Golden Cross EMA(50/200): What Two Trades Can Prove

A fast moving average crossing above a slow one is treated as a signal of a long-term bull trend. We test the canonical 50 and 200 period pair on daily bars.

Golden cross: the entry condition

Long while EMA(50) is above EMA(200), short while it is below. The rule is almost permanently in the market and changes position very rarely: slow averages cross only a handful of times in a two-year sample.

What the EMA(50/200) run showed

Run with the portal's own engine over 518 daily bars for each of 12 pairs, costs of 1 pip round turn, canonical parameters. The method and its limitations are set out in how we run backtests.

• Median return — 1.22%, profitable on 7 of 12 pairs
• Sharpe — 0.17, drawdown — 6.1%
• Trades — 2, win rate — 50%, time at risk — 62%
• Extremes: EURJPY +15.2%, NZDUSD -22.0%

Two trades per pair: why that settles the question

The median result is formally positive and more than half the pairs made money. But the number to look at is not return, it is trade count: a median of two per pair. Two trades are not statistics, they are two outcomes. The spread between pairs confirms it — from a substantial gain to a comparable loss, meaning the result depends entirely on whether the single cross happened to catch a real trend.

What this backtest actually teaches

This test is useful not for its result but as an illustration: a positive number on a sample of two trades means nothing. Had we selected this rule on that basis and traded it live, we would have been trading a coin flip. This is what self-deception in testing looks like — not falsified data, but an ignored sample size. For comparison: simply holding the position on the same data returned 4.79%, and this rule did not beat it — why a benchmark is not optional.

The limitation that decides everything here

The binding constraint is specific rather than general: two trades per pair. At that number the confidence interval around any estimate is wider than the estimate itself, and no amount of care elsewhere repairs it. Slow averages need decades, not two years. The general limitations — a short sample, correlated pairs, a signal tested without risk management — are listed in the methodology. The data is open and the run reproduces from a script in the repository.

This material is educational and is not individual investment advice. Backtested results do not guarantee similar results in the future. Trading forex carries the risk of losing capital.