Algotrader

Really complicated gambling

File Descriptions:

analysis.py

• Runs backtesting analysis metrics for portfolio performance.

testing.py

• Proof of concept file.
• Takes any ticker and date range and calculates 20-day SMA and Bollinger Bands.
• Trading Strategy:
  • Sell: If SMA is above the upper band.
  • Buy: If SMA is below the lower band.

pairs.py

• Pairs Trading:

  1. Generating Pairs:

     • Takes in a list of stocks.
     • Checks for cointegration of them over a set window for initial filtering.
     • Note: This initial filter isn't strictly necessary since a sliding window is used to constantly check for cointegration, but it helps in maintaining system performance.

  2. Generating Signals:

     • Moves the sliding window into the "live" data.
     • Continuously checks if cointegration is maintained.
     • Dynamically updates z-scores and residuals.
     • Generate signals based on z score crossing thresholds

  3. Backtesting:

  • Does the calculations for returns and percent change
  • Gives metrics like total return, CAGR, Sharpe ratio