Contrary to the narrative that DeFi is dead, a new report from CoinShares and Token Terminal shows deposits in tokenized real-world assets more than tripled, rising to $7.4 billion. The same report shows total DeFi deposits fell about 15%, while DEX spot volume collapsed about 70%. On its face, the signal is obvious: money is rotating from volatile crypto-native markets into real-world collateral. But the data never supports a single obvious story. The report also states that tokenized assets on-chain are worth more than $40 billion. That $32.6 billion gap between issued assets and active deposits is not an accounting footnote. It is the actual thesis.
The headline number, $7.4 billion, matters less than what it measures. This is not a wave of new users. It is a capital preservation trade wearing a blockchain costume. Data is the only witness that does not spin; it simply sits in the ledger and waits for someone to read it correctly.
Context: What The Report Actually Counted
Let me be precise about who published these numbers. CoinShares is a European digital asset manager. Token Terminal is an on-chain analytics platform. Both have commercial incentives to frame data in terms of institutional progress. That does not mean the report is wrong, but it means the analyst’s job is to strip the framing and test the methodology.
The report is not measuring the full tokenized-asset universe. It is measuring a specific subset: tokenized assets that are actively deposited into DeFi lending markets, liquidity pools, or yield vaults. The $40 billion tokenized asset figure is a broader issuance number. The $7.4 billion figure is an activity number. When you separate those two, the report becomes much more useful and much less euphoric.
In my audit workflow, I ask three questions before I trust any on-chain dashboard. What is the collateral? What is the claim? Who can stop the redemption? For a token like BUIDL, the collateral is a pool of cash, U.S. Treasuries, and repurchase agreements. The claim is a right to redeem the token at the fund’s net asset value. The party that can stop redemption is the fund administrator or the issuer. That is not a criticism of BlackRock. It is a structural fact.
The same logic applies to sUSDS, the yield-bearing savings asset from the Sky ecosystem. The yield does not appear because a smart contract invented it. It appears because the Sky collateral system generates returns, and those returns are passed through to sUSDS holders. The smart contract distributes the yield, but the real-world or crypto collateral behind Sky is what creates it. If that collateral becomes impaired, the distribution mechanism becomes a stress-test, not a source of value.
There is also a ticker in the report that I want to flag: JTRSY. I attempted to reconcile this ticker against public market datasets, token registries, and on-chain explorers. I could not verify it. That does not mean the token does not exist. It means the report contains an unverified label. A professional report should either provide the contract address or explain that JTRSY is a thematic basket. Passing along an unmatched ticker is the kind of data-hygiene failure that later becomes a footnote in an investigation.
Core I: The $40 Billion Overhang Is Not Demand
The most important number in this report is not $7.4 billion. It is the gap between $40 billion of issued tokenized assets and $7.4 billion of active deposits. That gap means more than 80% of tokenized assets are not currently deployed in DeFi. They are not being lent against. They are not supplying liquidity. They are not trading on DEXs. They are sitting in wallets and custody accounts as tokenized representations of traditional financial instruments.
Why does this matter? Because it changes the interpretation of growth. When a report says RWA deposits tripled, it is describing a smaller segment of the tokenized market: the segment that has actually been integrated into DeFi. That is a genuine development, but it is not the same as saying the tokenized asset market is booming. The $40 billion issuance figure is still mostly dormant capital.
Based on my audit experience, teams love to report gross asset values as if they reflect on-chain activity. They do not. In 2022, before the Terra collapse, the project had enormous notional value on its books. The collapse did not show up first in the market cap of the stablecoin. It showed up in reserve outflows and in the silent movement of large wallets. The same discipline applies here. Do not count what is issued. Count what is moving.
A tokenized Treasury product can issue $1 billion in tokens and never touch DeFi. The issuance alone does not create borrowing demand, fee revenue, or liquidation events. The active deposit figure does. So when I see $40 billion of tokenized assets and only $7.4 billion of active DeFi deposits, I conclude that the RWA-DeFi integration is still early. The infrastructure exists. The liquidity does not yet.
