What appears to be a one-click stablecoin deposit on a DeFi front end can hide a much larger machine underneath. In a long-form analysis translated by TechFlow from research by Sealaunch Intelligence, the authors argue that a single deposit may pass through 14 protocols, span multiple chains, and be levered and rebalanced several times before the user ever sees a yield number. The central question is straightforward: who is actually allocating that capital, who shares the risk, and who captures the economics along the way?
Vaults are not standalone products but part of a liquidity graph
The report describes a vault as a smart contract that accepts deposits and puts them to work, usually by allocating capital into DeFi protocols. Users deposit assets and receive share tokens, whose value rises over time if the strategy performs well.
Structurally, the authors say, a vault resembles a fund. It pools capital, shares returns, operates within a defined mandate, and has a manager that charges fees. The difference is in execution. Instead of relying on brokers and custodians, vaults allocate directly into DeFi protocols and, in many cases, into other vaults.
No vault exists in isolation, the report says. A better way to understand their role is to view them as nodes in a graph linking depositors, managers, destination protocols, held assets, and other vaults they own or are owned by. Two vaults that appear unrelated can still be connected through the same protocol, the same asset, the same manager running both portfolios, or the same chain.
The piece uses an island analogy: vaults may look separate above the waterline, but below the surface they are connected through a shared sea of liquidity.
In practice, a user may deposit USDC or USDT into a vault, which then spreads capital across Aave, Morpho, Fluid, Pendle principal tokens, Maple pools, tokenized Treasuries, and sometimes other vaults, often across several chains. For the depositor, it is a single action. Underneath, it is a moving set of allocations that keeps changing and rebalancing.
In crypto, firms that manage these portfolios are often called curators. The report argues that the label is too narrow. Their work includes credit risk management, market parameter construction, and asset allocation. They define mandates, choose venues, size positions, and bear responsibility when things go wrong. For that reason, the article refers to them as strategy managers, a term closer to the role of a portfolio manager in a multi-strategy fund.
To map the broader picture, Sealaunch Intelligence charted leading multi-protocol vaults. The sample covers more than 100 vaults representing over $14 billion in deposits and more than $6 billion in DeFi borrowing. Each line in the chart represents one vault allocating to one protocol. The study excludes single-protocol vaults and focuses only on vaults that deploy across multiple venues.
Allocation complexity is rising fast
Early vaults looked more like aggregators. They searched a small set of lending markets for the best yield, moved stablecoins between them, and offered users a blended rate.
That model has expanded sharply. According to the report, a vault today may hold or express exposure through:
- fixed-rate exposure via Pendle principal tokens
- tokenized Treasuries, private credit, and a growing set of tokenized equities
- liquid staking assets
- shares of other ERC-4626 vaults
- carry trades
- levered looping positions created by borrowing against collateral and redepositing
Across the sample, vaults have exposure to 14 protocols on average, while the broadest approach 50. The report says this breadth has little correlation with size. Vaults managing less than $100 million often allocate across more protocols than vaults twenty times larger.
As these ledgers widen, overlap increases. Two vaults with different strategies, different managers, and different depositors may still end up funding the same DeFi market. The article notes that a $30 million vault can appear in more markets than a $1 billion vault. That creates second-order exposure: part of one vault’s risk can come from other vaults that also fund the same venue, and that can show up in another vault’s rate.
Looping is one common example. A vault may hold a yield-bearing asset, post it as collateral, borrow stablecoins, buy more of the same asset, and repeat the process to increase leverage. That turns one deposit into several positions, so a vault’s footprint in a market can be much larger than the capital it originally raised. Exiting is not a single-step withdrawal either. It is a sequence.
The report says this structure could become more important as real-world assets and tokenized equities move on-chain at larger scale. Once those assets become active collateral, the same carry structures can be built on Treasuries or equity tokens rather than only on stablecoins. The authors say this matters for fintech firms and new banks that already hold idle customer balances. They point to ether.fi Cash as the clearest current example: users keep their assets, borrow against them, and then spend or reinvest the borrowed funds, while the underlying lending protocol functions as a financing venue.
Distribution is moving from retail to embedded channels to B2B
The article breaks vault distribution into three stages.
In the first stage, vaults waited for deposits from on-chain users who actively compared annualized yields and came to the product themselves.
In the second, strategy managers moved behind someone else’s front end. They still ran the vaults, but they also competed to win the distribution channel. The report cites Sentora allocating funds for users depositing through Kraken Earn. Partnerships like that can reach millions of users who may not know what a vault is, how their money is routed across DeFi protocols, or what strategy sits underneath.
In the third stage, the consumer disappears from the chain entirely. The client is another business: an asset manager, a treasury, or a fund bringing its mandate on-chain. The endpoint is no longer retail. Vaults become execution infrastructure sold to institutions that deploy either their own capital or client capital. Defined mandates and clear reporting fit structures those clients already understand, so vaults become an on-chain extension of existing operations. The report says this capital tends to arrive in larger tickets, more slowly, and leave more slowly as well.
For lending protocols, the outcome is the same across all three stages. Capital arrives through strategy managers, and the protocol is simply one position inside a vault portfolio. Buyers compare net yield after cap fees, oracle risk, and exit liquidity, and they can rotate positions in a single trade.
Once managers control real supply, liquidity itself becomes tradable
The report then turns to what it calls a three-sided trade. Once strategy managers control real capital supply, that supply becomes something issuers can buy access to. Asset issuers want distribution for their products. Managers control the balances needed to provide it. The incentives used to close the deal are passed through to depositors rather than kept entirely by the manager.
