The lending market doesn't need a lead model
Original title:CLAB Lending:
Original by: @0xJaehaerys, Gelora Research
Original language: EeeVee, SpecialistXBT, Block Beats
Editor: The DeFi community is experiencing a painful ghost after the thunderstorms of Stread Finance and USDX. The Curator model, introduced by agreements such as Morpho and Euler, was intended to address the fragmentation of mobility, but inadvertently re-invited the moral hazard of “humans” back into the chain. According to the author, the current lending agreement wrongly binds the "risk definition" to the "order-laying". By drawing on the traditional financial order book model, this paper has developed a new paradigm that does not require a master, automatic algorithm。
Evolution logic of the lending market
Looking back at the evolutionary history of chain transactions can inform our understanding of the lending market。
AMM based on a constant function (e.g. Uniswap) solves a fundamental question: How can a market be created without active business? The answer is to pre-set the "shape" of mobility using the constant function. Liquidity providers agree in advance on a set of strategies, which are automatically implemented by agreement。
This is effective in the area of transactions, which are relatively simple: a buyer and seller meet at a certain price. But borrowing is much more complicated. A loan has several dimensions:
Interest rate
Type of collateral
LOAN VALUE RATIO (LTV)
duration (fixed vs in demand)
Liquidation mechanisms
The combination of borrowing needs to meet all the dimensions mentioned above。
Early DeFi loans directly followed an AMM-like solution. Agreements like Compund and Aave preset interest rate curves, and lenders join a shared pool. This allows the lending market to operate without active lenders。
BUT THIS ANALOGY HAS A FATAL FLAW. IN DEX TRANSACTIONS, THE SHAPE OF A CONSTANT FUNCTION CURVE AFFECTS THE QUALITY OF EXECUTION (SLIDES, DEPTH); IN BORROWING, THE SHAPE OF AN INTEREST RATE CURVE DIRECTLY DETERMINES THE RISK. WHEN ALL LENDERS SHARE A POOL, THEY SHARE THE RISK OF ALL COLLATERAL RECEIVED BY THE POOL. LENDERS ARE UNABLE TO EXPRESS THEIR WILL TO ASSUME A PARTICULAR RISK ONLY。
In the area of transactions, the order book solves the problem: it allows the market to define its "curve shape". Each vendor offers prices at a comfortable price, and the order book consolidates these prices into a single market, but each comptoir still controls its own risk exposure。
Can you borrow in the same way? A project called Avon tried to answer that question。
Fragmentation of mobility
DeFi's first attempt to give the lender control was market segregation。
Agreements such as Morpho Blue and Euler allow anyone to create a lending market with specific parameters: specified collateral, loaned assets, fixed settlement loan value ratio (LTV) and interest rate curves. Lenders are placed in markets that are consistent with their risk preferences. Bad debts in one market will never spill over to another。
This is perfect for lenders, who get the risk segregation they want。
For borrowers, however, this leads to fragmentation。
IN THE CASE OF ETH-USDC LENDING, THERE MAY BE A DOZEN DIFFERENT MARKETS:
MARKET B: 3 MILLION LIQUIDS, 86% LTV, 5.1% INTEREST RATE
MARKET C: 2 MILLION LIQUIDS, 91% LTV, 6.8% INTEREST RATE
...and 9 other markets with lower liquidity
Users who want to borrow $8 million cannot be satisfied from the single market. They must compete manually, perform multiple transactions, manage decentralized positions and track different clearance thresholds. The theoretical best solution requires that loans be split into more than four markets。
No one actually does. Borrowers usually choose only one market. Funds are underutilized in fragmented pools。
Market risk segregation solves the problem of lenders but creates the problem of borrowers。
The limitations of the treasury
The treasury model seeks to bridge this gap。
The idea is that financial flows are managed by professional leaders. The lenders deposit funds in the vault, and the principals allocate the funds to the bottom markets, optimize the gains and manage the risks. Borrowers continue to face a fragmented market, but at least lenders do not have to move themselves。
It helped the lenders who wanted to lay down, but it introduced DeFi to something that was meant to be eliminated: discretion。
The principal decides which markets are funded and can be redistributed at any time. The lender ' s risk exposure changes with decision-making by the principal and cannot be predicted or controlled. As one Twitter user said, “The principal is with the borrower PvP, but the borrower does not even know that he is being harvested. I don't know