Core II: The Yield Is Imported, Not Native
The growth in RWA deposits is being driven by yield-bearing products. Remove BUIDL, sUSDS, and JTRSY from the dataset, and the narrative collapses. These are not assets that pay yield because of DeFi trading activity. They are assets that earn yield because of traditional financial mechanisms.
BUIDL is BlackRock’s tokenized USD liquidity fund, built on Ethereum through Securitize. It holds cash, U.S. Treasuries, and repurchase agreements. It pays a yield to token holders, updated daily as the fund’s net asset value changes. sUSDS is the yield-bearing version of Sky’s stablecoin, designed to pass through returns from the Sky system. JTRSY is harder to verify, but the report groups it with other yield products.
All of these products share one feature: the yield originates outside the DeFi protocol. A tokenized Treasury product earns yield because the U.S. government pays interest on debt. A money-market fund token earns yield because the fund holds short-term instruments. Sky earns yield from lending assets, including real-world collateral, and passes that return to sUSDS holders. DeFi is the distribution channel. It is not the source of the yield.
This distinction matters for sustainability. During DeFi Summer, high farm yields came from newly minted governance tokens. Many of those yields were subsidies printed by the protocol. They looked real until the emission schedule ended. Tokenized Treasuries are different. The underlying yield is an obligation of a real economic counterparty. But the counterparty is also the weak point.
If the fund manager blocks redemptions, the yield promise means very little. If the Federal Reserve cuts rates, the yield falls. A smart contract cannot renegotiate the central bank’s policy. It can only execute the rules that someone wrote after the outside world made the monetary decision.
Understanding how each protocol’s oracle, liquidation threshold, and reserve factor handles a fund NAV change is what I mean by decoding the algorithmic chaos of DeFi yield traps. The token contract is the least dangerous part of the system. The risk is in the gap between the token’s on-chain price and the net asset value of the underlying real-world portfolio.
Core III: The Lending Layer Is The On-Ramp
The report identifies Aave, Morpho, and Kamino as the protocols with the deepest RWA liquidity. That is not an accident. RWA assets are not primarily being swapped on DEXs. They are being used as collateral in lending markets.
Consider the trade. Buy one unit of a tokenized Treasury product. Deposit that token into Aave or Morpho. Borrow USDC or USDT against it. Use the stablecoin to buy more of the tokenized Treasury product. Deposit it again. The result is a leveraged carry position. It works as long as the yield on the tokenized product exceeds the cost of borrowing the stablecoin.
The structure feels like DeFi. The collateral is real-world. But the margin call is not. If the net asset value of the fund falls, the lending protocol’s oracle must immediately reprice the token. If the oracle delays, borrowers are temporarily overcollateralized on paper. If the oracle reprices too slowly, they become undercollateralized in practice. The protocol then faces a choice: liquidate at a stale price or absorb the risk.
Aave has a large multichain lending market, which makes it a natural first stop for institutions that want to deposit BUIDL and borrow stablecoins without creating bespoke bilateral agreements. Morpho offers vaults with granular risk parameters, which allows different lenders to take different views on the same RWA token. Kamino plays a similar role in Solana’s ecosystem. Together, these protocols are the bridge between custody and composability.
The problem is that the lending layer is also the failure layer. Aave and Morpho are not just neutral venues. They are where the value of a tokenized asset is tested every hour. If the issuer reprices a fund downward, the lending protocol must convert that repricing into a collateral factor. If it does not, the market will trade the token at a discount to NAV until someone forces a correction.
Core IV: The Danger Of A 220% Increase From A Tiny Base
The report notes that RWA spot volume rose roughly 220%. That number is true, but it is incomplete. The overall DEX spot volume fell roughly 70% in the same period. A 220% increase from a tiny base is not evidence of a liquid market.
