Onchain credit has crossed a meaningful threshold. DeFi loan volume now sits in the tens of billions of dollars across lending protocols, and onchain borrowing has overtaken centralized crypto-backed lending in market share. Tokenized institutional credit funds, onchain origination, peer-to-peer lending, and tokenized offchain credit are each finding product-market fit in their own ways. Capital is flowing in. Curators are underwriting risk. Money markets are integrating these assets as collateral. The asset class is real.

But growth from here gets harder without tools that price and transfer risk. Onchain credit's next phase isn't about issuing more credit, it's about making the credit that already exists more usable. That requires looking at the full stack:

Credit Issuance — where credit is originated and tokenized (Maple, Apollo, Figure, Cap)
Capital Allocation — where liquidity flows in (Sky and its agents, vault curators, yield-bearing products)
Infrastructure — the plumbing that makes credit composable (money markets, tokenization platforms, oracles, vault systems)
Risk Management — the layer that prices and transfers risk (liquidity protocols, tranching, risk analysis, reserve verification, coverage) The Risk Management layer is where onchain credit's growth gets unlocked or stalls. And within that layer, one specific tool is conspicuously underbuilt: interest rate hedging.

The Variable-Rate Problem

Almost every yield in DeFi is variable. Aave borrow rates, Morpho market rates, Ethena staking yields, LST yields, Pendle yield tokens — all of them float. Total DeFi TVL on Ethereum sits north of $70B, and the bulk of it generates variable yield through lending or staking. Variable rates are fine for short-duration trading and farming, but they are a structural blocker for the kinds of credit that matter at scale: long-duration, fixed-rate loans tied to real economic activity. Real estate financing, business credit, large purchases, treasury operations — these are decades-old categories of finance built almost entirely on fixed rates. They don't exist meaningfully onchain.

The reason is simple, and it isn't the protocols' fault. Lenders won't extend fixed-rate credit when they have no way to hedge what happens if rates move against them. Imagine offering a 10% fixed-rate two-year loan via Morpho V2 while variable market rates trade around 7%. If rates rise to 11% a year later, the loan still pays 10% but new origination prices higher, and the mark-to-market value of the fixed position drops. Without a hedging tool, the lender is naked long duration. Most rational lenders just won't do it.

This is the gap. It's why fixed-rate, long-duration credit — the largest category of credit in the world — has essentially no onchain presence.

What Traditional Finance Got Right

Interest rate swaps aren't a new invention. The traditional IRS market clears more than $460T in notional volume annually (per ISDA) and underpins a global debt market north of $250T (per the IMF), the majority of which is fixed-rate. The mechanism is straightforward: two parties agree to exchange fixed and floating rate payments on a notional principal over a defined period. Lenders use swaps to neutralize duration risk on fixed-rate books. Borrowers use them to lock in costs. Treasurers use them to manage cashflow. The product is so embedded in TradFi that fixed-rate credit at scale would be unimaginable without it.

None of this exists onchain in any usable form. Traditional swaps don't reference onchain rates like Aave or Morpho, and they don't accept digital assets as collateral. Pendle's Boros gestures at the primitive but is narrowly scoped to perpetual funding rates over short durations — useful for traders, not for a lender trying to hedge a two-year loan book.

So onchain lenders sit in a state where the asset class wants to grow into longer durations and fixed rates, but the hedging infrastructure that would make that possible isn't built yet.

Where Risk Management Sits Today

The Risk Management layer has a handful of distinct sub-categories worth naming, because each addresses a different kind of risk transfer:

Liquidity protocols (3F, Multiliquid, Fission, Keyring) address the inherent illiquidity of credit assets, providing exit paths during stress and enabling looping.
Tranching protocols (Cork, Strata, Royco, Pendle) split yield-bearing assets into senior and junior tranches, letting the spread itself price the underlying risk.
Risk analysis (Chaos Labs, Cicada, Credora, Block Analitica) provides the underwriting frameworks curators and money markets rely on.
Reserve verification (Accountable, LlamaRisk, vlayer) provides cryptographic and independent proof of what's actually backing onchain credit assets.
Risk coverage (Nexus Mutual, OpenCover, Catalysis, Symbiotic, Eigen) offers protection against tail-risk events.
Asset managers (M11, Five Sigma, Anemoy) bridge onchain capital and offchain credit operations. What's missing from this list is a primitive specifically for transferring interest rate risk. Tranching gets close — it prices and shifts certain risks — but it doesn't cleanly separate duration risk from credit risk, and it doesn't give a lender a precise way to neutralize the impact of rates moving against a fixed-rate book. That requires a swap.

Interest Rate Swaps as Onchain Infrastructure

A working onchain IRS protocol needs a few properties to actually serve credit markets:

Permissionless market creation. A single team picking which rates can be hedged and on what terms recreates the gatekeeping problem that DeFi exists to solve. Markets need to be creatable by anyone, with immutable risk parameters, so the market itself decides which references and structures are useful.

