I. Definition Evolution: From Crypto Leverage Tools to Mainstream Financial Infrastructure

On-chain lending is not a new concept. In 2020, Compound launched a liquidity mining mechanism, bringing DeFi from the geek circle into the public eye, thus opening the curtain on the "DeFi summer." At that time, on-chain lending was essentially a crypto-native high-leverage tool—users over-collateralized crypto assets to obtain liquidity, which was then invested in yield aggregators or liquidity provision, chasing annualized returns several times higher than traditional finance. This model operated smoothly in a bull market environment, but the 2022 collapse of Terra/Luna and the bankruptcy of FTX triggered a chain reaction that exposed the fragility of ultra-high collateralization rates and cascading liquidations. After two years of bear market reshuffling, on-chain lending has completed a key transformation from "leverage tool" to "allocation infrastructure." The driving forces behind this transformation come from three factors: first, the improvement of the regulatory environment—the MiCA framework taking effect in the EU and the SEC's gradual recognition of ETFs have cleared some compliance barriers for traditional funds entering the on-chain world; second, the wave of RWA assets going on-chain—real assets such as U.S. Treasury bonds, tokenized corporate bonds, and real estate revenue rights have begun to become core collateral for on-chain lending, changing the asset structure and user profile of on-chain lending; third, the exploration of interest rate marketization—from the initial pure floating interest rates to fixed-rate agreements (such as Notional, Yield Protocol) and then to hybrid interest rate systems (Pendle), the on-chain interest rate pricing mechanism is becoming increasingly mature and beginning to align with traditional financial markets.

By early 2026, the asset classification of the on-chain lending market has formed a clear three-tier structure: the bottom tier consists of stablecoin lending represented by USDC, DAI, and USDT, which is the largest market segment with the most controllable risks, where typical LTV can reach 80%-90%; the middle tier consists of volatile asset lending using mainstream crypto assets like ETH and BTC as collateral, with LTV usually controlled at 50%-70% to cope with the liquidation risks brought by price volatility; the top tier consists of RWA asset collateral lending, including tokenized U.S. Treasury bonds (Ondo Finance's OUSG), corporate credit (Maple Finance's private debt), and real estate revenue rights, which is becoming a new growth engine for on-chain lending, especially favored by institutional investors seeking compliant funding channels. From a regional distribution perspective, the user structure of on-chain lending is undergoing profound changes: the Asian market is dominated by individual investors and arbitrageurs, preferring high leverage and complex strategies; the European and American markets are showing a clear trend towards institutionalization, with higher demands for compliant custody, KYC verification, and audit transparency. This differentiation in user structure directly affects the functional design priorities of protocols in different regions.

II. Competitive Landscape: One Dominant Player and Diverse Strong Contenders

The competitive landscape of the on-chain lending market exhibits a typical "one dominant player and many strong contenders" characteristic. Aave holds an absolute dominant position with a TVL of approximately $32.9 billion, a figure that not only leads the second-place Compound (with a TVL of about $2.6 billion) by more than ten times but also accounts for over 50% of the total TVL in the lending sector. However, Aave's moat does not come from network effects or brand recognition—these are almost negligible in the open-source protocol world—but from its continuous technological iteration and ecosystem expansion capabilities. From Aave V1's floating interest rate model to V2's introduction of credit delegation and flash loans, and then to V3's Portal cross-chain liquidity and isolation mode, each generation of Aave's products has precisely addressed market pain points. The V4 version is expected to launch in mid-2026, further strengthening cross-chain liquidation capabilities and institutional-level compliance frameworks. In Aave's shadow, a number of differentiated protocols are seeking their own survival space. Morpho Labs has taken a unique evolutionary path—initially serving as an optimization layer for AAVE and Compound (enhancing capital efficiency through P2P matching), it has gradually developed independent Morpho Blue (no oracle, no governance lending) and Morpho Vaults (yield strategies managed by professional risk planners), transitioning from an "optimization layer" to an "independent protocol." Spark Finance, relying on MakerDAO's DSR (DAI Savings Rate) ecosystem, has established a solid user base in the stablecoin lending sector, and its technical synergy with Aave V3 makes it an important channel for institutional entry.

