I Backtested the EMA Crossover Strategy Over 20 Years... Here's What I Found
Summary
Highlights
Introduction and Context00:00:03
The video begins by addressing the popular claim that the EMA crossover is a profitable strategy. The presenter, Seneca, an engineer and quantitative researcher, explains that he tested the strategy across 28 Forex currency pairs on 4-hour and daily timeframes, and also examined variations with filters. He warns about data snooping, the danger of over-optimizing a strategy to find false positives. Seneca introduces himself and his channel, Lock Quant Research, where he rigorously tests trading strategies using historical data to provide evidence-based results.
Classic EMA Crossover Test00:02:22
The first test is the classic EMA crossover, where you buy when the 50 EMA crosses above the 200 EMA and exit on the opposite crossover. After testing over 1,100 trades, the strategy produced a significant overall loss and a steadily declining equity curve over 20 years, showing no reliable edge.
Price Confirmation Filter00:03:10
Next, a simple price confirmation rule was added: instead of entering immediately on every crossover, price also had to close on the correct side of the moving average. However, the results were almost identical to the original strategy, with no edge found.
Volatility Filter with ATR00:03:35
The volatility filter uses the Average True Range (ATR) to only allow trades when volatility is relatively high. This version showed a completely different equity curve, becoming more stable and ending with a positive return. While not conclusive, it was the first strategy that earned further testing.
ADX Filter00:04:46
The ADX filter, which measures trend strength, was tested with the idea that the EMA crossover might perform better in strong trends. The overall improvement was not convincing, but when filtering by currency groups, some slightly positive returns appeared. This also passed stage one and warrants further investigation.
Triple EMA Crossover00:05:31
The triple EMA crossover, using three EMAs like 20, 50, and 200, was tested but never got off the ground, showing no edge.
Conclusion and Call to Action00:05:56
After testing five variations, three showed little or no evidence of a reliable edge. Only the volatility filter and ADX filter passed stage one and earned further investigation. Full dashboards are available on the website. The presenter invites viewers to suggest strategies for future testing and emphasizes that he will show whether the data supports them or not.
Recently Summarized Articles
Related summaries
- I Tested The 9 20 EMA Strategy: Here Is The Truth
- Financial Management - Complete Chapter in 30 Minutes | Class 12th Business Studies 🔥
- Panama Canal Documentary On Building The Panama Canal Full Documentary
- A Book from 1857 Predicts Our Future a Little Too Well
- You're watching the replay of the 6ABC 11PM newscast; it's episode 14 of season 2.
- With a TV-G icon, here's the replay of the 9th episode of the 9AM newscast from CBS News Philly