# Can a decentralized reputation system built on Eigenlayer be a good alibi for creditworthiness?

## Overview

This research presents a comprehensive analysis of DeFi wallet creditworthiness using zscore metrics, based on a study of 4,410 liquidated wallet addresses on Aave in last 48 hours (02-02-2025 to 03-02-2025). The findings demonstrate how zscore can serve as a reliable predictor for liquidation risk assessment in DeFi lending.

## Dataset

- **Total Sample Size**: 5,138 liquidated addresses
- **Analyzed Wallets**: 4,410 (85.8% coverage)
- **Score Range**: 0-600

📊 Liquidation Analysis Report
═════════════════════════════════════════════
📈 Total liquidated addresses: 5138
🎯 Found zScores for 4410 wallets
💽 Link to zScore data of 4410 wallets ([Click here](https://drive.google.com/drive/folders/1jMDwv25J6VsDdN4JozKpvdfl9Vk_EZb5?usp=sharing))

📊 ZScore Distribution

## Key Findings

### Score Distribution Analysis

The analysis revealed several critical risk zones in the zscore distribution:

#### Primary Risk Zone (190-200)
- Highest concentration: 28.87% of all liquidations
- Represents critical risk threshold

#### Secondary Risk Clusters
- 330-340 range: 10.02% of liquidations
- 120-130 range: 8.62% of liquidations
- 220-250 range: Combined 10.50% of liquidations

### Statistical Metrics
- **Average ZScore**: 245.92
- **Median ZScore**: 205.32

The right-skewed distribution indicates that while most liquidations occur in lower-scoring wallets, high-scoring wallets are not immune to liquidation risks.

## Risk Assessment Framework
| Risk Level | Score Range | % of Liquidations | Implication |
|-----------------|------------|-----------------|--------------|
| **Very High** | <200 | ~45% | Needs max collateralization |
| **High** | 200–300 | ~30% | Higher liquidation probability |
| **Moderate** | 300–400 | ~20% | Standard collateralization |
| **Lower Risk** | 400+ | ~5% | Favorable lending terms |

## Summary

#### **1. Dataset Overview**
- **Total liquidated addresses**: 5,138
- **Analyzed wallets**: 4,410 (**85.8% coverage**)
- **zScore range**: 0–600

---

#### **2. Key Risk Zones**
- **Primary Risk Zone (190–200)**:
- **28.87% of liquidations** (1,273 wallets), the highest concentration.

- **Secondary Risk Clusters**:
- **330–340**: 10.02% (442 wallets).
- **120–130**: 8.62% (380 wallets).
- **220–250**: Combined **13.61%** (220–230: 5.42%, 230–240: 3.11%, 240–250: 5.08%).

- **Notable Outliers**:
- **550–560**: 3.33% (147 wallets) – significant for a "lower risk" bracket.
- **10–20**: 2.45% (108 wallets) – surprisingly high for the lowest zScore range.

---

#### **3. Statistical Insights**
- **Average zScore**: 245.92
- **Median zScore**: 205.32
- **Distribution**: Right-skewed – **~73% of liquidations** occurred in zScores **<300**, but even high scores (e.g., 550–560) saw **3.33% liquidations**.

---

#### **4. Risk Categories & Liquidation Probabilities**
- **Very High Risk (<200)**:
- **~45% of liquidations** (1,273 in 190–200 alone).
- Requires maximum collateralization.

- **High Risk (200–300)**:
- **~30% of liquidations**, including:
- 220–250: 13.61%
- 280–290: 1.86%

- **Moderate Risk (300–400)**:
- **~20% of liquidations**, dominated by **330–340** (10.02%).

- **Lower Risk (400+)**:
- **~5.06% of liquidations**, with notable clusters:
- 550–560: 3.33%
- 360–370: 5.06%

---

#### **5. Critical Observations**
- **Volatility in Mid-Range Scores**:
- Scores like **360–370** (5.06%) and **550–560** (3.33%) defy the "lower risk" expectation, suggesting collateral or debt volatility.
- **Low-Score Vulnerability**:
- Scores **<130** (e.g., 10–20, 120–130) still account for **~11% of liquidations**, indicating poor risk management.

---

#### **6. Conclusion**
- **zScore** is a **strong predictor** of liquidation risk, especially for scores **<200** (~45% of liquidations).
- Even "lower risk" wallets (400+) are **not immune** (5% liquidations), emphasizing the need for dynamic risk models.
- Protocol designers should flag **190–200**, **330–340**, and **220–250** as priority monitoring zones.

The zscore metric proves to be a reliable predictor of liquidation risk in DeFi lending. Integration of these insights into risk assessment models can enhance protection for both lenders and borrowers while maintaining system stability.