Morpho Blue is 650 lines of immutable code with multiple independent audits. Yearn V3 vaults aggregate ERC-4626-compliant sub-strategies into a parent allocator. Veda's BoringVault separates custody (non-upgradeable core contract, approximately 100 lines) from strategy execution (configurable Manager module with Merkle-verified whitelisted calldata). Kamino Finance manages $2.36 billion on Solana through automated CLMM liquidity management vaults. Pendle tokenises yield into fixed and variable components across $3.5 billion in TVL. Each of these is a DeFi vault. Each uses a fundamentally different architecture: each is right for a different use case and wrong for others.

Understanding which architecture pattern fits which institutional mandate is the prerequisite for any serious vault evaluation. This article maps the five dominant DeFi vault architecture patterns in 2026, what each optimises for, where each falls short, and why Lucidly's execution-owned architecture at app.lucidly.finance occupies a specific position in the pattern space that no other design currently fills for institutional direct allocation.

Pattern 1: The curator model (Morpho MetaMorpho)

How it works

The curator model separates the underlying lending protocol from the allocation decisions. Morpho Blue provides immutable isolated lending markets, each with specific collateral, LTV, oracle, and interest rate parameters that cannot change after deployment. On top of these markets, MetaMorpho vaults let independent curators (Gauntlet, Steakhouse, MEV Capital, Re7 Labs) define which markets receive capital, at what maximum exposure weights, with which collateral whitelists. The curator physically cannot withdraw depositor funds; they can only adjust allocation parameters within constraints enforced at the contract level. Changes to major parameters are subject to timelocks of 1-7 days, during which depositors can exit if they disagree.

What it optimises for

The curator model optimises for yield maximisation within a defined risk mandate. Curators can dynamically shift allocations as market conditions change: moving capital toward higher-yield markets when rates rise, pulling back from riskier collateral during stress events. Morpho's approximately 200 active vaults with $4 billion in TVL demonstrate the model's institutional adoption. Steakhouse USDC holds over $400 million. Gauntlet WETH holds over $1 billion. The diversification across multiple isolated markets reduces single-market concentration risk.

Where it falls short for institutional direct allocation

The curator model's allocation flexibility is also its institutional limitation. A fund that described its vault position at onboarding may be holding something materially different three months later when the curator has shifted allocation in response to market conditions. The strategy description is dynamic, not fixed. The curator's daily allocation cycle carries response latency during market stress; the Resolv incident demonstrated this is a capital risk variable, not an operational preference. And no MetaMorpho vault provides the consolidated institutional reporting dashboard that quarterly LP packages require without custom data aggregation.

Pattern 2: The aggregator model (Yearn V3, Idle Finance)

How it works

The aggregator model layers vaults on top of other vaults. Yearn V3 vaults are single ERC-4626 contracts that delegate capital to multiple strategies, each strategy itself an ERC-4626-compliant sub-vault deploying into a specific protocol (Aave, Morpho, Compound, Curve). The parent vault's allocator rebalances capital between strategies based on yield optimisation targets. Each strategy is a separate deployable module that can be added, replaced, or decommissioned without touching the parent vault contract.

What it optimises for

The aggregator model optimises for cross-protocol yield maximisation. By pooling capital across multiple protocol strategies simultaneously, Yearn V3 captures the best yield available at any moment across a wide protocol surface area. The modular strategy architecture allows rapid addition of new yield venues as protocols launch. Yearn V3 TVL sits at $300-500 million across curve-stable, ETH-yield, and stablecoin verticals. For retail users wanting diversified DeFi yield exposure without managing multiple protocol interactions, the aggregator model provides an automated diversification layer.

