Institutional Flows Push DeFi Lending Toward Risk-Isolated, Modular Credit Models

Institutional capital is beginning to reshape onchain lending, pushing DeFi credit markets away from monolithic liquidity pools and toward modular designs built around 'risk isolation'—a shift that could determine how quickly tokenized real-world assets (RWAs) scale on public blockchains.

That is the central conclusion of a new report from Tiger Research, which argues that both traditional finance and DeFi have repeatedly suffered from the same structural weakness: when many assets and risk parameters are commingled in a single pool, losses tied to a specific instrument can cascade into system-wide stress. As institutions increase their participation in onchain credit, Tiger Research says the main competitive axis is moving from simple 'liquidity provision' to the quality of 'risk segregation' and the capabilities of an emerging 'operations layer' that sits above core protocols.

The report anchors its thesis in a historical parallel. In September 2008, the Reserve Primary Fund—then the world’s third-largest money market fund—“broke the buck” after losses on Lehman Brothers debt that represented only around 1.2% of its portfolio. Redemption requests reportedly surged to roughly $40 billion within two days, forcing an unprecedented suspension of withdrawals. For Tiger Research, the episode illustrates how a concentrated impairment inside a pooled structure can trigger reflexive runs, liquidity spirals, and broader confidence shocks.

Traditional finance responded over time by separating risks into more granular compartments. The U.S. Securities and Exchange Commission overhauled money market fund rules in 2014, segmenting funds by investor base and liquidity profile and adding buffers tailored to each type. Hedge funds also migrated away from single-broker dependence toward multi–prime brokerage arrangements, distributing custody, collateral management, financing, and risk monitoring across multiple counterparties. The underlying logic was consistent: reduce single points of failure by layering and disaggregating key functions.

Early DeFi, by contrast, compressed those layers into a single smart-contract system. That approach improved automation and reduced operational overhead, but it also concentrated underwriting, collateral management, liquidation, and governance in a single protocol surface. In stressed markets, abrupt price moves or liquidation bottlenecks can therefore threaten the entire pool, encouraging governance to adopt conservative collateral rules. Tiger Research argues this is one reason lending markets historically gravitated toward only the most liquid assets such as Bitcoin (BTC) and Ethereum (ETH), limiting broader collateral diversity.

As RWAs have moved onchain—ranging from tokenized Treasuries to private credit—those limitations have become harder to ignore. RWAs often differ by trading hours, price-oracle requirements, legal constraints, and liquidation processes, making them difficult to manage under one standardized parameter set. Tiger Research says this is accelerating the migration toward architectures that isolate risk at the asset or market level, rather than aggregating everything into one shared pool.

The report points to Silo Finance as an early proof of concept: independent pools per asset, designed to prevent stress in one collateral type from infecting others. More recently, Tiger Research argues the market has begun to formalize a division between an immutable 'infrastructure layer'—the code that executes settlement and liquidation—and a more discretionary 'operations layer' responsible for asset selection, exposure limits, parameter tuning, and ongoing risk oversight. In the firm’s view, this separation is not merely a technical evolution but a prerequisite for scaling institutional participation.

Morpho is cited as a leading example of this model. Morpho Blue is structured as a base protocol where key market components—collateral asset, loan asset, loan-to-value ratio, price feed, and interest-rate model—are fixed at market creation, emphasizing predictability at the infrastructure level. Risk decisions, however, are increasingly delegated to Morpho Vaults, where designated curators set supply caps and manage asset selection. Tiger Research likens the arrangement to the division of labor in prime brokerage: code executes, while specialized entities make judgment calls on risk.

That tradeoff introduces new vulnerabilities. Because protocols do not inherently provide custody, institutions often rely on external custodians such as Coinbase ($COIN) and Anchorage. Meanwhile, outcomes can hinge on curator competence—an issue underscored by losses tied to xUSD and Stream Finance in 2025, which the report describes as illustrating how poor asset selection can quickly degrade vault performance. In response, Tiger Research observes that institutional flows have tended to concentrate with established risk managers and curators including Steakhouse, Gauntlet, and Sentora.

