SMA 21/50 Cross With ADX Filter Graduates With 12.7 Sharpe

We tested a simple moving average crossover strategy (21-period SMA crossing above the 50-period SMA for entry) with an ADX regime filter that only trades when ADX is below 50, screening out strong trend environments. The strategy graduated after passing all six gates with a deflated Sharpe ratio of 12.70 and a 0.0% probability of being a fluke after multi-test correction. Walk-forward training and validation both returned 2.198% average profit per trade with 100% win rates.

The concept is a reversal play on the classic golden cross. Most traders use SMA crossovers to chase trends. This gene inverts that logic by filtering out high-ADX regimes, meaning it only enters when the 21 crosses the 50 in low-trend, choppy conditions. The exit is net profit based, so it closes when a target is hit rather than waiting for a signal reversal.

We ran the backtest across ten coins (BTC, ETH, SOL, XRP, DOGE, BNB, ADA, AVAX, LINK, LTC) on 15-minute bars, with no session restrictions (trades around the clock). The deflated Sharpe ratio accounts for multiple testing and the number of backtests run before this gene emerged, which is why we trust the 12.70 figure more than raw Sharpe.

Now for the honest caveats. First, the strategy has logged zero forward trades in live paper trading so far. That means every number above comes from backtest, not real fills. Second, a 100% win rate in both training and validation is a red flag for sample size. We do not have the trade count stored at the gene level, but a perfect win rate usually means fewer than a dozen trades. Third, the ADX filter is doing heavy lifting. If the market shifts into a sustained trending regime across all ten coins, this strategy will sit idle for weeks.

We are watching this one closely in forward testing. The statistical gates it passed are rigorous, but the zero-trade forward count and the 100% backtest win rate make it fragile until we see real fills.

Check the full survivor registry at stratproof.com/survivors to compare this gene against other graduates, or visit stratproof.com/prove to test your own regime-filtered crossover ideas.