From Funds to Vaults: The Evolution of Investment Vehicles

While mutual funds aggregated dispersed savers behind a professional mandate, ETFs added a secondary market and a mechanical arbitrage loop to keep price anchored to Net Asset Value (NAV). Both generations reduced friction at the point of entry. Neither fully eliminated the institutional layer sitting between the asset and the investor: the custodians, clearing houses and counterparty desks whose function was essential and whose cost was non-trivial. Onchain vaults make the same wager, but they clear the institutional overhead.

Three phases of adoption 

Institutional tokenization has passed through three distinct phases, each with a different relationship between the technology and the capital behind it. The first phase was representational: assets recorded onchain without any corresponding change to how they were financed or distributed. Demand did not materialize because a digital wrapper that cannot plug into anything is not an improvement on paper. The second phase found its footing: tokenized short-duration government debt attracted yield-seeking capital from crypto-native participants, and a genuine product-market fit emerged around accessible, transparent, onchain yield. The third phase, currently in progress, is where the architecture becomes consequential. Tokenized assets are accepted as collateral within DeFi lending infrastructure, composing with yield strategies and distribution platforms in ways that were structurally impossible in prior generations.

The collateral mechanism is what makes composition viable. When a lending protocol's risk function underwrites a tokenized asset, it activates a financing loop: the token is posted as collateral, stablecoin liquidity is borrowed against it, exposure scales, and the borrowing demand itself generates yield for depositors in the vault. That yield attracts further deposits, which deepens the market, which in turn makes the collateral destination more credible to the next issuer considering tokenization. The logic is identical to repo and prime brokerage. The onchain version simply removes the bilateral negotiation from each step.

The stack beneath onchain vaults 

Early DeFi lending protocols operated as unified pools: a single shared book, parameters set by token-weighted governance, losses distributed across all depositors regardless of which collateral triggered them. The model served its purpose in a period when the primary challenge was attracting liquidity at all. As the capital base matured and the collateral universe expanded, its limitations became structural. An undifferentiated pool cannot price risk granularly. The next generation of protocols addressed both failures by separating the lending engine from the allocation function. Isolated markets at the protocol layer, strategy and risk management delegated upward to vault curators operating within programmatic constraints.

A vault is a blockchain program, a smart contract that accepts a deposit, deploys it according to a programmatic or curator-defined strategy, and returns a token representing the depositor's proportional claim. The structure is identical to traditional fund units. The only fundamental difference is the rails. Settlement is atomic and continuous. The ledger is public. The functions that traditional structures delegate to intermediaries - margin enforcement, collateral valuation and liquidation - are embedded in the contract at inception, executing without discretion or delay. The same design that compresses cost and latency also concentrates risk.

That architectural shift produced a value chain that maps directly onto the traditional asset management stack.

• An asset originator runs the underlying strategy and brings the exposure to market. This peforms the function of any fund manager, unchanged. 
• A structuring layer packages that exposure into an onchain instrument conforming to the standards the rest of the stack can read. 
• Where a traditional structurer might produce a CLO or a credit-linked note, the onchain equivalent produces a composable token. 
• The lending protocol provides the financing infrastructure – an automated system that manages borrowing, interest accrual and liquidation based on predefined parameters, entirely removing the need for independent judgment. 
• A risk management layer, the most consequential in the chain, determines which assets qualify as collateral, sets the terms on which they are accepted, and allocates liquidity across the markets it underwrites. This is the prime brokerage margin desk, rebuilt in code.
• Finally, distribution platforms aggregate depositor capital and route it into allocation vaults, abstracting the underlying mechanics entirely. End investors receive a yield figure, not a collateral schedule.

The onchain vaults landscape in numbers 

The four protocols that dominate onchain vault infrastructure, Morpho, Fluid, Euler (EUL) and Kamino (KMNO), collectively custody a material share of institutional DeFi activity.  

(Source: Token Terminal)

Morpho is the market's dominant TVL vehicle, with $9.9bn in locked assets as of March 2026. This figure reflects its first-mover position in curator-managed vault infrastructure and its role as the primary onchain destination for tokenized real-world asset collateral. Its fully diluted valuation of $1.8bn makes it by far the most expensively priced protocol in the group. The complication is that Morpho captures no protocol revenue. The token prices future monetization entirely, a reasonable bet given the platform's structural centrality, but a pricing that demands significant faith in an eventual fee switch materializing at scale.

(Source: Token Terminal)

Fluid and Kamino sit in a middle band, each carrying fully diluted valuations slightly above the $200mn threshold against annualized revenues in the $9-10mn range, implying FDV/revenue multiples of approximately 22-24x. Fluid has been the more aggressive TVL grower across the past 12 months, scaling from under $2bn to a peak above $5.8bn by October 2025 before retracing with the broader market. Kamino's revenue has compressed steadily since its January 2025 peak of $4.5mn per month and now runs below $700,000. This contraction likely reflects both reduced activity on Solana and the gravitational pull of competition from EVM-native vaults, attracting fresh institutional flows.

Euler is the outlier that a conventional screen would surface immediately. A $26mn fully diluted valuation against annualized revenue of approximately $6.5mn, computed from the three-month trailing average, implies a revenue multiple of 4x and a TVL/FDV ratio of 41x, against Morpho's 5.4x. By any earnings-based framework, the implied discount relative to peers is steep.

(Source: TradingView)

Token performance across all four protocols has been punishing since the October 2025 peak. Morpho has held up best in relative terms, down roughly 26% from its November base, a function of its TVL resilience and dominant mind share among institutional integrators. Fluid and Kamino have given back approximately 51% and 71% respectively over the same period. Euler's token has declined 91%. Read against its revenue trajectory, this move looks less like a fundamental deterioration than a liquidity-driven reset of a low-float asset in a risk-off market.

The clock that runs against the vault

Valuation is only one dimension of the risk picture in this sector. A more structural concern runs through the vault architecture itself. Every vault is configured at a point in time: collateral types approved, exposure limits defined, yield targets set against a specific reading of market conditions. Markets do not hold that configuration valid indefinitely. Borrowing demand shifts across markets. Collateral risk profiles migrate quietly at the asset level. The yield environment that justified an allocation last week may look entirely different today, yet the vault's parameters have not moved.

The gap between a vault's current configuration and the one its risk profile actually demands is where performance deteriorates silently. It is compounded when capital crowds into the highest-yielding markets without distinguishing gross yield from risk-adjusted return, a pattern that scales exposure precisely when the underlying risk is least well-priced. The two dynamics reinforce each other: configuration lag creates the vulnerability, and concentrated inflows ensure the repricing is disorderly when it arrives. Every market participant looks to exit at the same time when risk gets recognized, drying up liquidity and cascading losses.

The Resolv exploit from last week drew in Morpho, Fluid, and Euler simultaneously, a reminder that shared infrastructure creates shared exposure. All three protocols remain operational, and the episode has accelerated the development of more sophisticated risk monitoring tooling across the sector. What it demonstrated more precisely is that the ex ante enforcement model has no mechanism for human judgment to intervene between a deteriorating parameter and its consequences. The infrastructure is still fragile and can be unforgiving, especially when it comes from features embedded at the infrastructure level.

The curated vault model faces a competitive tension that the market has not yet resolved. Fee compression on standardized strategies pushes curators toward either thinner margins or greater risk budgets. Neither path is obviously superior, and recent history suggests the latter has been the more commonly chosen. The sector's discipline is still being heavily stress-tested in real time.