From a technological route perspective, on-chain lending protocols are diverging along three paths. The first path is the "liquidity aggregation" route (P2Pool), represented by protocols including Aave, Compound, and Kamino Finance, whose core idea is to pool lender funds into a shared pool and dynamically adjust interest rates based on utilization rates to achieve efficient capital allocation. The advantage of this route is ample liquidity and a simple user experience, while the disadvantage is relatively low capital efficiency (lenders cannot negotiate terms directly with borrowers). The second path is the "peer-to-peer matching" route (P2P), represented by protocols including Notional Finance and Myso Finance, whose core idea is to provide direct matching opportunities for lenders and borrowers, achieving a fixed-term, fixed-rate lending experience. This route has advantages in interest rate stability but relatively insufficient liquidity, suitable for borrowers with clear funding plans. The third path is the "permissionless pool" route, represented by protocols including Euler Finance (V2 version) and Ajna Finance, whose core idea is to completely hand over the risk management rights of the protocol to the market—no oracle price feeds, no governance voting, borrowers and lenders set parameters themselves, and bear the risks. Although this route has a higher degree of decentralization, it also faces higher user education costs and potential smart contract risks.

III. Core Risks: The Triple Dilemma of Liquidation, Credit, and Cross-Chain

The risk landscape of on-chain lending is far more complex than traditional finance. Unlike the banking system, on-chain protocols lack deposit insurance, a central bank as a lender of last resort, and regulatory guidance—when a crisis strikes, the liquidation mechanism becomes the only price discovery mechanism, and this "ruthless mechanization" often amplifies declines during market panic. The liquidation waterfall is the most typical systemic risk in on-chain lending. On March 12, 2020, "Black Thursday," the price of Ethereum plummeted 37% in a single day, triggering large-scale liquidations in MakerDAO. Due to insufficient liquidity, liquidation auctions experienced extreme phenomena of zero-price transactions, with the actual liquidation price of ETH collateral being only 50%-60% of the market price. Similar events replayed during the UST/LUNA collapse in May 2022, where multiple high-leverage positions in Aave and Compound were forcibly liquidated, further exacerbating market selling pressure. To address the risk of liquidation waterfalls, various protocols have adopted different strategies: Aave V3 introduced the "Efficiency Mode," allowing borrowers to optimize collateral efficiency for specific assets; the Isolation Mode places high-risk assets in independent pools to prevent the risk of a single asset from spreading to the entire protocol; Ajna Finance completely abandoned oracles, using the supply-demand relationship of collateral and debt for automatic pricing, placing the responsibility for price discovery entirely on the market.

Credit default risk is the second dilemma of on-chain lending. Unlike the "machine execution" model of over-collateralization, uncollateralized or under-collateralized on-chain credit lending inherently faces assessment challenges. Goldfinch and Maple Finance have adopted a hybrid model of off-chain KYC verification + on-chain settlement, scoring borrowers through real-world credit assessment agencies (such as Blackstone Credit Partners, Van Eck, etc.) to solve the problem of information asymmetry on-chain, but this "centralized endorsement" fundamentally contradicts the permissionless spirit of DeFi. In November 2022, the crypto trading firm Orthogonal Trading declared default, leaving approximately $36 million in bad debt on the Maple Finance platform, exposing the fragility of on-chain credit lending—when borrowers are institutions rather than individuals, their asset allocation and risk management capabilities vary widely, raising doubts about the reliability of "credit assessments." A deeper contradiction lies in the fact that on-chain credit lending attempts to replicate the traditional financial credit assessment system in a decentralized world, but this path faces inherent tensions between regulatory compliance (GDPR, KYC/AML) and on-chain anonymity. Establishing an effective credit assessment mechanism while protecting user privacy will be a core issue for the long-term development of on-chain credit lending.

