Decentralized finance has a marketing problem, and it is not the one most people think. The problem is not that on-chain yield is too low or too complicated. It is that it is too often presented as a number on a screen, detached from the question every lender has asked for centuries: who is paying me, and why?
At Caldrion, we believe on-chain yield is credit, and that it should be underwritten as credit. That belief shapes how we approach decentralized finance in our digital assets portfolio, and it reflects the habits we bring from lending in private markets, where every loan sits on our own balance sheet. This note explains how we think about yield in decentralized finance, what we look for, what we avoid and why we believe a lender’s discipline is the right lens for this part of the market.
Yield is payment for a service
We start from a simple premise. Sustainable yield is payment for something useful. In decentralized finance, we see three services that genuinely earn it.
Providing liquidity. Automated market makers (AMMs) let anyone exchange one asset for another against a pool of assets supplied by liquidity providers. Traders pay fees for that convenience, and the fees go to the providers. The service is real: without liquidity, there is no market.
Lending. Lending protocols match borrowers and lenders through software, with loans secured by collateral that the protocol can liquidate if its value falls. Borrowers pay interest because they want leverage, liquidity or a short position. The lender is paid for providing capital.
Securing a network. Many blockchains are secured by validators who stake assets and can lose part of them for misbehavior. Stakers earn rewards for providing that security, and liquid staking lets them do so while keeping a tradable claim on their stake.
Each of these services has a clear payer and a clear reason to pay. We believe that is the first test any on-chain yield must pass. If we cannot trace a return to a real source, fees, interest or staking rewards, we assume it is compensation for a risk we have not identified, or a subsidy that will end.
If we cannot explain where a yield comes from, we do not take it.
The subsidy trap
Much of the yield advertised in decentralized finance has historically come from incentives: new tokens issued by a protocol to attract liquidity or users. Incentives can be a legitimate way to bootstrap a network, much as a new business might discount its products to win customers. But they are not income in the way fees and interest are. They are paid in the protocol’s own token, often from a supply schedule that dilutes existing holders, and they tend to end or shrink once the protocol has what it needs.
We believe the market still too often confuses the two. A pool that pays a high headline yield may be paying mostly in incentive tokens whose value depends on continued demand for the protocol. When incentives fall, liquidity leaves, the token falls and the yield turns out to have been a temporary loan from the protocol’s future.
So when we look at any yield opportunity, we separate it into its parts: the portion paid by real users for a real service, and the portion paid in incentives. We underwrite the first as income. We treat the second, if we accept it at all, as a speculative position in the incentive token, sized and judged on its own merits.
Underwriting a lending protocol
Because Caldrion also lends in private markets, we approach on-chain lending with familiar questions, translated into a new setting.
What secures the loan? On-chain lending is usually over-collateralized: borrowers must post collateral worth more than they borrow. We study which assets are accepted as collateral, how their values are set and how quickly they can be sold in a falling market. Over-collateralization is only as good as the liquidity of the collateral when it matters.
How does enforcement work? In private credit, enforcement means legal rights and the ability to take control of collateral. On-chain, it means liquidation mechanics: the rules that let third parties repay a borrower’s debt and take collateral at a discount when a loan becomes undercollateralized. We study how those mechanics behaved during past periods of stress, whether liquidators were able to act in time and what happened to lenders when they were not.
Where do prices come from? Lending protocols rely on oracles to know the value of collateral. An oracle that reports a wrong price, even briefly, can trigger liquidations that should not happen or prevent ones that should. We look at oracle design, sources, update frequency and fallback arrangements.
Who can change the rules? Many protocols can be upgraded, and some have administrative keys that allow changes without holder approval. We want to know who holds those powers, what they can change and how quickly. A lender should know whether the terms of its loan can be rewritten.
What does the code say? We require audited smart contracts and read the audits critically: their scope, their findings and what was fixed. We also look at how long the code has run in public with meaningful value at stake, which we treat as an additional form of testing.
These are the same questions any careful lender asks: collateral, enforcement, valuation, documentation and counterparty. The answers are written in code rather than contracts, but the discipline is the same.
Underwriting liquidity provision
Providing liquidity to an automated market maker earns trading fees, but it also exposes the provider to the price movements of the assets in the pool. When the relative prices of pooled assets diverge, a liquidity provider can end up worse off than if it had simply held the assets, a cost the market calls impermanent loss. In volatile pairs, that cost can exceed the fees earned.
We believe this is one of the most misunderstood risks in decentralized finance, and it is why we focus on pools where we can limit directional price risk. Liquidity in pools of stable digital dollars, or of assets designed to track one another closely, earns fees with far less exposure to divergence. We still underwrite the assets in each pool: a stable digital dollar is only as stable as its reserves and redemption mechanics, and assets designed to track one another can separate under stress.
We also consider how fees are generated. Pools with steady, organic trading volume from real users are more attractive to us than pools whose volume depends on incentives or on a small number of large traders.
Underwriting staking
Staking is often described as the closest thing decentralized finance has to a risk-free rate. We believe that description is misleading. Staking rewards are payment for securing a network, and they carry real risks: validators can be penalized for errors or misbehavior, staked assets may be locked for a period before they can be withdrawn and liquid staking introduces its own smart contract and market risks.
We underwrite staking by asking who operates the validators, how they are monitored, what penalties apply and how quickly we could exit. For liquid staking, we also look at how closely the liquid token has tracked its underlying stake under stress, and at the depth of the markets in which it trades. And we remember that staking rewards are paid in the network’s own token: they are income in units of an asset whose price can move a great deal.
Diversification and limits
No amount of underwriting removes the possibility that a protocol fails. Code can contain flaws that audits miss; oracles can be manipulated; governance can make poor decisions. We plan for that.
We diversify across protocols, so that no single failure can impair the portfolio. We prefer over-collateralized positions. We use insurance coverage against protocol failure where it is available, while recognizing that coverage has limits of its own. We set position limits in advance and review them as protocols change. And we keep the yield engine in its proper place in our barbell: a source of income with low correlation to public markets, not a place to reach for returns.
We believe decentralized finance will reward lenders who behave like lenders: patient about returns, precise about collateral and humble about what they do not know.
Why the opportunity persists
If on-chain yield is credit, why does it pay? We believe the answer lies in the structure of the market. Decentralized finance remains young and fragmented. Many participants cannot hold digital assets, cannot use on-chain protocols or face constraints that prevent them from providing liquidity when it is most needed. Demand for leverage and liquidity tends to rise sharply during rallies and stress alike. In those conditions, a patient, well-secured provider of capital can be paid for providing something scarce.
We also believe the opportunity is changing. As institutional participation grows and regulation becomes clearer, we expect yields on the simplest strategies to compress over time, and we expect more of the opportunity to sit in work that requires real underwriting: understanding a protocol’s mechanics, assessing collateral and managing counterparty and code risk. That is the kind of work we are built to do.
How this shapes our approach
Our decentralized finance strategy has a clear role: market-neutral income, earned by acting as a liquidity provider to the decentralized economy while limiting directional price risk. We work in automated market makers, lending protocols and liquid staking. We use audited smart contracts, prefer over-collateralized positions and diversify across protocols. We take only yield we can trace to a real source.
Protocol teams building lending markets, liquidity venues or staking infrastructure that meets our criteria are welcome to share their work with us. We aim to give initial feedback within 72 hours of receiving a complete package.
This article reflects Caldrion’s views as of November 2026. It is not an offer of securities or investment advice.