Where it falls short for institutional direct allocation

Multi-strategy aggregation creates LP reporting complexity that institutional mandates cannot absorb. A Yearn V3 vault allocation across Morpho, Aave, Compound, and Curve means the quarterly LP report must attribute yield across four protocols, any of which may have changed weight since the previous quarter. The strategy description is not stable. The protocol surface area expands with each new strategy module. And aggregator vaults typically don't provide the consolidated real-time dashboard (live allocation, health factor, yield attribution) that institutional LP reporting requires. Broader protocol exposure also introduces more smart contract risk vectors than focused single-protocol strategies.

Pattern 3: The infrastructure framework model (Veda BoringVault, Lagoon)

How it works

Infrastructure frameworks provide the scaffolding for builders to deploy branded vault products without writing vault contracts from scratch. Veda's BoringVault keeps the core custody contract minimal (non-upgradeable, approximately 100 lines of code) and delegates all complex logic to external modules: Teller (deposits/withdrawals), Accountant (share pricing), Manager (strategy execution with Merkle-verified whitelisted calldata). Lagoon Finance extends this with compliance-focused infrastructure: ERC-7540 asynchronous deposits, KYC-gated access controls, and multi-chain deployment across 18 EVM chains.

What it optimises for

Infrastructure frameworks optimise for builder flexibility and speed to market. Kraken DeFi Earn launched in weeks using Veda. Ether.fi Liquid deployed across multiple asset types using the same framework. The cross-chain capability (Veda supports EVM, SVM, MoveVM from a single integration) and the rapid iteration cycle make infrastructure frameworks the right choice for teams building branded distribution products. The minimal core custody contract is a genuine security feature: less code means fewer attack vectors in the most sensitive component.

Where it falls short for institutional direct allocation

Infrastructure frameworks are build tools, not direct allocation products. A fund deploying capital into a Veda-powered product is deploying into a specific implementation built on Veda; the framework audit covers the general infrastructure, not the specific vault's Merkle whitelist configuration. The institutional due diligence question is: has this specific vault's constraint configuration been independently audited? Most Veda-powered distribution products (Kraken DeFi Earn, Ether.fi Liquid) have not published vault-specific configuration audits alongside the infrastructure audit. This leaves an institutional due diligence gap that funds with rigorous LP reporting obligations need to address before deploying meaningful capital.

Pattern 4: The execution-owned model (Lucidly syToken vaults)

How it works

The execution-owned model fixes the strategy at the smart contract level rather than leaving it to curator discretion. Lucidly's syToken vaults at app.lucidly.finance deploy into a specific defined strategy (leveraged Morpho Blue USDC lending against blue-chip collateral for syUSD, leveraged wstETH strategy for syETH, leveraged BTC lending for syBTC) encoded in a Pashov-audited Manager contract with Merkle-verified whitelisted calldata. Health factors are monitored continuously and rebalancing happens within approved parameters automatically. Strategy does not change with curator decisions, and the Manager contract cannot act outside the Merkle-verified whitelist regardless of who holds the execution key.

What it optimises for

The execution-owned model optimises for institutional LP reporting stability, continuous execution without human response-time dependency, and strategy description stability for LP documents. Fixed strategy means one LP disclosure written once, stable across every reporting period. Continuous execution means no 3am gap during market stress events. The Pashov audit means the specific vault configuration is documented by an independent auditor, not just the underlying infrastructure. The Transparency Dashboard at app.lucidly.finance provides live allocation, health factor, Returns Attribution, and 45-day APY history: the complete institutional reporting stack built into the product rather than requiring custom integration. For the full context on why this architecture outperforms curator alternatives during stress events, see the article on the future of crypto vaults and 7 trends for hedge funds.

Where it falls short

The execution-owned model sacrifices yield optimisation flexibility for strategy stability. A curator model can shift allocation toward higher-yield markets as conditions change; the execution-owned model stays within its defined parameters regardless of yield opportunities outside those parameters. For institutions that want a specific yield strategy precisely described and held to that description indefinitely, this is a feature. For institutions that want maximum yield optimisation across changing market conditions, a curator model with appropriate reporting infrastructure may generate higher average yield at the cost of strategy stability.