Cross-chain security is the third dilemma. Aave's Portal feature, Morpho's cross-chain deployment, and Ajna's multi-chain expansion—the cross-chain layout of leading protocols is pushing the boundaries of on-chain lending from a single chain to a multi-chain ecosystem. However, the complexity brought by cross-chain expansion also exponentially amplifies security risks. The 2022 Ronin Bridge attack (resulting in a loss of $625 million) and the Harmony Horizon Bridge attack (resulting in a loss of $100 million) revealed how the security vulnerabilities of cross-chain bridging can transmit to the DeFi ecosystem. When Aave's V3 protocol introduces assets from chains like BNB Chain, Avalanche, and Arbitrum into its lending pool, these assets actually need to be transferred across chains through cross-chain bridges, which often have weaker security than the chains themselves. More troublesome is the dependency on price oracles for cross-chain assets—when an oracle on a certain chain experiences anomalies or delays, positions using that asset as collateral on that chain may face the risk of not being liquidated in time. This "barrel effect" means that the overall security of on-chain lending protocols depends on the weakest link among all the chains they expand to. For investors, paying attention to the cross-chain expansion strategies and bridging security of protocols is a key dimension for assessing long-term risks.

IV. Innovation Trends: Fixed Rates, RWA, and the Wave of Institutionalization

Despite the numerous risks, the innovation engine of on-chain lending has never stopped. Between 2024 and 2026, three forces are reshaping the rules of the game in this sector. The first force is the breakthrough of fixed-rate lending. The traditional P2Pool model is essentially a floating rate—interest rates dynamically adjust with the utilization rate of the pool, and borrowers may face pressure from skyrocketing interest costs when market rates rise rapidly. For enterprises and institutions seeking stable financing costs, this uncertainty is unacceptable. Notional Finance was the first to launch fixed-term, fixed-rate lending products, allowing borrowers to lock in interest rates for the next 12 months or even longer when creating loans, while lenders achieve term matching by purchasing corresponding yield certificates (fCash). Pendle Finance has taken a different approach by tokenizing yield rights—splitting the future yield of an asset into "principal tokens" (PT) and "yield tokens" (YT), allowing lenders to lock in certain yields by purchasing PT while transferring interest rate volatility risks to willing speculators holding YT. These two routes jointly promote the process of market-oriented pricing of on-chain interest rates.

The second force is the explosive growth of RWA lending. By early 2024, BlackRock's tokenized fund BUIDL surpassed $5 billion in scale, and Ondo Finance's OUSG (U.S. Treasury bond yield token) exceeded $1 billion—these compliant assets have begun to be introduced into on-chain lending protocols as core collateral. Compared to the extreme volatility of crypto assets like ETH and BTC, U.S. Treasury bonds have the triple advantages of low volatility, good liquidity, and regulatory compliance, becoming a "green channel" for institutional funds to enter on-chain lending. Protocols like Maple Finance, Pendle, and Flux Finance have supported lending with tokenized U.S. Treasury bonds as collateral, allowing users to obtain liquidity with Treasury bond positions while retaining the yield of the bonds. Aave has also specifically designed an "Institutional Market" (Horizon Institutional Market) for RWA assets in its V4 version, providing on-chain lending services for compliant borrowers registered under the SEC framework. By early 2026, the scale of on-chain RWA lending has surpassed $18.5 billion, and it is expected to exceed $50 billion by 2027.

The third force is the acceleration of institutionalization. Unlike the anonymity, permissionless nature, and complex strategies preferred by DeFi natives, institutional funds demand compliance, auditability, and controllable risks. RWA lending platforms like Centrifuge and RWA.xyz have specifically designed product frameworks to meet institutional needs: KYC/AML verification, off-chain credit assessments, custodian liquidation, regulatory reporting—these traditional financial infrastructures are being "transplanted" onto the blockchain. A deeper transformation is that the entry of institutions is changing the competitive landscape of on-chain lending. Traditional DeFi players are accustomed to leveraging, flash loans, and arbitrage strategies to extract value from protocols, while institutional funds tend to prefer a conservative strategy of "hold-borrow-hold again." This difference in strategy will lead to fundamental changes in the funding structure and interest rate curves of lending protocols—more long-term locked funds, more stable interest rate levels, and lower speculative liquidations. For protocols, how to serve institutional users well while not losing retail liquidity is a long-term balancing challenge.

