Constant Finance says fixed-rate lending is gaining speed again in DeFi, but locking a rate does not automatically mean borrower and lender demand is actually matched. In a long-form post titled How do borrower and lender needs line up once the rate is fixed?, Stephen argues that the missing piece is time: how long capital is truly needed, and what happens when that real usage period is shorter than the contract maturity.
The article frames lending as one of DeFi’s core financial primitives. By writing collateral, borrowing and liquidation rules into public smart contracts, protocols make capital flows, debt positions and risk exposure verifiable on-chain while allowing the same assets to be combined across different protocols. After years in which floating-rate markets drove on-chain credit expansion, the piece says the market is now asking a more specific question: can a borrower know, before opening a position, how much interest the loan will actually cost?
Fixed-rate lending is back in focus
According to the article, fixed-rate lending has become a fresh area of DeFi construction in 2026. The idea is not new. Stephen points to Yield Protocol and Notional Finance, both of which brought fixed-rate, fixed-term structures on-chain in 2020. The article also cites Term Labs being selected for a16z CSX in 2024 as a sign that early-stage interest in the structure had not faded. More recently, Morpho launched Midnight as a fixed-rate protocol alongside its floating-rate market Blue.
Those efforts respond to a basic borrower need, the article says: users want to know their funding cost before they put capital to work. That matters most in staking, Launchpool participation, LP farming and arbitrage, where the return side is already uncertain. If the cost side is moving too, many trades never get opened.
Still, Stephen argues that a fixed rate does not mean demand has been matched. In his view, fragmentation comes less from user diversity itself and more from a market structure that requires every need to be met by a precise mirror counterparty. In a market that does not allow early repayment, every borrower must find someone willing to take the exact opposite side with the same amount and the same term. Real markets rarely look like that.
Floating-rate markets made funding accessible, but not predictable
The article says on-chain borrowing has long been handled mainly by floating-rate protocols such as Maker, Aave and Venus. Those markets made financing more widely available, but they did not always let borrowers lock in the cost at the moment a position was opened. Borrowing rates on Aave and Venus move with supply and demand, while Maker’s stability fee can change through governance.
Stephen uses BNB Launchpool as an example. The most direct way to join such an event is to buy BNB and participate. But once an announcement is out, BNB often rises quickly; when the event approaches its end, the price may fall back. In that case, even if a user receives new tokens, the spread between buying and selling spot BNB may wipe out the gain.
The second route is to post other assets as collateral, borrow BNB for the event, then repay it after the campaign ends. The article says Venus has served this type of demand on-chain. The problem is that the rate visible when the borrower opens the position is not necessarily the rate paid during the event. A borrower may start at 2% annualized, only to see the rate jump to 6% or even 7% once many users pile in. By the end, token rewards and interest expense can cancel each other out.
That, the article says, is exactly why fixed-rate lending is attractive: returns may still move, but costs at least become calculable. Even so, calculable cost does not mean the actual funding window is matched.
The rate may be fixed, but the term can still be wrong
Most fixed-rate lending markets today still organize liquidity around maturity dates, according to the article. If a borrower enters one of those maturity markets and holds the position to expiry, the rate is indeed fixed. The issue is that real-world borrowing needs are not distributed by maturity bucket. A Launchpool campaign may require only seven days. A staking strategy may last 14 days. An arbitrage window may be open for just a few days.
The article gives a simple example: a borrower needs funds for seven days but the available fixed-rate market matures in 30 days. By day seven, the trade is over and the position has been unwound, so the capital could in principle be returned. But the loan contract still has 23 days left. At that point the real problem is no longer fixed versus floating. It is a term mismatch. The borrower either accepts a known but potentially too-long funding cost, or exits early and meets the market again at the exit price.
Stephen argues that early exit costs cannot be fully locked in on day one. In one type of fixed-maturity debt design, a borrower may repay the full face value of the obligation before maturity. The amount is defined, but the contract does not automatically subtract financing cost for the unused 23 days. Worse, if no one in the market is willing to take over that debt directly, a borrower who used funds for only seven days may still need to pay for all 30 days to get out. In that setup, the rate looks fixed on paper, but the borrower’s real cost extends beyond the actual usage period.
