Every valuation starts with a boring question: what return can I earn before taking the specific risk of this asset?
That number is the risk-free rate. In traditional valuation it anchors the cost of equity, the cost of debt, and the discount rate used for cash flows. In DeFi, the same idea is useful, but the label is dangerous. A stablecoin lending APY, staking yield, or treasury-backed token yield may be a useful reference point. It is not automatically risk-free.
The job is to separate the base price of time from the risks that got bundled into the yield.
What A Risk-Free Rate Must Do
A risk-free rate has to match the cash flow being valued.
- Currency: dollar cash flows need a dollar base rate; peso cash flows need a peso base rate; ETH-denominated cash flows need an explicit choice between valuing in ETH terms or converting to dollars first.
- Inflation: nominal cash flows need a nominal discount rate. Real cash flows need a real discount rate.
- Horizon: a one-month bill does not anchor a ten-year valuation without adding reinvestment risk.
- Default: the issuer cannot add credit risk to the rate. If a sovereign bond has default risk, the default spread has to be stripped out before calling the remainder a risk-free rate.
This is why “use the government bond yield” is incomplete. The government, currency, maturity, and default assumptions all matter.
For a long-term dollar valuation, the usual benchmark is a long-term U.S. Treasury rate. For a long-term euro valuation, the cleaner benchmark is the lowest-risk euro sovereign rate, not the average rate across every country using the currency. For a local-currency bond issued by a government with default risk, the rate has to be adjusted down by the sovereign default spread before it can function as a base rate.
The DeFi Translation
DeFi does not remove the need for a base rate. It makes the decomposition more visible.
An on-chain yield can include:
- time value of money
- borrower demand
- token incentives
- liquidity risk
- smart contract risk
- oracle risk
- bridge risk
- stablecoin peg risk
- custody, issuer, and regulatory risk
- governance risk
Calling the whole number “risk-free” hides the most important part of the analysis. A lending market paying 6 percent is not saying the base rate is 6 percent. It may be saying the user is being paid a base rate plus compensation for liquidity, collateral quality, protocol risk, incentive sustainability, or temporary leverage demand.
The cleaner habit is to call it a base-rate proxy until the extra risks have been named.
Equity Risk Premium
The equity risk premium is the extra return investors demand for owning risky residual claims instead of the risk-free asset.
In the simplest cost of equity model:
The formula is less important than the discipline behind it. The risk-free rate pays for time in a currency. The equity risk premium pays for exposure to the risky market. Beta, or any other relative-risk measure, scales that market risk to the asset being valued.
The premium is not a law of nature. It changes when investors become more risk-averse, when macro uncertainty rises, when liquidity disappears, or when the market’s expected cash flows change.
Historical premiums are useful context, but they are noisy. The answer can change materially depending on the lookback period, the country, the averaging method, and whether the chosen market happened to be a long-term winner. A forward-looking premium is usually cleaner: start with today’s market price, estimate expected cash flows, solve for the discount rate that makes those cash flows equal the price, then subtract the risk-free rate.
Protocol Tokens Are Not Automatically Equity
The equity risk premium only applies cleanly when the token is actually exposed to residual economics.
A protocol token may look equity-like if holders have a credible claim on fees, buybacks, burns, distributions, or governance-controlled cash flows. It looks less like equity when the token only coordinates voting, subsidizes usage, or floats on narrative without a path to economic capture.
That distinction matters for valuation. If the token is a residual claim, the discount rate needs a market risk premium plus protocol-specific risk adjustments. If it is not a residual claim, forcing it into an equity model can turn the valuation into decorative math.
A Practical DeFi Discount Rate
For DeFi work, I prefer writing the hurdle rate in plain language before turning it into a number:
Then make each term earn its place.
- Define the cash flow owner: equity holder, token holder, lender, LP, validator, or vault depositor.
- Choose the numeraire: USD, stablecoin units, ETH, BTC, or another token.
- Match the horizon: overnight liquidity, fixed maturity, perpetual protocol cash flow, or uncertain exit.
- Pick the cleanest base-rate proxy available.
- Identify the risks embedded in the observed on-chain yield.
- Add risk premiums only for risks not already reflected in the cash-flow forecast.
- Stress the result when incentives disappear, liquidity leaves, governance changes parameters, or the peg breaks.
The main mistake is double counting. If protocol cash flows are already haircut for expected exploits, bad debt, or liquidity exits, do not also add the full same risk again in the discount rate without saying why.
What This Means For DeFi Lab
The public DeFi question is not “what APY is highest?” It is “what risk is this APY paying me to hold?”
Risk-free rates and equity risk premiums are useful because they force that question into the open. The base rate says what time is worth. The premium says what uncertainty is worth. A serious DeFi product has to explain both before its yield, token, or vault can be valued with any discipline.