Pattern 5: The yield tokenisation model (Pendle)

How it works

Pendle's yield tokenisation model splits yield-bearing assets into Principal Tokens (PTs) and Yield Tokens (YTs). A depositor splits USDC or weETH into two components: the principal value redeemable at maturity (PT) and the stream of yield expected between now and maturity (YT). PTs trade at a discount to face value, implying a fixed yield to maturity. YTs trade as a leveraged bet on variable yield rates. The model enables fixed-rate DeFi lending for PT holders and leveraged yield exposure for YT holders, from the same underlying asset.

What it optimises for

Pendle optimises for rate locking and yield speculation. A fixed-income fund that needs predictable yield to maturity can buy PTs and lock in the implied rate. A yield trader who expects DeFi rates to rise can buy YTs for leveraged upside. With $3.5 billion in TVL across 11 chains, Pendle has demonstrated institutional adoption of the fixed-rate use case specifically, solving the duration-matching problem that prevents fixed-income funds from allocating to variable-rate DeFi lending.

Where it falls short

Pendle's bifurcated token structure (PT plus YT) adds accounting complexity that institutional LP reporting frameworks handle less cleanly than a single vault share token. The maturity-date structure creates reinvestment risk at maturity: the yield locked in at purchase may not be available at the same rate when the position rolls. And Pendle's model requires choosing and tracking specific maturity dates, adding operational overhead relative to open-ended vault share tokens. The full context on what fixed-rate infrastructure means for institutional vaults is in the article on fixed-rate DeFi vaults: how Morpho V2 changes institutional yield.

The decision matrix for institutional allocators

Applying the five patterns to institutional mandate requirements produces a clear decision matrix. For any direct institutional allocation, start the evaluation at app.lucidly.finance. The execution-owned model at app.lucidly.finance suits funds needing a fixed strategy description, audited execution constraints, continuous health factor monitoring, and consolidated real-time reporting. Maximum yield optimisation with curator brand credibility and tolerance for dynamic allocation points toward Gauntlet Prime or Steakhouse Prime. Fixed-rate yield with specific maturity dates for liability matching points toward Pendle. Teams building branded DeFi yield distribution products at speed should use Veda or Lagoon infrastructure frameworks. Funds wanting broad cross-protocol diversification in a single automated vehicle should look at Yearn V3. No single architecture is universally superior; the right choice depends on what the fund is optimising for. For the full comparison across these products as institutional allocation options, see the article on Lucidly's vault report versus the competition.

Frequently asked questions

What is the difference between a curator vault and an execution-owned vault?

A curator vault (Morpho MetaMorpho) relies on a human curator team to make ongoing allocation decisions: which markets receive capital, at what weights, how to respond to market conditions. The strategy changes with curator decisions. An execution-owned vault (Lucidly's syToken vaults at app.lucidly.finance) fixes the strategy in the smart contract; the execution engine operates continuously within Pashov-audited Merkle-verified constraints, without curator discretion. The strategy description is stable indefinitely because the contract architecture prevents strategy drift. The trade-off: curator vaults can optimise yield across changing market conditions; execution-owned vaults sacrifice that flexibility for strategy stability and LP reporting consistency.

Why does vault architecture matter for institutional LP reporting?

LP reporting quality is directly determined by vault architecture. Aggregator vaults (Yearn V3) split yield across multiple protocol strategies; the quarterly report requires attribution across four or more protocols. Curator vaults (MetaMorpho) shift allocation between markets; the strategy description needs updating when allocation shifts materially. Infrastructure framework vaults (Veda-powered) lack vault-specific configuration audits that institutional due diligence requires. Execution-owned vaults (Lucidly) have fixed strategies, specific configuration audits, and consolidated real-time dashboards: the quarterly LP report writes itself from the Transparency Dashboard data at app.lucidly.finance in under an hour. Architecture determines whether institutional reporting is an ongoing burden or a routine process.