Risk Grades A, B, C: How Apyee Scores Stablecoin Lending Pools
Risk grade is not a marketing badge. Here is the exact five-factor rubric Apyee uses — audits, operating history, TVL, oracles, governance — and how the grade actually changes your allocation.
A risk grade is supposed to tell you, in one letter, whether a yield is worth the volatility. Most DeFi dashboards either skip the grade entirely or hand-wave one from TVL. Apyee uses a five-factor rubric, weighted, that we publish openly. Here is what the grades mean.
The five factors
Every strategy in the Apyee universe is scored on:
- Audit depth — number of distinct top-tier audits, their recency, and the firms involved (Trail of Bits, OpenZeppelin, ABDK, SigmaPrime, Spearbit, Cantina, PeckShield)
- Operating history — months running on mainnet without a core-logic exploit. Front-end and oracle bugs are noted separately; only protocol-level drains count here
- TVL scale and stickiness — current TVL and the 90-day stability of that TVL. A pool that doubled in a week is treated more cautiously than one that has held steady at $1B for a year
- Oracle and liquidation design — Chainlink price feeds, time-weighted average prices, multi-source aggregation, and the existence of a working liquidation engine
- Governance and pause authority — whether the protocol can pause itself in an emergency, whether parameters change only by timelock, whether the team has demonstrated competent incident response
Each factor scores 0–5. The total determines the grade.
What the grades mean
Grade A (20+ points) — battle-tested, large-TVL, multi-audited, zero core exploits. The names in this tier are Aave V3, Compound V3, Spark, and Morpho Blue. Yields tend to be lower because the market knows they are safe. These strategies anchor any large position.
Grade B (12–19 points) — solid track record, one or more audits, but younger or smaller TVL. Examples: Fluid, Kinza, Venus on BSC. Yields tend to be 1–2% higher than Grade A; the extra return is your compensation for less operating history.
Grade C (≤11 points) — newer, single audit or no audit, smaller TVL, or known to have had a non-core exploit. The Apyee Keeper rarely includes Grade C strategies in the active rotation. When it does, the cap is tight (≤10% of allocation) and the position is reviewed weekly.
A concrete example
Aave V3 on Ethereum:
| Factor | Score | Reasoning |
|---|---|---|
| Audit depth | 5 | 4 top-tier firms, refreshed at every major upgrade |
| Operating history | 5 | 36+ months on mainnet, no V3 core exploit |
| TVL scale/stickiness | 5 | $15B+ TVL, stable for 18 months |
| Oracle/liquidation | 5 | Chainlink + TWAP fallback, tested in two black swans |
| Governance/pause | 4 | Timelocked governance, working pause guardian |
| Total | 24 | Grade A |
Kinza on BSC:
| Factor | Score | Reasoning |
|---|---|---|
| Audit depth | 3 | Single audit (PeckShield), thorough but not multi-firm |
| Operating history | 3 | 14 months on mainnet, no exploits, but young |
| TVL scale/stickiness | 2 | $80M TVL, fluctuates 20% week to week |
| Oracle/liquidation | 4 | Chainlink-backed, liquidations tested in normal markets |
| Governance/pause | 2 | Centralized admin keys (BSC norm), no timelock |
| Total | 14 | Grade B |
A Grade B protocol like this earns a smaller allocation than a Grade A anchor, and is rebalanced out faster if any signal turns negative.
How grades affect your allocation
The Keeper does not just look at APY when allocating capital. It cross-references the grade:
- A Grade C strategy with 8% APY may lose to a Grade A strategy with 5% APY when the rubric penalizes the risk-adjusted comparison
- The 40% cap on any single strategy applies regardless of grade
- For Grade B and Grade C combined, total allocation never exceeds 50% of the vault
These rules are not negotiable per-user. The vault is a pooled product; every depositor gets the same risk profile.
The honest limitation
A risk grade is an opinion expressed in a number. It can be wrong. The grade does not protect you from:
- A novel exploit class that breaks an audited contract
- A US Treasury redemption queue that locks RWA for weeks
- A coordinated MEV attack that drains liquidity in seconds
What the grade does is make the trade-off legible. If you choose a Grade C strategy chasing 12% APY, you should know exactly which boxes you traded off.
Apyee publishes the grade of every strategy on its detail page. If you disagree with a score, write to [email protected] — grades are revisited quarterly and faster when new evidence appears.