Financial markets aren’t just math — they’re emotion.
Fear. Greed. Panic. FOMO. Confidence. Hesitation.
Every candle tells a psychological story.So I decided to answer one powerful question:
Does market sentiment actually influence trader performance?
To explore this, I combined two datasets:
Bitcoin Fear & Greed Index (daily sentiment)
Hyperliquid Historical Trader Data (real trades: PnL, position size, timestamps)
Once merged, the insights were surprisingly clear.
📁 Datasets Used
1. Bitcoin Fear & Greed Index
The famous index that quantifies market psychology on a scale of 0–100.
Score Range Market Mood
0–24 Extreme Fear
25–44 Fear
45–54 Neutral
55–74 Greed
75–100 Extreme Greed
Columns:
- date
- value
- classification
- Hyperliquid Trader Executions
Contains:
- Timestamp
- Execution Price
- Size USD
- Side (Buy/Sell)
- Closed PnL
This tells how the trader performed each day.
🧹Step 1 — Data Cleaning & Preparation
We converted UNIX timestamps → dates and merged the datasets.
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
sns.set(style="whitegrid")
trades = pd.read_csv("historical_data.csv")
sentiment = pd.read_csv("fear_greed_index.csv")
trades['Timestamp'] = pd.to_datetime(trades['Timestamp'], unit='ms')
trades['date'] = trades['Timestamp'].dt.date
trades_clean = trades[['Timestamp','date','Execution Price','Size USD','Side','Closed PnL']]
trades_clean.rename(columns={'Closed PnL':'pnl'}, inplace=True)
sentiment['date'] = pd.to_datetime(sentiment['date']).dt.date
sentiment_clean = sentiment[['date','value','classification']]
merged = pd.merge(trades_clean, sentiment_clean, on='date', how='left')
Now every trade has the market sentiment associated with it.
🔍 Step 2 — Exploratory Data Analysis
📊 Fear & Greed Index Over Time
Helps visualize the emotional highs and lows of the market.
📈 Trader PnL Over Time
Shows profitable and rough periods.
💡 Average PnL by Sentiment
avg_pnl = merged.groupby('classification')['pnl'].mean().sort_values()
💥 Key Result:
Best performance: ✔ Greed
Worst performance: ❌ Extreme Fear
The trader clearly performs better when markets are trending and confident.
📏 Position Size by Sentiment
avg_size = merged.groupby('classification')['Size USD'].mean()
Insight:
Traders take larger positions during Greed phases.
This highlights a psychological pattern:
- Confidence → bigger trades
- Uncertainty → smaller positions
🎯 Win Rate by Sentiment
merged['win'] = (merged['pnl'] > 0).astype(int)
win_rate = merged.groupby('classification')['win'].mean()
Pattern:
Win rate peaks during Greed, drops during Fear.
📉Correlation Between Sentiment & PnL
corr = merged[['value','pnl']].corr()
🧠 Step 3 — Key Insights
✔ 1. Performance increases in Greedy markets
Trends become cleaner and easier to ride.
✔ 2. Losses spike during Extreme Fear
Market becomes chaotic.
✔ 3. Trader takes larger positions when confident
Risk-taking aligns with sentiment.
✔ 4. Win rate highest in Greed phases
Clear sign that trend-following works better here.
✔ 5. Sentiment can filter bad trades
Avoiding Extreme Fear days = fewer drawdowns.
🔥 Step 4 — A Simple Sentiment-Aware Trading Strategy
You can turn these insights into a practical system:
🟥 Avoid trades when Fear Index < 20 (Extreme Fear)
Market too volatile → losses increase.
🟩 Increase position size when Greed Index > 60
Strong trends → higher win rate.
🟨 Trade normally in Neutral zones
Mean reversion works better here.
This simple sentiment filter can significantly improve risk-adjusted returns.
🎯 Conclusion
This analysis proves:
Market sentiment directly impacts trader performance.
Greed = opportunity
Fear = danger
By incorporating the Fear & Greed Index into decision-making, traders can:
- Avoid bad market conditions
- Maximize trend opportunities
- Understand their own psychological biases
- Improve long-term consistency
This combination of data + psychology = a real trading edge.
Top comments (0)