Lumibot: Sentiment Analysis Strategy for Backtesting

[shaziaakhan]

Implementing a backtesting strategy for a trading bot that leverages sentiment analysis to inform buy and sell decisions.

Functionality:

• Backtesting: Evaluates the strategy's performance using historical data.
• Sentiment-Based Trading: Analyzes news headlines to estimate positive or negative sentiment for a chosen stock (default: SPY).

Process:

1. Gathers Sentiment: Fetches news headlines for the target stock within a specific timeframe.
2. Estimates Sentiment: Analyzes the headlines using a custom estimate_sentiment function to determine sentiment and its confidence level.
3. Trading Logic (High Confidence Only):
   • Only considers trades when the sentiment analysis has extremely high confidence (>0.999)
   • If sentiment is positive and confidence is high:
     • Sells existing holdings if previously bought.
     • Places a buy order with a "bracket order" (including take-profit and stop-loss prices).
   • If sentiment is negative and confidence is high:
     • Sells existing holdings if previously bought.
     • Places a sell order with a bracket order.
4. Position Sizing: Allocates a portion of available cash (cash at risk percentage) for each trade.

Files added:

bot.py
utils.py
requirements.txt

[shaziaakhan]

@Ayushpanditmoto it has been more than a week! please merge this PR, thanks.