Analyzing how trading behavior (profitability, risk, volume, and leverage) aligns with market sentiment (Fear vs. Greed)

---

Overview

This project explores the relationship between market sentiment and trading behavior, using the Fear & Greed Index alongside historical trade data. The goal was to identify whether emotional market conditions such as Fear and Greed influence how traders perform in terms of profitability, participation, and risk-taking behavior.

The workflow follows a structured, multi-stage process:

1. Data cleaning and preparation
2. Independent exploration of sentiment and trading data
3. Integrated analysis of trading behavior under different emotional regimes

---

Folder Structure

DS_Kaveri_Biswas/
│
├── notebooks/
│   ├── Book1_Data_Preparation.ipynb
│   ├── Book2_Independent_EDA.ipynb
│   └── Book3_Merged_Analysis.ipynb
│
├── data/
│   ├── sentiment_cleaned.csv
│   ├── data_cleaned.csv
│   └── merged_dataset.csv
│
├── summaries/
│   ├── Summary_Book1_Data_Preparation.md
│   ├── Summary_Book2_EDA.md
│   ├── Summary_Book3_Merged_Analysis.md
│   └── Final_Project_Summary.md
│
├── outputs/
│   ├── sentiment_trend.png
│   ├── pnl_distribution.png
│   ├── volume_vs_sentiment.png
│   └── risk_efficiency_chart.png
│
├── report/
│   └── ds_report.pdf
│
└── README.md

---

Phase 1: Data Preparation (Book1)

Goal: Prepare both datasets for analysis.

• Standardized all timestamps (UTC alignment).

• Removed duplicates, invalid rows, and missing timestamps.

• Aggregated to daily granularity to merge market sentiment and trading data.

• Created date_key for consistent merge logic.

• Exported cleaned datasets:

  • sentiment_cleaned.csv
  • data_cleaned.csv
  • merged_dataset.csv

Outcome: Three high-quality datasets, each consistent and merge-ready for behavioral analysis.

---

Phase 2: Independent Exploration (Book2)

Sentiment Data

• Covered 2018–2025, showing recurring emotional cycles of Fear and Greed.
• High autocorrelation (~0.95) indicating persistence in sentiment.
• Emotional extremes align with major volatility events, confirming sentiment’s predictive potential.

Trading Data

• Spanned Sep 2024–Apr 2025 with 7,000+ trades.
• Most trades near breakeven; few large wins/losses dominated results.
• Daily activity spiked in early 2025 — suggesting high market volatility.

Outcome: Understanding both emotional context (sentiment) and behavioral response (trading) separately helped define merging logic and hypothesis for integrated analysis.

---

Phase 3: Integrated Analysis (Book3)

1. Profitability vs Sentiment

• Mean profit rises with optimism (Neutral → Fear → Greed → Extreme Greed).
• Median profit ~0 across all phases, showing heavy-tailed outcomes.
• Win rate peaks in Extreme Greed (≈47%) and drops in Neutral (~35%).
• Risk also rises sharply, showing overconfidence bias during Greed.

2. Volume & Participation vs Market Sentiment

• Trading volume highest during Fear and Extreme Fear.
• Fear-driven spikes often precede sentiment changes, indicating panic-based reaction.
• Greed periods show fewer but more confident trades.

3. Risk & Efficiency

• PnL volatility and risk-taking behavior surge in emotional markets.
• Traders behave defensively under Fear (larger, concentrated bets).
• Greed produces frequent, smaller, confidence-driven trades.

Outcome: Emotions don’t dictate trade direction but intensity. Fear inflates participation and volatility; Greed improves success rate but not consistency. Sentiment-driven awareness can guide smarter position sizing and leverage management.

---

Key Visual Outputs

Saved in /outputs/:

• sentiment_trend.png: Evolution of Fear–Greed sentiment over time
• pnl_distribution.png: Distribution of profitability across sentiment regimes
• volume_vs_sentiment.png: Trading volume behavior under Fear vs. Greed
• risk_efficiency_chart.png: Risk–return efficiency comparisons across emotional states

---

Report and Summaries

All supporting explanations are documented in the /summaries/ folder. The complete compiled report (ds_report.pdf) is included in /report/.

---

Final Insight

Market emotion governs intensity, not direction. Fear triggers overreaction and volume surges; Greed fuels confidence and profit variance. Neutral sentiment brings calm and balanced performance. Combining emotional indicators with behavioral metrics can significantly improve trading strategy design.