thousands of assets. And its “borrow” function aggregates lending paths, offering two varieties based on a user’s risk tolerance: “safe” and “degen.” That lets users easily see which lending protocols support a particular token. It also makes it easy to compare lending terms to get the best deal when borrowing crypto. 
Finally, LlamaSwap introduced a feature called “gas refuel” last year. It ensures traders never get stuck without the crypto needed to execute trades. 
Staying safe

Hacks continue to plague the DeFi ecosystem, and DefiLlama founder 0xngmi has called his approach to building LlamaSwap “extremely paranoid.”
Some users have executed nine-figure swaps, 0xngmi said on X earlier this year. It demonstrates an incredible level of trust in the product, and for that reason he personally audits updates to the LlamaSwap frontend and the contracts that users approve. 
“To this date our users have NEVER had an approval hack, despite doing almost $100 billion in volume over the years and being the most-used meta aggregator,” he wrote.
“DefiLlama is extremely cautious with which aggregators are listed,” Patrick Scott, DefiLlama’s chief revenue officer, said. 
“Swaps are executed directly against the smart contracts of each aggregator, so there's no additional smart contract risk of using LlamaSwap.” - Patrick Scott
That also means LlamaSwap users are protected from hackers that exploit aggregators’ websites, as happened early this year when the CoW Swap website was compromised. 
That isn’t the only way LlamaSwap protects users. It also prevents them from executing terribly disadvantageous trades. Sadly, there was a prime example of this back in March, when someone used Aave’s website to swap $50 million in USDT for AAVE tokens.
The trader likely expected to receive about $50 million in AAVE tokens. Instead, extreme slippage on the multi-step trade meant the trader received just $36,000 worth of AAVE. 
Aave founder Stani Kulechov said the trader had been warned they would suffer “extraordinary slippage” and moved ahead with the trade anyway. Kulechov also said Aave would attempt to return the $600,000 in fees the transaction generated. 
LlamaSwap takes a different approach. 
“We directly forbid swaps like this in the UI and you wouldn't be able to make them,” 0xngmi wrote in March. 

There was a less extreme example just this month, when a trader swapped $2 million in Ether for $14,000 in LIT using 0x, a popular aggregator.

But sometimes, users do have to protect themselves. 

Scammers have tried to swindle LlamaSwap users by creating bogus websites that look and feel like LlamaSwap but serve only to drain users’ wallets. Google has been slow to remove suspect websites, so anyone using LlamaSwap needs to be vigilant — always make sure you’re using swap.defillama.com. It can also be accessed by clicking on the “DEX aggregator” button on DefiLlama’s main page.

Traders have taken notice. In 2025, LlamaSwap processed $50 billion in trades. In another sense, however, it has flown under the radar. Uniswap, the world’s largest decentralized exchange, has processed more than $36 billion in trades in the past 30 days alone.

Our latest research

What the last two generations of reserve proofs missed

Blue Wave, Same Regulators: What a Democrat-led Congress Means for Crypto Regulations

What HTTPS did for the Internet, Zama is doing for Onchain Finance

Tokenized Stocks Have Volume. Do They Have Liquidity?

Gate Money: From Crypto Exchange to All-in-One Financial Account

Prediction or roulette? Kalshi’s trades are moving to the edges of the price curve

Decentralization, Permissionlessness, Neutrality and Censorship Resistance: DeFi’s Founding Principles Put to the Test

How Flare uses data to expand what capital can do

Ethena's Expansion Playbook: Payments, Basis Trading, and Wall Street's ENA Bet

Card Compare dashboard tracks crypto cards’ cashback, fees, and more

Sandwich attacks and the limits of private routing: what THORChain does differently

What percentage of DeFi hacks could AI reproduce today?

What the last two generations of reserve proofs missed

What HTTPS did for the Internet, Zama is doing for Onchain Finance

Gate Money: From Crypto Exchange to All-in-One Financial Account

Decentralization, Permissionlessness, Neutrality and Censorship Resistance: DeFi’s Founding Principles Put to the Test

Ethena's Expansion Playbook: Payments, Basis Trading, and Wall Street's ENA Bet

Sandwich attacks and the limits of private routing: what THORChain does differently

Blue Wave, Same Regulators: What a Democrat-led Congress Means for Crypto Regulations

Tokenized Stocks Have Volume. Do They Have Liquidity?

Prediction or roulette? Kalshi’s trades are moving to the edges of the price curve

How Flare uses data to expand what capital can do

Card Compare dashboard tracks crypto cards’ cashback, fees, and more

What percentage of DeFi hacks could AI reproduce today?