V. Participation Strategies: Three Value Clues and Risk Alerts

For investors and practitioners focusing on the on-chain lending sector, the current market offers three clear value participation clues. The first clue is the extension investment in the Aave ecosystem. Besides directly holding AAVE tokens, paying attention to Morpho Labs (as an independent protocol optimizing Aave, its Morpho Blue is establishing a new paradigm of oracle-free lending), Spark Finance (a stablecoin lending protocol deeply integrated with MakerDAO, benefiting from the expansion of the DSR ecosystem), and the new features brought by the Aave V4 upgrade (such as institutional markets and cross-chain liquidation) are more risk-adjusted return options. Historical data shows that whenever Aave releases significant version upgrades or achieves historical highs in TVL, AAVE tokens often experience significant excess returns.

The second clue is the beta opportunities in the RWA lending sector. Ondo Finance (OUSG), Maple Finance (institutional credit), and Centrifuge (real asset financing) represent three different entry paths into RWA. Ondo's advantage lies in its deep integration with BlackRock's BUIDL fund and the stable yield source from compliant U.S. Treasury bonds; Maple's advantage is its established credit profiles from real institutional borrowers (Coinbase Ventures, Framework Ventures, etc.); Centrifuge's advantage is its real demand for physical asset financing and lower default rates. For investors seeking exposure in the RWA sector, a diversified allocation strategy is recommended to avoid the black swan risks of a single protocol.

The third clue is the structural opportunities in fixed-rate innovative protocols. Pendle Finance and Notional Finance represent two different fixed-rate paths: Pendle achieves "yield separation" through yield tokenization, suitable for advanced users who understand the Lego logic of DeFi; Notional achieves "interest rate locking" through traditional fixed-term loans, more suitable for institutional users seeking stability. Notably, Pendle's TVL achieved a tenfold growth in 2024, expanding from less than $100 million to over $1 billion, and the high volatility of its YT tokens also provides space for arbitrage and speculative strategies.

While chasing opportunities, three types of risks need to be closely monitored. The first is smart contract risk— the TVL scale of lending protocols makes them high-value targets for hackers; the 2023 Euler Finance attack, which resulted in a loss of $197 million, warns us that even leading protocols may have undiscovered contract vulnerabilities. The second is liquidity concentration risk—when a certain collateral (such as stETH or Lido's staked ETH) occupies too high a proportion of a protocol's TVL, extreme volatility of that collateral may trigger systemic liquidations. The third is regulatory policy risk—the "permissionless lending" feature of on-chain lending protocols may be deemed by regulators as unregistered securities issuance or illegal financing, especially under the MiCA framework in the U.S. and EU, where compliance costs will significantly increase. For allocation ratios, it is recommended to control on-chain lending exposure at 20%-30% of the overall DeFi allocation, prioritizing mature protocols that have undergone multiple audits, have robust TVL, and transparent team backgrounds.

VI. Conclusion: Infrastructure Value and Investment Clock

On-chain lending is the sector in DeFi that most closely aligns with the definition of "infrastructure." It does not pursue extreme leverage ratios like perpetual contracts, nor does it rely on token incentives for false prosperity like liquidity mining, nor does it face cyclical asset shortages like the NFT market—it derives its value from real financing needs, stable interest income, and gradually established institutional trust. Behind the $64.3 billion TVL are countless financing, deposit, and risk management activities of individuals and institutions; this scale effect of "grassroots finance" is the simplest yet most powerful value proposition of DeFi. Looking ahead, the investment clock for on-chain lending is transitioning from the "concept validation phase" to the "institutional acceptance phase." The influx of RWA assets, the establishment of institutional markets, and the improvement of compliance frameworks are all driving this sector from a playground for crypto natives to an extension battlefield of traditional finance. In this transformation process, finding the balance between "DeFi-native innovation" and "institutional compliance needs" will be key to determining the rise and fall of various protocols. For long-term investors, the on-chain lending sector deserves strategic allocation, with core positions focusing on the core assets of the Aave ecosystem, while satellite positions can moderately participate in alpha opportunities in RWA and fixed-rate innovations, while maintaining respect for smart contract risks and discipline in position management.