Testing MACD Filters on Five Years of US_TECH100 Data

Summary

This study tests whether regime, higher-timeframe trend, and trading-session filters improve a MACD histogram crossover strategy on H1 US_TECH100 broker data. The baseline uses classic MACD settings, ATR-based stop and target distances, and a fixed maximum holding period. The author adds filters incrementally so their individual contributions can be compared across the same instrument and entry logic.

In the reported five-year sample, regime filtering and subsequent EMA200 alignment reduced trade counts without improving expectancy; regime filtering also removed volatile-regime trades that were comparatively productive in the raw strategy. Adding a US-session window produced the strongest reported change: fewer trades, higher expectancy, and lower maximum drawdown. The author suggests thinner liquidity and weaker overnight signals as possible explanations, but does not establish causation. Results are specific to one broker’s US_TECH100 feed, timeframe, and parameter choices; they should not be assumed to transfer to other instruments or sessions. The article includes code intended to reproduce the analysis.

Key ideas

• The test compares progressively filtered versions of the same MACD crossover strategy.
• The regime and higher-timeframe filters did not improve the reported expectancy in this sample.
• The session restriction was associated with fewer trades and better reported risk-adjusted results.
• Raw MACD performance varied by regime, with volatile conditions outperforming the trend category in the reported breakdown.
• The findings are limited to one instrument, broker feed, timeframe, and test design.

Tags

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