Optimizing Take-Profit and Stop-Loss Levels with a Train-Test Backtest

Summary

This project tests whether adjusting percentage-based take-profit (TP) and stop-loss (SL) levels can improve strategies using simple or exponential moving-average crossovers and RSI signals. For each strategy and market series, it searches TP and SL combinations on the training portion of daily US equity and forex data, then applies the chosen pair unchanged to the test portion. Test returns are compared with buy-and-hold and a fixed TP/SL control; trades hold at most one position at a time.

The reported results are mixed. Some forex SMA and RSI combinations outperform the comparisons, while optimized exits often fail to improve stock results, can trail the control, and produce no clear overall conclusion for EMA. The author flags possible survivorship bias in the stock sample. The exercise shows how exit parameters can be evaluated while holding entry signals constant, but selecting the best pair from training data risks overfitting, and the document provides no detailed performance figures in the text. Its limited asset set and period also constrain how broadly the findings apply.

Key ideas

• The project searches training data for TP and SL levels while keeping each strategy’s entry rules fixed.
• It evaluates SMA crossover, EMA crossover, and RSI signals on daily equity and forex series.
• The selected exit levels are carried unchanged into a held-out test set and compared with buy-and-hold and a fixed-exit control.
• Reported results vary by market and strategy, with no consistent benefit from optimized exits.
• The stock results may be affected by survivorship bias, and training-set optimization may not generalize.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.