In theory, each side gets something useful:
- depositors receive yields above the base market rate
- issuers gain access to supply they could not directly purchase on their own
- strategy managers can offer more competitive products without using their own balance sheet
In practice, the report says, the arrangement only works as long as someone keeps paying. Supply acquired through incentives is rented for the life of the budget. When incentives stop, rates fall back toward the underlying market level.
Sometimes that is acceptable because the incentives bought enough depth and integration for demand to persist at a lower unit cost. Sometimes nothing durable was built, and the rented liquidity leaves as soon as the program ends.
The authors argue that issuers have the strongest reason to understand which of those outcomes they are buying, because they are the party capable of paying for it. Incentives are, in effect, a slice of issuer profit redirected through vaults to end users in exchange for distribution.
The harder question is where to place those incentives. They can sit at the top of the stack, paid directly to the vault and visible in the APY breakdown. They can also sit lower down, inside a market the vault allocates to or inside an asset the vault holds. As the stack deepens, the number of possible layers multiplies, and each layer creates another place where the payment can be captured by depositors the issuer did not intend to target. Put incentives at the wrong layer, the report says, and all that is purchased is a number that lasts as long as the budget. Put them at the right layer, and what is bought is depth or integration that can remain after the budget ends.
Thin markets are more exposed to rebalancing shocks
Strategy managers are now among the largest sources of supply in DeFi lending markets, according to the report, and their moves are large enough to reprice markets quickly. When one manager rebalances, utilization drops in the market losing funds and borrowers see rates change immediately. In the market receiving funds, utilization rises and depositors see yields compress.
How strongly a market feels that shock depends on two things: how much of its liquidity sits inside vaults, and how much of that vault supply moves at once.
The report defines the first as market exposure, E, calculated as total vault supply divided by total market liquidity. The second is the effect of a partial withdrawal of those allocations. Rate changes, the authors write, are proportional to three factors: exposure, the share of vault supply that actually moves, and the steepness of the rate curve at the current utilization point.
The variables are defined as follows: Aᵢ is the supply provided by vault i to the market, L is total market liquidity, m is the fraction of vault supply that actually moves, and k is the slope of the rate curve at the current utilization level.
Two markets can pay the same rate and still have very different exposure. A $2 billion market with 10% of supply coming from vaults may see only a small rate move even if all of that capital exits. A $50 million market with 70% vault supply can reprice violently when a single manager rebalances. Borrowers who were paying a reasonable rate in the morning may be the first to feel the shock.
That makes liquidity depth a competitive advantage rather than a vanity metric. As strategy managers get larger and move faster, the report says, this matters more. A deep market can absorb a large allocation, price it, and survive its exit. A thin market may display a more attractive headline rate, but the round-trip experience is worse because the same flow that pushes rates up can tear the market apart on the way out.
This dynamic favors shared-liquidity designs over isolated markets. In isolated markets, a single vault can account for most of the supply, making E structurally high and causing every rebalance to hit other participants. In shared-pool or hub-and-spoke designs, the same allocation enters a much deeper book and represents only a small share of it, so the same decision barely moves rates. Isolated markets isolate credit risk but concentrate liquidity risk. Shared pools and hybrid designs do the reverse. As supply concentrates in fewer managers, the report argues, the cost of the isolated model is rising.
Managers, protocols, chains, and issuers are all being forced to adapt
The final section lays out where the authors think this structure is heading.
First, strategy managers are consolidating, and some will likely be acquired. Winning distribution is a business development function, the report says, and business development naturally favors a small number of names. An on-chain asset manager may prefer to buy a team that already has client mandates, tools, and a track record rather than build from scratch. Others may be disintermediated instead, as their largest clients bring the mandate back in-house to save on management fees.
Second, protocols need more active protocol management. Rate curves, caps, and reserve factors were designed for a deposit base made up of many small decisions that entered and exited in roughly similar ways. The shape of capital flows is no longer that simple.
Third, chains need to compete for allocation rather than just users. Vaults abstract the chain away. A user may deposit on Ethereum while the capital ultimately lands on Base. In that setup, attracting users does not automatically mean attracting capital. What matters is whether a chain offers the conditions required by delegated strategies: acceptable collateral assets, venue depth, and incentive terms that make allocation possible.
For issuers, the key question is what exactly they are paying for. Incentives buy allocation, the report says, but allocation is not the same thing as a durable deposit base. The distinction between supply that remains after the budget ends and supply that leaves with it determines whether the issuer paid for distribution or merely for a temporary number.
Connection growth is outpacing simple growth in size
The report closes by arguing that several trends are usually covered separately and each one sounds positive on its own. More assets are moving on-chain, from Treasuries and private credit to tokenized equities. More capital is being allocated through vaults rather than direct deposits. More vaults are launching, and each one runs more strategies than the previous generation. More strategy managers are competing for that capital, while the larger ones continue to get larger.
Viewed together, however, the picture changes. Every new asset can be reached by every vault whose mandate allows it. Every new vault becomes another allocator into the same set of markets. Every new manager becomes another actor whose decisions can affect the rates paid by people they have never met. The report’s conclusion is that connections are growing much faster than any single count, which is why the graph is becoming denser faster than it is becoming larger.
Understanding that liquidity map can help different participants identify where the next deposit may come from. A protocol can see which strategies are capable of allocating to it and how much room they still have. An issuer can identify which managers are already able to hold its asset, which the report describes as the shortest path to distribution. A chain can see what conditions are still missing before a given strategy can deploy there. The risk question follows from the same map: who else is in your market, and what would need to happen somewhere else for them to leave?
The article notes that Sealaunch Intelligence has operated independently since 2021 and provides research on on-chain capital markets for protocols and institutions, covering on-chain credit, yield, RWA, and stablecoins.