The article describes another route as well: buying the relevant debt position in a secondary market to offset the original liability. In that case, the buyback cost usually falls if market rates rise, and may increase if market rates fall. The final execution price also depends on quotes, depth and liquidity. The key point, Stephen writes, is that under an early exit, the borrower’s realized rate is not determined by the signed APY on day one. It is determined by the market APY at the time of exit.
So fixed-rate lending has not failed, the article says. But for a user who exits on day seven, the final cost still cannot be pinned down precisely on day one. For someone who needs seven days of funding, the market is offering a longer-dated contract, not a financing product tailored to the exact usage window.
Fragmentation remains a structural issue
If fixed-rate lending continues to revolve around maturity dates, the article says, it must face a structural problem: once actual usage time and market maturity diverge, a fixed rate does not guarantee a fixed final cost. Early repayment or a secondary-market exit can bring rate risk back onto the borrower’s balance.
Stephen describes that as fragmentation. Liquidity is split across maturities, rates, collateral types and protocols, while borrowing demand is distributed according to actual usage periods. The two do not naturally line up one-for-one.
The piece argues that current industry responses have not escaped the framework of precise matching. They mostly replicate it in more places. A curator model adds a layer on the demand side, absorbs fragmentation and redistributes it to lenders. Multi-market quote pooling aggregates pricing on the supply side. But in the article’s telling, these approaches solve where liquidity sits, not whether the requirement for exact matching should exist in the first place. Borrowers still have to fit into a specific maturity, a specific rate, a specific collateral mix and a specific protocol market. More channels for matching do not remove the need to match exactly. The article’s conclusion is blunt: that does not solve fragmentation; it makes distribution even more fragmented.
Put differently, fixed-rate lending has delivered more certainty on rates than on actual funding time. Borrowers care less about what rate exists at a given maturity and more about a simpler question: if I need capital for only seven days, what will I end up paying? If the market cannot answer that in advance, fixed-rate lending remains a partial solution.
The article says the real test is whether a borrower whose actual usage period is shorter than the contract maturity can still pay the original rate only for days actually used, and whether the lender’s capital can then move into the next suitable match without friction. If not, fragmentation is still there.
Fixed rates and early repayment are not the same thing
Stephen also lays out a second axis of the problem: whether funding cost itself is knowable. Under floating rates, borrowing cost is an unknown variable. For strategies that depend on spread, the return leg may be modeled, but if the cost leg is open-ended, the model is not enough to make a decision. The result is either no borrowing at all or a very wide safety margin, leaving only high-return trades able to clear.
The article sums it up this way: fixed rates turn the unknown into the known; early repayment turns the known into a ceiling.
It then makes a distinction. Fixed rate and early repayment are two separate steps and should not be conflated. A fixed-rate product without the ability to repay early merely swaps rate risk for lock-in risk. The asymmetry remains; it just points in a different direction. Add early repayment, and the borrower gets a different profile: if rates rise, that does not matter; if rates fall, the borrower can refinance more cheaply; the funding cost can only end up lower than, or equal to, what was implied at signing.
That right is not free, the article says. Early repayment is effectively an option sold by the lender to the borrower. Options have a price, and that price is embedded in the fixed rate. For that reason, a fixed-rate loan with early repayment rights should, in theory, price above an otherwise identical fixed-rate loan without them. A higher borrowing rate in that context should not automatically be read as worse pricing, Stephen writes. The product includes an additional right, and that right turns funding cost from a random variable into a known ceiling at the time of signing.
Constant Finance’s design: pay for the days actually used
Constant Finance says it is trying to address that missing layer at the mechanism level. Under its design, the borrower and lender agree on both a fixed rate and early repayment terms when the loan is matched. The borrower can repay principal before maturity within the allowed window, pay interest at the original rate for the actual usage period, and stop the interest clock at that point.