This asymmetry is reflected not only in strategy but also in the accuracy of the underlying interface. Morpho’s UI sometimes shows “three million dollars available for liquidity”, but when implemented, the funds at low interest rates are minimal, and most of them are in high interest-rate zones。
Transparency is compromised when mobility coordination depends on human decision-making。
The distribution of funds adjusts market liquidity to their timetable, rather than to immediate market demand. The vault tried to resolve the issue of borrowers' fragmentation by “rebalancing”, but rebalancing required Gas fees, relying on the will of the principal and often lagging behind. Borrowers continue to face suboptimal interest rates。
Dissociate risk from setup
The loan agreement confused two distinct modules。
User definition of risk: Different lenders have different perceptions of collateral quality, leverage ratio。
The method of arranging the loan: it's mechanical. It does not require subjective judgement from the user, but only efficient routing。
The pool model binds the two and the lender loses control。
The isolation pool model separates the definition of risk, but abandons the set-up, and borrowers have to search manually for the best path。
Through the role of the principal, the model of the treasury was re-engineered, but the presumption of trust in the principal was introduced。
Could it be automated without introducing discretion (manual intervention)
The field order book did this. A market offer is defined, the order book is aggregated in depth, and the combination is definitive (best price preferred). No one decides where to order, and the mechanism determines everything。
CLOB LENDING APPLIES THE SAME PRINCIPLES TO CREDIT MARKETS:
Lenders define risk through isolation strategies。
The strategy is to issue quotations to shared order books。
Borrowers interact with uniform liquidity。
Combination occurs automatically, without the intervention of the principal。
Risks remain in the hands of lenders and coordination becomes mechanized. There is no need to trust third parties。
Double Layer
Avon lends the order book through two distinct levels。
Policy Layer
A "strategy" is a sequestered borrowing market with fixed parameters。
THE POLICY CREATOR DEFINED THE FOLLOWING PARAMETERS: COLLATERAL/LOANING OF ASSETS, LIQUIDATION OF LTV, INTEREST RATE CURVE, PREDICTOR, CLEARING MECHANISM。
Once deployed, the shape of the interest-rate curve is immutable. The lender knows exactly the rules before deposit。
Money will never move between strategies。
IF YOU PUT IN STRATEGY A, YOUR MONEY WILL STAY IN STRATEGY A UNTIL YOU COLLECT. THERE ARE NO PRINCIPALS, NO REBALANCING, NO SUDDEN CHANGES IN RISK EXPOSURES。
While there are still people (tactical managers) who set parameters, they are fundamentally different from the principals: the principals are the money distributors (deciding where the money goes), the strategic managers are the real risk managers (defining the rules without moving the money), reference Aave DAO. Decision-making on the allocation of funds remains in the hands of lenders。
How does the system adapt to market changes? Changed by competition, not parameters. If risk-free interest rates soar, this forces old strategies to be phased out (fund outflows) and new ones created (fund inflows). The term "discretion" has been changed from "where money should go" (prior decision-making) to "what strategy I should choose" (lender decision-making)。
Bridging Layer
The strategy is not to serve borrowers directly, but rather to issue quotations to shared order books。
The order book consolidates all policy offers into a single view. Borrowers see the depth of integration of all strategies to accept their collateral。
When the borrower orders, the blending engine:
FILTER QUOTES FOR COMPATIBILITY (COLLATERAL TYPE, LTV REQUIREMENT)。
Sort by interest rate。
From the cheapest。
Settlement in an atomic transaction。
If a strategy meets the entire order, it is fully packaged; if not, the order is automatically split into multiple strategies. The borrower was aware of only one transaction。
Important tip: The order book reads only the policy state, cannot modify it. It only coordinates visits and does not have the power to allocate capital。
THE GOSPEL OF RWA
DeFi has been facing structural contradictions in institutional adoption: compliance requires isolation, but isolation stifles mobility。
Aave Arc tried the Wall Garden model, and compliance participants had their own pool. The result was low liquidity and interest rate differentials. Aave Horizon has tried the "semi-open" model (KYC for RWA issuers, but with credit waivers), which is progress, but institutional borrowers are still unable to reach the $32 billion liquidity of Aave's main pool. A number of projects explored license-type rollup. KYC process completed at the infrastructure level. This approach, which applies to some examples, may result in decentralized network-level mobility. Compliance users on chain A cannot access the mobility on chain B。
