Skip to content
Financial Context Wiki

DeFi Yield Risk Scorecard

A transparent framework for separating cash yield, token emissions, protocol risk, liquidity, leverage and exit costs before allocating capital to DeFi.

Updated

For educational purposes only; not investment advice. DeFi positions can lose principal, become illiquid, or incur losses beyond displayed yield.

Direct answer

Score a DeFi position from evidence, not from its headline APY. Identify the exact chain, pool or market, token contracts, implementation version, oracle, collateral rules and withdrawal path. Then convert every return into a common horizon and denomination, subtract all costs, stress adverse prices and liquidity, and record what remains as executable net cash flow. A score is a decision aid, not a probability of loss or a guarantee of recovery.

How it works

  1. Pin the product, chain, contract addresses, proxy implementation, deployment block, deposit asset, reward asset, lock, withdrawal cap and recipient path.
  2. Decompose yield into borrower interest, trading fees, staking or validator income, MEV, and token emissions. Keep token price appreciation separate from cash yield.
  3. Normalize APR, APY, compounding interval, token decimals and time horizon. Subtract gas, protocol/performance/withdrawal fees, borrow interest, slippage and taxes in the reporting denomination.
  4. Stress price, depeg, utilization kink, oracle stale/deviation, collateral LTV/LT, health factor, liquidation penalty, bad debt and LP impermanent loss.
  5. Inspect implementation, reentrancy and token behavior, proxy/admin/upgrade/pause roles, oracle rounds and sequencer feeds, bridges, custody and nested aggregators.
  6. Test the executable entry and exit with the intended size: quote depth, price impact, queue or lock, withdrawal cap, gas, settlement delay and failure or pause behavior.
  7. Publish factor scores, weights, confidence and missing-data penalties; apply correlation stress and define position limits, monitoring triggers and an exit plan.

Examples

  • A $10,000 position at 8% simple APR for 90 days on a 365-day basis earns 10000*0.08*90/365 = $197.2603 gross. After $20 of fees, net cash yield is $177.2603; this excludes token-price and principal risk.
  • An 80% annual reward rate compounded monthly implies (1+0.80/12)^12-1 = 116.9425212972% APY, while daily compounding implies 122.3593468213%. A displayed APR is not an APY or a promise of reinvestment.
  • A reward balance of 8,000 tokens falling from $1 to $0.30 is worth $2,400 instead of $8,000 before fees. Token quantity is unchanged, but market value lost 70%.
  • In a 50/50 LP, if token A moves from $100 to $70, the price ratio is r=0.70; the constant-product LP value relative to holding is 2*sqrt(r)/(1+r)=0.9843059136, about -1.5694086430% impermanent loss before fees and emissions.

Risks

  • Yield-source misclassification: interest, fees, emissions and price gains are not interchangeable cash flows.
  • Emission cliff or governance change removes the reward subsidy.
  • Reward-token price drawdown destroys the displayed return.
  • APR, APY and compounding assumptions do not match the product.
  • Raw units, decimals or reward denomination are decoded incorrectly.
  • Gas, protocol, performance, withdrawal and tax costs are omitted.
  • Borrow rates or funding costs jump with utilization.
  • A utilization kink creates a liquidity crunch or withdrawal queue.
  • LP impermanent loss exceeds collected fees.
  • Oracle staleness, deviation, decimals or sequencer downtime misprices collateral.
  • LTV, liquidation threshold, health factor, penalty or bad debt rules are misunderstood.
  • A contract bug, reentrancy path or non-standard token corrupts accounting.
  • A proxy implementation, admin, pause or emergency role changes behavior.
  • Bridge, custodian, underlying issuer or nested aggregator fails.
  • Withdrawal locks, caps, queues or pauses block the planned exit.
  • Thin depth, slippage, MEV or failed transactions make the quote non-executable.
  • Depeg and correlated collateral losses defeat diversification.
  • Score weights, missing-data treatment or stale observations create false precision.
  • Monitoring, alerting or fallback execution fails during stress.
  • Tax, accounting, jurisdiction and denomination choices hide the realized loss.

Common misconceptions

  • The highest APY is the highest return. It may be mostly emissions, leverage or uncompensated tail risk.
  • A stablecoin pool is risk-free. Depeg, blacklist, oracle, contract, liquidity and issuer risks remain.
  • An audit removes smart-contract and economic risk. It is a bounded review of a version and scope.
  • APY is cash yield. It may be a token-denominated, hypothetical reinvestment rate before costs.
  • Diversifying across protocols creates independent risk. Shared bridges, oracles, custodians and collateral can correlate losses.

Sources

Navigation

Search the wiki...