Between June 20–24, 2025, cryptocurrency values took a sharp downturn, triggering one of the year’s largest liquidation cascades across DeFi lending protocols. While price drops were the spark, the deeper story lies in user behavior, credit risk, and how platforms like Aave and Morpho Blue structured that risk.

At Zeru, we used our onchain reputation engine (zScore) to analyze wallet-level risk signals before and after the event — and what we uncovered tells us not just who was liquidated, but why.

• Over 90% of liquidations on both platforms came from users borrowing stablecoins against volatile assets — a predictable but risky pattern.

• Morpho users skewed riskier, with 62% of liquidated wallets having zScores under 300.

• Aave’s liquidations were more diverse, including mid- and even high-score borrowers who got caught by leverage or mistimed moves.

• Withdrawals before the crash were common on Aave (60%) — and a key factor in increasing liquidation risk.

• zScores dropped across the board post-liquidation, reinforcing that one mistake can quickly erase creditworthiness.

We analyzed zScores (Zeru’s behavioral credit metric) across both platforms. Here’s what we found:

| Platform | Avg Pre-Liquidation Score | Median Pre-Liquidation Score |
|----------|---------------------------|------------------------------|
| Aave     |           343.7           |             330.9            |
| Morpho   |           321.6           |             244.2            |

While both protocols saw liquidations primarily in the 200–400 score band, Aave's liquidated users had stronger historical profiles. Morpho, by contrast, had a more bottom-heavy distribution — with most liquidated users falling under 300.

This tells us that risk scores were predictive — but also that historical safety isn't immunity. Even users with 800+ scores were liquidated if they took a wrong turn.

Credit risk isn’t just about what you borrow — it’s how you manage it.

• On Aave, 60% of liquidated users withdrew collateral before the crash, weakening their buffer.

• On Morpho, only 27% did — suggesting a more passive, but also less reactive, user base.

And across both platforms, over 90% of liquidated users borrowed stablecoins against volatile crypto assets.

This asset mismatch (e.g. ETH-in, USDC-out) magnified exposure during price drops. Risk wasn’t random — it was structural.

One of the more revealing metrics is how risk scores changed post-liquidation:

| Platform | Avg Post-Liquidation Score | Median Post-Liquidation Score |
|----------|----------------------------|-------------------------------|
| Aave     |            308.2           |             294.2             |
| Morpho   |            288.3           |             219.8             |

Scores fell across the board, but especially on Morpho, where the median dropped ~25 points — confirming that users already in high-risk categories took the brunt of the impact.

Yet, even high-score users weren’t immune. The score distributions on both protocols were right-skewed — meaning a small number of “safe” users still got hit. Why? Likely due to poorly timed collateral withdrawals, over-leverage, or unattended positions.

For DeFi protocols:

• Implement real-time volatility alerts and withdrawal warnings

• Introduce collateral diversification tools or “portfolio safety” scores

• Customize zScore thresholds to filter for behaviorally sound users

For users:

• Don’t treat stablecoin loans backed by volatile assets as low-risk

• Avoid reducing collateral buffers during turbulent conditions

• Know that reputation isn’t static — it adapts to your latest actions

Credit reputation is a living signal — it must reflect real behavior, not assumptions.

The June liquidation wave proved that zScore (and similar metrics) can help us anticipate risk — but also that scores with long lag times are not enough. We need dynamic models, real-time risk flags, and more education for borrowers.

Because in DeFi, trust isn’t earned once — it’s maintained through every transaction.

zScore APIs are live, and our team is onboarding design partners across DeFi, DAOs, NFTs, insurance, and agentic systems.

Whether you’re building a lending market, airdrop campaign, governance DAO, or identity solution — reputation will define how real your users are.

If you're ready to build with verifiable trust, get in touch on Telegram, Discord, X or at support@zeru.finance.