For a loan with unchanged principal and simple-interest calculation, the article gives the formula as:
Interest = loan principal × agreed annual rate × actual days used ÷ agreed day-count basis.
If the funds were used for seven days, then interest is calculated over seven days. The borrower does not have to wait for maturity, search a secondary market for an acceptable exit price or absorb financing cost for the remaining 23 days.
The article argues that this creates a layer of flexibility between contractual maturity and actual capital usage. Even if the contract ends later, it can still serve a shorter borrowing need as long as the lender accepts the early repayment terms. For borrowers, that may bring back opportunities that were previously rejected because the term did not fit or the exit cost was impossible to know in advance.
What happens to lender capital after early repayment
The article spends equal time on the lender side. Once borrowers gain flexibility, lender needs also have to be addressed. What lenders really care about, Stephen writes, is whether capital can keep finding borrowing demand that fits their conditions. One loan ending early does not necessarily mean the entire lending plan should end with it.
According to Constant Finance’s design, if the original lending order is still active and the market contains enough executable borrowing orders that match the lender’s conditions, repaid capital is immediately and automatically rematched. Interest then accrues based on the conditions of the next matched loan. The lender’s quoted APY acts as a floor rather than a market order, and after early repayment the capital returns to the order book instead of remaining idle.
That means funds originally intended for a longer lending horizon can first serve a seven-day borrower, then move to the next borrower after repayment. In the article’s framework, short-term usage and longer lending plans can be connected through consecutive matching.
There is a limit. Automatic rematching still depends on suitable demand being available. The piece notes that after a broader market rate decline, there may be no sufficient demand at the original terms, and some funds may still wait. Lenders therefore have to account for both early repayment and redeployment when they set quotes. The minimum-rate constraint governs the conditions of the next loan, not a guaranteed return during any waiting period.
Three types of demand the article says this could unlock
The article closes by naming three categories of borrowers or strategies it believes could benefit.
The first is beta strategies: staking, arbitrage and liquidity mining. In Stephen’s framing, traders pursuing alpha may not care much about a 1% to 2% rate move, but beta strategies can live or die on that same 1% to 2%. A seven-day staking event may have no fixed-rate product with a matching maturity, while a floating-rate position could reverse against the user mid-trade. Rational users may simply choose not to borrow. If funding cost moves from a random variable to a ceiling known at signing, those trades can become executable.
The second is certainty at the decision point. With early repayment and refinance available, the borrowing decision can narrow to one test: if the trade is profitable when signed, take it. If rates rise later, the borrower is unaffected. If rates fall, refinancing can reduce the cost.
The third is longer-duration borrowers, such as users borrowing stablecoins against BTC collateral. If the rate ceiling is locked, the funding cost is capped. The article says that even if a borrower expects the collateral to rise 10% this year, they can still borrow now because the financing cost is already known.
Constant Finance’s pitch: make loans happen that otherwise would not
The article ends with a broader claim. A market can have capital available to lend and people willing to borrow, yet still fail to clear a trade acceptable to both sides. Often the reason is concrete rather than abstract: an opportunity lasts only a few days, while financing forces the borrower to carry costs for far longer; or the headline rate looks acceptable, but the price of exiting early will not be known until later. Demand like that is often abandoned before an order is ever placed.
Stephen argues that if a fixed-rate system still requires mirror matching inside a maturity market, it has not solved fragmentation; it has only redistributed it. Real progress, in the article’s view, does not come from adding more matching channels. It comes from making mismatched demand executable.
Constant Finance says it is trying to connect rates, actual funding time, early repayment, refinance and capital redeployment so that more capital can meet real use cases and some loans that would otherwise never happen can take place.
The article says Constant Finance’s beta test is now live. It adds that readers who want more detail on its fixed-rate lending and early repayment mechanism can visit the project’s official website for product materials and a litepaper, and can also follow its official Twitter (X) account for beta progress, product updates and project news.
The post also includes a disclaimer saying markets carry risk, investors should be cautious, and the article does not constitute investment advice.