Any reference rate, any duration. Onchain credit isn't one rate. It's Aave borrow rates, Morpho market-specific rates, Ethena staking yields, LST yields, Pendle PT/YT yields, tokenized fund distributions, and on. Each warrants its own market. Durations need to range from hours (for traders) to years (for credit hedgers).

Pooled liquidity via AMMs, with one-way markets. Order books work for liquid, high-frequency markets. For the long tail of rates and durations, AMM-based liquidity makes more sense. Markets should be one-way (sell fixed, receive floating) so that curators can actively manage their own directional rate exposure while still serving demand from the buy side. Liquidity flows in through curator-managed vaults — sophisticated allocators earn fees for active risk management, and passive depositors earn yield on the same assets they already hold. Buyers get aggregated liquidity and immediate execution without waiting for a counterparty.

Oracle-anchored settlement. Pricing swaps off implied rates from market trades works for short-dated, liquid markets but breaks down for longer durations where trades are sparser and manipulation is easier. Long-duration swaps benefit from oracle-anchored base rates that creators can opt into.

Onchain collateral. USDC, LSTs, PTs, vault receipt tokens — the same assets participants already hold need to work as margin. Anything else fragments the user experience.

These aren't theoretical requirements. They map directly to what credit markets need to mature, and they're the design choices we've made building @KairosSwap.

Critically, an IRS protocol designed this way doesn't compete with money markets like Morpho, Aave, or Kamino — it complements them. Lenders extending fixed-rate loans on those platforms get a hedge they previously couldn't access. Curators allocating into onchain credit collateral get a way to convert variable yield to fixed. The credit infrastructure that already exists becomes more usable, not displaced.

What This Unlocks

The point of an IRS primitive isn't the swaps themselves — it's the credit products they make possible.

Lenders can offer fixed rates. A curator extending a 10% fixed-rate two-year loan via Morpho V2 enters an offsetting swap that pays ~8% fixed and receives the floating Morpho market rate. If rates rise to 11%, the loan still earns 10% and the swap nets a 3% profit. If rates fall, the swap loses but the above-market 10% loan is worth more. The lender's interest rate exposure is neutralized, the borrower gets what they actually wanted (a fixed rate), and the protocol unlocks a category of credit it couldn't previously offer.

Digital Asset Treasuries can issue fixed-yield bonds. A DAT staking ETH can hedge the variable staking rate through a Kairos-style market and lock in fixed yield over multiple years. That fixed cashflow is bondable. Bonds raise capital. Capital buys more ETH. This is how TradFi treasury operations work, and it's the missing piece for DAT companies that want to operate as actual yield-generating businesses rather than passive holders.

Yield traders get capital-efficient leverage. Instead of looping through Aave, a Pendle user can post a YT or PT as collateral in a market referencing the same underlying rate and gain levered exposure through a swap. Margin and a small swap fee replace a full borrow rate. More capital efficient than the loop, and it unlocks size on positions that looping can't reach.

Vault curators can offer fixed-rate products. A vault generating variable yield can use swaps to convert that yield into a fixed payment to depositors. Predictability becomes a product feature instead of an aspiration.

Each of these use cases extends the surface area of onchain credit. None of them require new lending protocols, new oracles, or new chains. They require a hedging primitive sitting in the risk management layer, alongside tranching and risk coverage and liquidity protocols.

Together they form a flywheel: lenders gain confidence to originate fixed-rate, long-duration credit; borrowers get payment predictability; curators deploy more capital into onchain credit collateral; and the asset class moves closer to behaving like a real fixed-income market.

The Composability Argument

The most underrated property of onchain interest rate swaps isn't any single use case — it's that they become a building block. Once a protocol exposes a clean API for pricing and executing swaps, every other team in the stack can compose with it. Money markets can let lenders hedge at origination. Vault platforms can convert variable yield to fixed. DAT companies can embed it in their treasury operations. DeFi terminals can offer levered yield positions natively.

This is the dynamic that made tokenization platforms valuable: not the tokens themselves, but the composability they unlocked across DeFi. A risk management primitive is more valuable when it's connective tissue across the rest of the stack than when it's a destination product.

What Comes Next

This isn't theoretical. @KairosSwap has been live in capped-TVL beta on Base since late October, with over $300M in notional swap volume to date. The demand for onchain rate hedging is real, and the design choices above hold up in production.

The onchain credit ecosystem has spent the last two years proving credit can be issued, allocated, and serviced onchain. That work is still ongoing, but the harder problem is now in view: making this asset class behave like a real fixed-income market, with the risk management infrastructure to support fixed-rate credit at scale.

Liquidity protocols, tranching, risk analysis, reserve verification, and coverage all play roles. So does interest rate hedging. The lenders who underwrite the next $100B of onchain credit will need it. The DATs that issue the next generation of onchain bonds will need it. The traders building levered yield strategies want it now.

Onchain credit doesn't need more origination to grow. It needs better tools to manage the risk of what's already there.