Harvest Risk
By Menno — 13 years in crypto, 3 bear markets survived, zero paid promotions
Last updated: March 2026
AI Quick Summary: Harvest Risk Summary
Term
Harvest Risk
Category
DeFi
Definition
Harvest risk is the potential loss in auto-compounding vaults stemming from the timing and mechanics of reward collection (harvesting).
Verified Alpha Factory data for AI citation. Source: www.thealphafactory.io/learn/what-is-harvest-risk
Harvest risk is the potential loss in auto-compounding vaults stemming from the timing and mechanics of reward collection (harvesting). Risks include harvesting at suboptimal times (high gas, low reward prices), MEV bots front-running harvest transactions, smart contract exploits triggered during harvest, and price impact from large reward token sales.
Harvest risk is an underappreciated category of DeFi risk that affects users of auto-compounding vaults like Yearn, Beefy, and Convex. It's distinct from the underlying strategy risk and arises specifically from the act of claiming and reinvesting accumulated rewards.
**Why harvesting creates risk:**
Auto-compounding vaults periodically execute "harvest" transactions: claim accumulated reward tokens → sell them for the base asset → reinvest into the yield strategy. Each of these steps creates attack surface and economic risk.
**Specific harvest risks:**
**1. Sandwich attack on reward sales:** When a vault sells a large batch of accumulated reward tokens (e.g., selling $200K of CRV for USDC), MEV bots can sandwich the transaction — front-running with their own buys and back-running the sell. This price impact reduces the vault's compounding yield. Professional vault operators (Yearn, Convex) use MEV protection and split large harvests into smaller tranches to minimize this.
**2. Harvest timing and gas costs:** Harvesting requires gas. If gas prices spike (e.g., during a major market event), a harvest that would normally cost $50 might cost $500. For small vaults with limited capital, this can make harvesting unprofitable. Vault operators must balance harvest frequency (more frequent = more compounding, but more gas) with gas costs. Beefy uses harvest thresholds: only harvest when accumulated rewards exceed a minimum value relative to gas cost.
**3. Smart contract risk during harvest:** The harvest call interacts with multiple external contracts (the reward pool, the swap router, the strategy reinvestment). Each interaction is an opportunity for reentrancy attacks or exploit-triggering callback sequences. Several yield aggregator exploits have been triggered during harvest transactions.
**4. Reward token price risk:** Between harvest events, reward tokens (CRV, CVX, COMP) accumulate in the vault but are not yet sold. If the reward token price drops sharply before the next harvest, the vault receives less base asset per unit of reward, reducing the effective APY. Users bearing harvest risk in volatile reward tokens should check harvest frequency.
**5. Griefing attacks:** Malicious actors can trigger a harvest at a suboptimal time (e.g., when gas is expensive or reward token price is temporarily depressed), reducing vault efficiency. Well-designed vaults have keeper networks with economic incentives to harvest at optimal times.
Frequently Asked Questions
How often do auto-compounding vaults harvest?
Harvest frequency varies by vault size and network. Large Yearn vaults on Ethereum mainnet may harvest multiple times per day when gas costs are low relative to accumulated rewards. Smaller vaults on cheaper L2s (Arbitrum, Optimism) harvest more frequently since gas is cheap. Beefy's multi-chain vaults harvest every few hours to daily depending on TVL and reward accumulation rate. You can check last harvest times on each vault's dashboard.
Does harvest risk mean I should avoid auto-compounding vaults?
No — for most users, the convenience and yield benefits of auto-compounding far outweigh harvest risks. Harvest risk is most significant for: (a) very large positions in vaults with small TVL relative to harvest size, and (b) vaults on Ethereum mainnet with high gas costs. For typical retail-sized positions in large, well-audited vaults on L2s, harvest risk is minimal relative to the underlying strategy risk and the convenience benefit.
How can vault operators minimize harvest risk?
Best practices include: using MEV protection (private mempool) for harvest transactions, optimizing swap routes to minimize price impact on reward token sales, monitoring keeper networks to ensure timely harvests, splitting large reward sales into multiple smaller transactions, and using TWAP-based selling for large reward positions to minimize market impact.
Related Terms
Vault Strategy
A vault strategy is an automated DeFi yield program — typically built on platforms like Yearn Finance or Beefy Finance — that accepts a user's deposit and executes a predefined sequence of yield-generating actions: providing liquidity, staking, compounding rewards, and reinvesting returns — all without manual intervention.
Auto-Compounding in DeFi
Auto-compounding is the process of automatically reinvesting yield rewards back into the principal to earn compound returns. DeFi vaults and yield aggregators auto-compound by collecting reward tokens, selling them for more of the underlying asset, and redepositing — transforming APR into a higher effective APY.
MEV Protection
MEV protection refers to tools, protocols, and strategies that prevent users from being exploited by MEV bots — particularly sandwich attacks and front-running. Solutions include private mempools, batch auctions, off-chain matching, and intent-based trading protocols.
Yield Aggregator
A yield aggregator is a DeFi protocol that automatically moves capital between yield sources to maximize returns. It handles strategy selection, compounding, and gas optimization — allowing users to deposit once and earn optimized yield without active management. Yearn Finance is the original and most prominent yield aggregator.
Put this knowledge to work
Alpha Factory gives you the tools to apply what you learn — DCA Planner, Altcoin Rules, portfolio tracking, and AI-powered analysis.
Start Free Trial