AI-Powered Algo Trading: Build, Backtest & Trade Strategies! with Mr. Saddam Shaik.

In this Face2Face episode, Vivek Bajaj speaks with Saddam, a Hyderabad-based bank manager and self-taught algo trader who has been active in the markets since 2016–17 and using algorithms since 2019–20. Remarkably, Saddam hasn’t placed a manual trade in over four years. His focus is now primarily on research, strategy development and validation.

The central theme of the conversation is how AI tools such as Claude are making systematic trading accessible even to people without coding knowledge. Saddam demonstrates how AI can help traders download historical market data through broker APIs, understand API documentation and build integrations using simple natural-language prompts. He also takes a trading strategy from a random YouTube video, extracts its rules and uses AI to code and backtest it within minutes.

The resulting analysis includes trade logs and performance metrics such as CAGR, Sharpe, Sortino and Calmar ratios. He then explains his own systematic investing frameworks. His ETF Momentum Rotation strategy ranks 54 ETFs using weighted 1-, 3-, 6- and 12-month momentum and selects the strongest candidates. His Nifty 500 momentum models progressively add RSI, 52-week-high and broader market regime filters to improve risk management and reduce drawdowns. Saddam also discusses capital allocation between ETF momentum, stock momentum, cash and F&O strategies.

A particularly important lesson comes from the popular 9:15 short-straddle strategy. His backtesting shows that once realistic slippage and execution costs are included, much of the apparent profitability disappears. The key takeaway: AI can make strategy creation and backtesting easier, but realistic assumptions, proper validation and disciplined risk management remain essential.