The 4 Breakdown Axes
The risk breakdown explains why a pool received its score.
Each axis shows:
- a score out of 100;
- its weight in the final score;
- readable notes;
- sometimes a penalty or bonus.
1. Protocol
The Protocol axis measures protocol strength.
Signals include:
- number of public audits;
- protocol age;
- protocol TVL;
- hack or incident history;
- open-source status when available;
firstSeenfallback when DefiLlama does not provide a reliable date.
Example notes:
Audits: 2 audits;Age: established;Protocol TVL: very high TVL;History: 1 past hack/exploit;Open-source: yes.
2. Asset
The Asset axis measures token quality and pool exposure.
Signals include:
- tiers of underlying tokens;
- stablecoin presence;
- blue-chip or exotic assets;
- single-asset or multi-asset exposure;
- impermanent loss risk;
- pool size.
A single-asset pool usually has a better IL profile than a volatile LP. A stable-stable or same-peg LP is penalized less than a volatile-volatile LP.
Examples:
Tokens: USDC=A+, WMON=A;Exposure: single-asset (no IL);Pair: stable-volatile;Pool size: thin pool TVL <$50k.
3. Yield
The Yield axis measures yield quality.
The model distinguishes:
- base yield;
- native yield from the underlying asset;
- rewards or incentives;
- extreme APY;
- APY marked as an outlier.
Yield that mostly comes from real revenue or native staking scores better than yield mostly driven by temporary incentives.
Examples:
Yield: real revenue / staking;Yield: mostly token incentives;APY > 200%;DefiLlama flagged APY as outlier.
4. Liquidity
The Liquidity axis measures how easily a user can exit the position.
Signals include:
- protocol family;
- product type;
- instant withdrawal or withdrawal queue;
- bridge or cross-chain dependency;
- TVL depth.
Examples:
AMM: exit limited by slippage;lending: instant withdraw, pause-able;LST with unbonding queue;cross-chain: bridge dependency;Pool depth: exit slippage risk.
Manual Overrides
Monad APR can apply a manual override when automatic data is not enough, for example after a known exploit.
An override can:
- add a visible reason;
- cap the score;
- show a red callout in the breakdown.
Overrides are intentionally conservative.