📊 Bitcoin Market Sentiment vs Trader Performance Analysis

This project explores how Bitcoin market sentiment — measured by the Fear & Greed Index — influences trader performance on the Hyperliquid platform. Using real execution data and sentiment scores, we uncover patterns, build predictive models, and offer strategic insights to optimize trading decisions.

---

📁 Datasets Used

1. Bitcoin Fear & Greed Index

• Source: Alternative.me
• Columns: date, value, classification (e.g., Extreme Fear, Fear, Neutral, Greed)

2. Hyperliquid Historical Trader Data

• Source: Proprietary execution logs
• Columns include:
  • account, symbol, execution price, size, side, timestamp, start position, event, closedPnL, fee, leverage, etc.

---

🎯 Objective

To explore the relationship between trader performance and market sentiment, uncover hidden behavioral patterns, and deliver insights that can drive smarter trading strategies.

---

🧪 Methodology

✅ Data Processing

• Cleaned and merged sentiment data with trader logs
• Converted timestamps and aligned formats
• Encoded categorical variables (e.g., sentiment, side)

🧠 Feature Engineering

• PnL_per_token: Profit per token traded
• efficiency: PnL relative to execution price
• Time-based features: hour, day_of_week

📊 Exploratory Analysis

• Sentiment distribution
• PnL trends across sentiment phases
• Correlation heatmaps

🤖 Predictive Modeling

• Random Forest Regressor to predict Closed PnL
• Features: execution price, trade size, sentiment, time features
• Evaluation: R² score, RMSE

---

📈 Key Insights

• Traders tend to perform better during Greed phases, with higher average PnL and efficiency.
• Extreme Fear correlates with lower execution efficiency and smaller trade sizes.
• Time-of-day and day-of-week features show subtle but consistent performance patterns.
• Sentiment is a valuable signal when combined with trade metadata.

---

📦 Files Included

File	Description
sentiment_trader_analysis.ipynb	Full Jupyter notebook with code, charts, and models
historical_data.csv	Raw trader execution data
fear_greed_index.csv	Daily sentiment scores
merged_trader_sentiment_analysis.csv	Final merged dataset with engineered features
sentiment_strategy_summary.csv	Aggregated insights by sentiment phase

---

📌 How to Run

1. Clone the repo:

   git clone https://github.com/yourusername/bitcoin-sentiment-trader-analysis.git
   cd bitcoin-sentiment-trader-analysis