The order book model provides a third route。
ANY ACCESS CONTROLS (KYC, GEOGRAPHICAL LIMITATIONS, QUALIFIED INVESTOR CHECKS) CAN BE IMPLEMENTED AT THE STRATEGIC LEVEL. MATCHING ENGINES IS ONLY FOR MATCHING。
If both a compliance strategy and a permit-free strategy provide compatible terms, they can fill the same loan at the same time。
Imagine an enterprise bank mortgaged to borrow $100 million of its national debt:
30 MILLION FROM KYC STRATEGY (AGE FUND LP)
20 MILLION FROM STRATEGIES REQUIRING CERTIFICATION OF QUALIFIED INVESTORS (FAMILY OFFICE LP)
50 MILLION FROM A POLICY OF COMPLETE EXEMPTION
Funds have never been mixed at source and institutions are compliant, but liquidity is uniform globally. This breaks the dead end of compliance or isolation。
Multi-dimensional set-up mechanisms
The order book is tied to one dimension only: price. The highest bid matches the lowest selling price。
The loan order book must be combined in multiple dimensions:
Interest rate: Must be below the ceiling acceptable to the borrower。
LTV: THE BORROWER'S MORTGAGE RATE MUST MEET THE STRATEGIC REQUIREMENTS。
Asset compatibility: Currency matching。
Liquidity: The market is sufficiently liquid。
BORROWERS CAN MATCH MORE STRATEGIES BY OFFERING MORE COLLATERAL (LOWER LTV) OR ACCEPTING HIGHER INTEREST RATES. THE ENGINE WILL FIND THE CHEAPEST PATH IN THIS BINDING SPACE。
For large borrowers, care needs to be taken. Of Aave, $1 billion in liquidity is a single pool. And in orderbook lending, $1 billion could be spread across hundreds of strategies. A $100 million loan will quickly consume the entire order book, starting with the cheapest strategy and gradually filling in the most expensive one. The slide point is obvious。
The system based on the pool also had a slide point, albeit in different forms: the surge in usage would push up interest rates. The difference is in transparency. In the order book, the slide points are visible in advance. In the pool, the slide point is shown after the transaction is executed。
Floating interest rate and re-price
DeFi borrows at floating interest rates. Interest rates change as utilization rates change。
This presents the challenge of synchronization: if the strategic utilization factor changes, but the quotations on the order book are not updated, the borrower will be traded at the wrong price。
Solution: Continuous recharging。
Once the policy status changed, new offers were issued to the order book. This requires very high infrastructure performance:
Extremely fast out of the hour。
Very cheap transaction costs。
Atom status read。
That's why Avon chose to build on MegaETH. This structure is unfeasible because Gas is too expensive on the Internet。
Existing friction:
If market interest rates change, but the strategy's fixed curve is not adapted, there will be a "Dead Zone" – the borrower is too expensive to borrow, and the lender has no return. In Aave, the median curve is automatically adjusted, and in CLOB mode this requires lenders to draw down manually and migrate to new strategies. This is at the cost of obtaining control。
Multi-strategic position management
When a loan is filled with multiple strategies, the borrower actually has multiple strategic positions。
While the interface looks like a loan, the bottom level is independent:
INDEPENDENT INTEREST RATE: THE INTEREST RATE FOR COMPONENT A MAY INCREASE AS A RESULT OF THE INCREASED UTILIZATION OF STRATEGY A, WHILE COMPONENT B REMAINS UNCHANGED。
Independent health ratio (Health Ratio): When currency prices fall, more restrictive LTV components are partially liquidated. You're not going to explode at once, but you're going to go through a series of partial liquidations like "smuggling."。
In order to simplify the experience, Avon provides a unified position management (one-click collateral, automatic weight allocation) and one-key refinancing function (one-click loan, old and new, always locking on the best interest rates in the market)。
Conclusions
DeFi has gone through several stages:
Pool Agreement (Pooled): The borrower was given depth but the lender was deprived of control。
Isolated: Gives the lender control, but cuts off the borrowing experience。
Vaults: Trying to bridge both but introducing the risk of artificial decision-making。
ORDER BOOK LENDING (CLOB): THE ABOVE MODE OF DECOUPLING WAS ACHIEVED. RISK DEFINITION RIGHTS REVERT TO LENDERS AND ARE BROKERED THROUGH ORDER BOOK ENGINES。
This design principle is clear: human intervention is no longer required when a combination can be achieved by code. Markets can regulate themselves。
RECOMMENDED READING:
Multicoin Capital: Financial Technology 4.0
Rewrite the 18-year script, the American government's suspension is over