Institutional adoption is also becoming more visible in product design. Tiger Research says Coinbase and Binance have incorporated Morpho Blue into parts of their lending infrastructure. SG Forge has deployed MiCA-compliant stablecoins EURCV and USDCV, while Apollo has used its private credit fund ACRED as onchain collateral. Asset manager Bitwise has also participated directly in risk curation, highlighting how the 'operations layer' is becoming a key interface for regulated and professional participants. Tiger Research frames the dynamic succinctly: tokenization increases access to assets, but onchain credit rails are what make those assets usable within lending markets.

Aave (AAVE) is taking a different route. Rather than full isolation, Aave V4 is pursuing a hub-and-spoke architecture intended to balance containment with capital efficiency. In this design, the hub centralizes liquidity and accounting, while spokes operate as asset-specific markets with independent parameters. If a spoke experiences stress, its impact is constrained within credit lines defined by the hub—an approach Tiger Research compares to how universal banks manage exposure limits across business units.

The structure may be especially advantageous for RWAs, the report argues, because new tokenized markets often struggle to bootstrap liquidity from scratch. By leaning on hub liquidity, spokes can be launched without needing each asset to attract a standalone pool of lenders on day one. Aave’s institution-oriented RWA initiative, Horizon, reflects that direction: issuers handle KYC and AML, risk specialists such as LlamaRisk propose diligence and parameters, Chainlink (LINK) supplies price feeds, and the protocol focuses on execution. The result is a dispersion of responsibilities away from a generalized DAO and toward function-specific specialists.

Euler (EUL) V2 represents a third model: a flexible system combining isolation and composability. After a roughly $197 million exploit in 2023, the project opted for architectural redesign rather than a simple recovery. Tiger Research highlights two core components. EVK enables the creation of credit vaults with asset-specific risk settings, while EVC links collateral and debt positions across different vaults into a single account framework—aiming to preserve independence while allowing controlled connectivity. The report compares the concept to multi-strategy hedge fund “pods,” where teams run distinct risk books under a broader framework.

Use cases for Euler are expanding quickly, according to Tiger Research. Through collaboration with Ondo Finance, Euler launched a “STEY” market that accepts tokenized equities such as SPYon, QQQon, and TSLAon as collateral while offering PayPal’s stablecoin PYUSD as a borrowable asset. In another example, KPK’s USDC Prime RWA vault allows collateral including VanEck’s tokenized Treasury product VBILL and Securitize’s AAA-rated CLO product STAC. The overarching significance, Tiger Research says, is that these structures can accommodate RWAs with differing pricing conventions and liquidation rules—constraints that single-parameter pools handle poorly.

While Morpho, Aave, and Euler differ in emphasis—Morpho on speed and choice, Aave on capital efficiency, and Euler on flexible risk balancing—the report argues they share a common direction: separating execution from judgment and engineering bespoke risk conditions per asset class. In Tiger Research’s view, that convergence signals a broader market re-rating in which differentiated value accrues less to base protocols and more to whoever can manage risk most effectively above them.

Tiger Research expects the next phase of competition to center on the 'operations layer'—including risk design, collateral underwriting, regulatory workflows, and track record—much as prime brokerage became commoditized infrastructure while hedge fund performance became the differentiator. The firm estimates total assets under management in onchain curation vaults at about $7.4 billion, suggesting the segment is already material and still early in its institutional adoption curve.

The report also underscores a critical difference from traditional prime brokerage: onchain infrastructure is broadly permissionless. Where legacy finance concentrated power among a handful of large intermediaries, DeFi allows virtually anyone to build markets and strategies atop protocols such as Morpho or Euler. If this modular shift continues, Tiger Research concludes, the long-term winners in onchain credit may be less about who owns the protocol and more about who can consistently deliver robust 'risk management' at scale.