Let me make the point concretely. If the baseline RWA volume was $2 million per day, a tripling produces $6 million per day. In a market that once cleared billions of dollars per day, $6 million is a rounding error. I do not know the exact baseline from the report, but the presentation matters. Percentage changes are useful, but professional analysis should also show dollar volume. Without that, the 220% figure is closer to marketing than measurement.
The danger of low-base volume is liquidity fragmentation. A tokenized Treasury product is not designed for 24/7 trading. It is designed to be held. If a liquidation event hits, the DEX order books and lending pools that look deep could still be shallow relative to the size of the fund. That creates the classic exit liquidity problem. When the NAV drops, the protocol will try to sell the token into a market that is far less deep than the issuer’s balance sheet. Price impact will be extreme.
This is why I separate issuance from activity. A $40 billion tokenized market sounds robust. But if only $7.4 billion is active in DeFi, and only a fraction of that is actually traded, the true liquidity is much smaller than the headline suggests. In a stress event, the liquidation engine will not care about the fund’s AUM. It will care about the depth of the order book at the moment the token is sold.
If the trade reverses, reconstructing the timeline of a rug pull exit will not begin with a suspicious wallet. It will begin with a failed redemption request and an oracle that did not reprice the token until six hours later. The blockchain will record every step. The narrative will not.
Core V: Trust Assumptions And The Real Security Model
A tokenized U.S. Treasury is not a trustless asset. The token contract on Ethereum can be immutable. The chain can resist censorship. But the value of the token depends on a custodian holding securities, a fund administrator calculating NAV, and a legal process that permits redemption. This is not a critique of BlackRock. It is a structural feature of RWA tokenization.
When I audit a protocol, I ask what happens if the redemption function fails. For a crypto-native asset like ETH, if the token contract fails, the asset still exists on the chain. For a tokenized Treasury product, if the issuer fails to redeem the token, the token becomes a claim against a legal entity. The chain cannot compel the entity to pay.
That creates a hybrid trust model: decentralized settlement layer, centralized collateral layer. That model can work, but it is not equivalent to a smart-contract-native collateral pool. Investors who do not understand the difference are buying protocol risk parameters without understanding issuer legal risk.
The report does not disclose the audit status of the relevant smart contracts. It does not disclose which oracles are used to price BUIDL or sUSDS in lending markets. It does not explain what happens if the fund administrator pauses redemptions during a market crisis. These are not minor omissions. They are the exact failure points that determine whether RWA-backed lending survives its first real stress test.
The market is currently calm, so the missing details do not seem urgent. But calm markets reward the analyst who prepares for the first broken promise. The collapse of Terra was not caused by a single bug. It was caused by a series of assumptions about liquidity, redemption, and trust that failed in sequence. RWA markets are building the same stack. The difference is that the underlying assets are more stable. The trust model is still centralized.
Contrarian: RWA Deposits Are Rising Because DeFi Is Falling
Now the contrarian part. RWA deposits are not rising because DeFi is succeeding. RWA deposits are rising because DeFi is falling.
Consider the environment. DEX spot volume is down roughly 70%. Total DeFi deposits are down roughly 15%. Risk appetite is shrinking. Where would a rational institution put money? Into a volatile lending pool with opaque liquidations, or into a tokenized money-market fund that pays yield and can be used as collateral in Aave? The answer is obvious.
The RWA growth we are seeing is not a new population entering DeFi. It is the existing population abandoning volatile DeFi in favor of a safer, more boring version of finance. The report can be read as RWA in DeFi growing alongside DeFi. But the same data can be read as RWA cannibalizing DeFi.
If the RWA deposit base were additive, we would expect total DeFi deposits not to fall by 15%. The fact that the total declined while the RWA subset grew suggests money left other DeFi venues. That is substitution, not complementarity. The migration is real. The innovation is smaller than it appears.
Correlation is not causation. Aave may have more BUIDL deposits, but if its total stablecoin borrow demand is falling, protocol fee revenue could still be declining. Deposits are not revenue. Utilization is revenue. A deposit that simply sits in a lending market and earns yield from a money market fund does not pay Aave. The value that accrues to AAVE, MORPHO, or KMNO depends on the fee model and on borrow volume. The report does not provide enough data to confirm that tokenized collateral has produced higher protocol revenue. Therefore, do not infer that the RWA tripling is automatically bullish for governance tokens.
