Morpho Midnight Solves Fixed-Rate Lending Cold Start with Intent-Based Matching

Author: Castle Labs

Compiled by Deep潮 TechFlow

DeepChain Overview: The on-chain lending market, nearly $28.5 billion, is almost entirely composed of variable-rate loans, causing borrowing costs to spiral out of control during extreme market conditions. This article breaks down how Morpho Midnight tackles the most challenging cold-start problem in fixed-rate lending through intent-based matching and tradable positions—a must-read for professionals in the DeFi lending space.

This excerpt is from our research report "Market Structure and Protocol Design" on fixed-rate lending, highlighting emerging design approaches in this space.

Today, the active loan volume in the lending sector reaches $28.5 billion, with nearly all demand coming from floating-rate lending. This works well during stable periods; however, during stress events, as the utilization curve shifts, borrowing rates can surge dramatically. Sharp increases in interest rates force some borrowers to exit or deleverage, making the overall credit market inefficient.

In fact, DeFi money markets solve a problem that traditional credit cannot: instant borrowing backed by collateral.

But there is still one unresolved issue: before the loan ends, you cannot determine the cost of the debt.

This is precisely where many products are currently focused: shifting toward fixed-rate, fixed-term lending products. In such a market, lenders receive a fixed yield and know exactly how much their deposits will earn, while borrowers know precisely how much they will need to pay.

Demand in this market can be broadly categorized into three types of participants:

Term-matching borrowers: funds, treasuries, RWA issuers, and basis/arbitrage trading desks, who require debt maturities to align with asset maturities, redemption windows, or strategy cycles.

Deterministic Borrowers: Users who engage in cyclical leverage, leverage yield strategies, and traders who may not care about exact maturity dates but require stable borrowing costs to avoid margin compression.

Lenders/Curators: Vaults, market makers, and allocators who wish to autonomously select tenor, collateral, and returns, rather than passively accept the outcomes driven by utilization rates.

The iteration of early fixed-rate lending faces three major issues:

Liquidity fragmentation: The fixed-rate market segments liquidity by maturity, interest rate, collateral, and term, making matching significantly more difficult than in a single floating-rate pool.

Cannot exit early: Once the loan term begins, it is difficult for the lender to exit before maturity unless there is secondary liquidity, a redemption pathway, or another buyer. This is not an issue in floating-rate lending.

Cold start problem: Lenders are unwilling to lock up collateral without earning returns, just to wait for a counterparty to appear.

As more institutional capital enters the market and more sophisticated strategies, such as cyclic leverage, emerge, the user base has evolved, driving growing demand for fixed-term markets. One of the primary issues with on-chain lending is the uncertainty of variable interest rates; with fixed-rate lending, users know their yields and costs from the outset. Additionally, it enhances the user experience by compelling protocols to directly price terms, collateral quality, exit liquidity, and refinancing risk.

In this article, we will examine the designs adopted by established floating-rate lending protocols, including Morpho, Jupiter, and Kamino. Together, these three platforms have $6.83 billion in active loans and have recently entered the fixed-rate and fixed-term market with the above concerns in mind.

As a longstanding protocol for floating-rate lending, Morpho launched Morpho Midnight in July. This is an intent-based, zero-coupon bond lending protocol where lenders and borrowers express their intentions, with positions represented in debt units (each representing an obligation to repay one loan token before maturity) and credit units (claims on repaid loan tokens). Midnight’s solution enables loans to be tradable, providing term flexibility and predictable underwriting for institutions. Interest rates are determined by the prices at which borrowers and lenders trade fixed-term credit and debt units.

At Midnight, lenders and borrowers post "quotes" without locking up funds; instead, they express their intent to borrow or lend at specific prices, maturities, and collateral configurations within a given market.

Funds are only deployed at settlement (upon callback), resolving the cold start problem—lenders commit capital only after a match is executed, improving capital efficiency. This also helps attract more liquidity, as the Morpho team noted:

By enabling users to earn variable interest rates on protocols like Morpho Blue, you can eliminate the typical opportunity cost of waiting for orders to be matched, while creating stronger incentives to place orders and increasing the total liquidity available to users.

Another challenge facing fixed-rate markets is capital fragmentation, as each maturity, collateral type, and interest rate band can become a separate market. In Midnight, capital is not locked up during the intent phase; users can post cross-market quotes: “Since the same capital can be quoted to multiple markets simultaneously, the total liquidity a single market maker can offer users = available capital × number of markets.”

Since its launch in July 2026, Midnight Market has reached $3 million in active loans. While the numbers are still small, the team expects this to change soon, as it also inherits Morpho’s existing network effects and ecosystem. For example, Morpho Vaults currently hold over $4 billion in assets. Once the vault adapter is released, these funds will be able to begin quoting on Morpho Midnight and play a crucial role in building deep liquidity.

The most interesting problem Midnight solves is early exit. In older or less liquid term markets, borrowers and lenders often have limited options to exit before maturity. Midnight improves this by making positions interchangeable: lenders can sell credit units, and borrowers can buy debt units to reduce their outstanding debt.

Although Midnight can be understood as the underlying architecture for fixed-rate loans, an access layer has already been built on top of it: Tenor Finance. DeFi Frontier refers to Tenor as “Midnight’s HIP-3.”

Tenor inherently inherits all the functionality of the underlying layer, Morpho Midnight, and builds additional features on top of it:

Auto-renewal and fallback options: Tenor introduces automatic position renewal to prevent liquidation upon expiration. It uses independent keepers to roll over loans to new fixed-rate terms before expiration. If no fixed-rate match is found, the position can directly revert to the variable-rate pool on Morpho Blue.

