Trader Behavior vs Market Sentiment

By Amirtha Ganesh R Date: December 2025

🔗 Google Colab Notebook

Main Analysis Notebook:
[https://colab.research.google.com/drive/14tb6KqEB8ijX3mXJUUKfoLAGxHqsLSlU?usp=sharing]

Access: Set to "Anyone with the link can view"

📊 Analysis Summary

This project analyzes the relationship between trader behavior (profitability, risk-taking, volume) and market sentiment (Fear/Greed) using:

• Historical Trader Data from Hyperliquid (211,224 trades)
• Bitcoin Fear & Greed Index (2,644 daily sentiment labels)

Key Findings:

1. Fear periods generate 2.5× higher daily PnL despite lower activity
2. Greed periods show 87% higher risk ratios but lower efficiency
3. Neutral conditions yield highest PnL per trade (63.82 USD)
4. Top 10% of traders capture 63% of profits during Fear vs 58% during Greed

🛠️ Technologies Used

• Python 3.12 (Google Colab)
• Pandas, NumPy - Data processing
• Matplotlib, Seaborn - Visualizations
• Statistical Analysis - Sentiment comparison, quintile analysis

📝 Datasets

1. Fear & Greed Index: Daily Bitcoin sentiment classifications
2. Hyperliquid Trader Data: Trade-level execution records (May 2023 - May 2025)

📄 Report

The ds_report.pdf contains:

• Data preparation methodology
• Comprehensive statistical analysis
• Actionable trading insights
• Limitations and future work

---

Contact: amirthaganeshramesh@gmail.com / +91 6374707500