The token economics of Aave, Morpho, and Kamino are not fully disclosed in the report. No unlock schedule, no team allocation, no treasury balance. Without those numbers, any statement about whether a token’s price is justified by RWA growth is speculative. What I can say is this: the deeper the RWA collateral pools grow, the more important it becomes for protocol governance to adjust collateral factors, loan-to-value ratios, and liquidation strategies. Governance tokens capture value only if they translate collateral growth into fee generation or a return of value to holders. The report does not establish that this has occurred.
There is also a conflict-of-interest concern. CoinShares has a commercial interest in institutional adoption of digital assets. Token Terminal sells data to institutions. A report titled RWA growth is also a marketing document. That does not mean the data is fabricated, but it means the report should not be treated as an independent audit.
The Hidden Regulatory Risk
Securities law is the unexamined risk in every RWA story. Under the U.S. Howey test, a tokenized money-market fund likely satisfies the four elements: an investment of money, a common enterprise, an expectation of profit, and profits derived from the efforts of others. If BUIDL is a security, every DeFi protocol that lists BUIDL as collateral is handling a security.
That brings the protocol’s governance token, front-end, and contributors into the same regulatory orbit as the fund manager. The legal risk may be manageable for licensed entities, but it is not manageable for an anonymous DAO with no KYC layer. A regulator that wants to stop RWA adoption does not need to ban blockchain. It can simply rule that unregistered DeFi lending markets cannot accept unregistered securities as collateral.
The system would not collapse. It would go permissioned. And a permissioned RWA market is a very different investment thesis from the one the headlines suggest. Institutional money is not entering this market because it wants to evade the law. It is entering because it believes the legal wrapper cleans the asset. But the DeFi layer that accepts the asset is often not as clean as the fund itself.
I am also watching the redemption mechanism. Tokenized funds have terms that govern when redemptions are processed. Some process daily. Some weekly. Some can delay if markets become stressed. In a crisis, the fund may honor a redemption request at a price that is worse than the token’s market price. When that happens, every protocol using the token as collateral must decide whether to mark the token at the new realizable value or at the stale NAV. That decision will determine which liquidations are triggered and who absorbs the loss.
The report does not address this scenario. It does not address the speed of oracle updates. It does not address the legal jurisdiction of the tokenized fund. It does not address what happens if the issuer is headquartered in a country that imposes capital controls. For an institutional-grade report, these are significant gaps.
Takeaway: The Next Signal Is Not Another 220% Jump
The market is in chop. The old crypto-native growth engine is still contracting. RWA deposits are a counter-cyclical signal, but the business model is not yet proven.
The next data point I am watching is not total tokenized assets. I am watching active deposits on Aave and Morpho, utilization rates in the BUIDL and sUSDS lending pools, and the stablecoin borrow rate relative to the tokenized product yield.
If the tokenized Treasury yield is 5% and the stablecoin borrow rate is 3%, the carry trade works. If the Fed cuts rates and the tokenized Treasury yield falls to 2.5%, the carry trade flips. At that moment, borrowers will not want to borrow stablecoins against BUIDL. They will want to redeem BUIDL. RWA deposits in DeFi may fall as quickly as they rose.
The data will show this before the narrative does. Active deposits will drop. Borrow rates will spike. The protocols with the deepest liquidity will suddenly become the narrowest exit doors. The next real test is not another tripling of deposits. It is the first redemption request under stress, the first oracle deviation, the first time a fund administrator delays a payout.
After Terra, after the DeFi fee collapse, and after every narrative cycle, I still tell people the same thing: read the chain, not the headline. The chain recorded the growth. It will also record the exit. Reconstructing the timeline of that exit will not be a mystery. It will be a matter of watching the blocks.