Investment Thesis: Enhancing Algorithmic Trading Using RSI Indicator

 

Thesis Statement

 

Our algorithm tradings strategy trades stocks, crypto, funds, and futures (commodities). It divides the percentage of the portfolio equally between the four, giving 25% to each. Trades are only performed during a bullish momentum in the market, that is it's creating higher highs in biweekly time periods, and are triggered based on the Relative Strength Index (RSI): entering a position if RSI exceeds 60 and is increasing (signaling a bullish momentum), or when RSI is below 20 and is decreasing (signaling that it's been very oversold). We exiting if RSI falls below 40 and is decreasing (signaling a bearish momentum), or if RSI is above 80 and increasing (signaling that is been very overbought). We set the SPY as benchmark and a 30 day warmup.

Areas to review and improve

Allocation Percentage

For simplicity, the model allocates the same 25% to each asset type. It can be tested to see which asset type performs the best and have the algorithm change the allocation percentages based on performance.

Adaptive Indicators and Robust Volume Filters

Fixed RSI thresholds may fail in varying market conditions, leading to false signals. Adaptive RSI thresholds, adjusted for historical volatility, and additional indicators improve signal accuracy and market clarity. Enhanced volume filters using multiple time frames or volume percentile ranks ensure accurate liquidity measurement, identifying genuinely investable assets.

Comprehensive Benchmarking and Efficient Execution

A single asset benchmark, like SPY, doesn't reflect a diversified portfolio's performance. Using a comprehensive asset index better evaluates strategy effectiveness and market alignment. Limit orders reduce slippage, ensuring better execution prices, while sophisticated exit strategies with trailing stops and additional indicators optimize timing and enhance overall performance.

Author

Quant League Competitions

January 2025