On-chain OTC protocol: Tenor enables users to request quotes and broadcast customized over-the-counter offers, which can be shared with whitelisted counterparties for direct negotiation.

Organization Tools and Access Control: Tenor enables institutions to organize accounts with role-based permissions. Through these accounts, they can deploy customized, access-restricted credit markets that limit who can borrow or lend based on compliance or KYC requirements.

Tenor reduces maturity friction by adding auto-renewal and fallback options, allowing fixed-term positions to extend more smoothly as long as matching liquidity exists or fallback conditions are met. Building on this, its customizability makes it more institutional-friendly; over the long term, the team expects the platform to be used by users on one side being asset managers and on the other side being corporations.

Jupiter Offerbook, launched by Jupiter Exchange, entered public testing in June 2026, around the same time as the release of the Morpho Midnight whitepaper. Jupiter Lend, a floating-rate product introduced last year, marked Jupiter’s first foray into the lending space. Now, with Offerbook, they are entering the fixed-term market.

Offerbook is an intent-based lending protocol featuring no price-based liquidations and supporting fixed-term lending for long-tail assets.

Loan terms on the platform are short, typically ranging from 1 to 30 days. Upon maturity, if the borrower fails to repay, the lender directly claims the collateral without any liquidation occurring. This type of market enables the use of any long-tail collateral—whether NFTs, RWA, or any asset without active price discovery—as long as the lender is willing to underwrite it. It is a unique solution because it replaces continuous price-based liquidations with maturity-based collateral transfer, helping to create specialized markets that support assets otherwise difficult to finance.

On Offerbook, users can post intentions to lend or borrow, which are displayed in the app; liquidity is matched when an offer is accepted. Since users only commit when a match occurs, they are free to use their funds elsewhere until the matched order executes, solving the cold start problem. This creates opportunities for both lenders and borrowers to earn yields before finding a perfect term match.

Since its launch, Jupiter Offerbook has achieved $450,000 in active loans. Although their model is unique, it has been difficult to validate the market and generate demand due to scalability constraints tied to lenders’ willingness to directly underwrite collateral.

Kamino recently released its whitepaper on a fixed-rate lending protocol. Rather than building a standalone fixed-rate market, it has integrated fixed-rate reserve pools within Kamino Lend. This design offers distribution advantages: borrowers can see clear term structures, while lenders can quote specific rates and terms without fully exiting the floating-rate system, making fixed-rate lending an additive feature.

Each pool on the platform is defined by an interest rate and a term—for example, borrowing USDC at different rates and different terms. All these varying rates and terms form a grid.

Through this grid, Kamino enables borrowers and lenders to express their desired transaction positions along both price and time dimensions. Borrowers post borrowing intentions, specifying collateral, size, maximum interest rate, and term. Lenders post conditional liquidity, specifying the interest rate, term, and amount they are willing to provide. The grid becomes the execution layer: borrowers draw from predefined combinations of interest rates and terms to access available fixed-rate liquidity.

Lenders do not match directly but quote on a structured grid with predefined rates and terms (e.g., 4.5% for 1 month, 5% for 3 months), creating visible term structures and yield curves for different assets. By leveraging Kamino’s infrastructure, borrowers can post intent to wait for matching liquidity or directly draw from available fixed-rate liquidity on the grid. Based on this, if liquidity permits, Kamino can automatically roll loans over to the next term, similar to Tenor; if no fixed-rate liquidity is available, it can revert to floating rates. This addresses maturity issues and supports loan continuations, reducing the need for borrowers to manually manage each maturity.

To withdraw, lenders must go through the withdrawal queue. If a lender cannot withdraw immediately due to liquidity already being deployed, they will enter a first-in, first-out queue and be repaid when the loans in this reserve pool mature. This design ensures that the maximum waiting time for lenders is limited by the reserve pool’s term.

Although matching occurs between both parties, the funds do not sit idle—they continue to earn yield from a floating-rate reserve pool, helping to address the cold start problem.

Fixed rates do not eliminate any of the risks exposed by variable-rate borrowing over the years, but they make the cost of debt clear.

This is exactly what DeFi lending has been missing.

Floating rate pools are powerful because they make borrowing instantly available, but they compress everything into a single utilization curve. In contrast, fixed-rate markets allow borrowers to price for term, lenders to choose term and collateral risk, curators to allocate capital across terms, and applications to bundle more predictable credit products. We are also beginning to see initial iterations of predictable credit products, such as Aave’s Stable Vaults, launched in July.

This is important because DeFi lending is expanding. It now supports revolving credit, basis strategies, treasury management, RWA-linked assets, and consumer-facing applications. These users need more than just liquidity—they need clear and fixed financing terms.

We expect competition in this space to intensify, with more innovative solutions emerging to scale fixed-rate lending.

Current adoption is also relatively low, and floating-rate lending still dominates the market; however, the goal is to grow the overall pie, as these products can address many use cases that current DeFi lending cannot serve.

In addition, these products are designed to address the challenges faced by early protocols in the space, and they offer stronger distribution capabilities because the corresponding floating-rate lending products have matured. For example, funds in the floating-rate market can be quoted in the fixed-rate market while still earning yield and maintaining efficiency.

As these products mature, we should see a significant number of strategies that were previously impossible, along with a new flywheel in the lending space.