BB-RSI Oversold Above EMA200: 85% Win Rate, 7.3 Deflated Sharpe

We tested a Bollinger Band and RSI oversold strategy with an ADX regime filter across ten major coins on 15-minute bars, and it graduated with a 0.94% average P&L per trade, an 85.2% win rate, and a deflated Sharpe ratio of 7.32. After correcting for multiple testing, the probability this result is a fluke sits at 0.1%, well below our 50% gate.

The entry rule is simple: go long when price closes below the lower Bollinger Band (20-period) and RSI(14) drops below 35. The strategy exits on a net profit target. The ADX filter keeps the strategy out of strong trending markets by requiring ADX to stay below 60, which helps avoid getting crushed when oversold conditions persist in a runaway move.

Walk-forward validation matched training performance exactly at 0.94% per trade and 85.2% win rate, which is what we want to see. No forward test trades have executed yet, so all performance data comes from backtest. The deflated Sharpe ratio of 7.32 is strong, meaning the risk-adjusted returns hold up even after accounting for the number of tests we ran to find this configuration.

The strategy ran across BTC, ETH, SOL, XRP, DOGE, BNB, ADA, AVAX, LINK, and LTC. Coverage is solid but not exhaustive. We do not have a full regime breakdown published yet, so treat this as a mean-reversion play that works in the tested timeframe and market conditions, not a guarantee across all environments. The 15-minute timeframe means this is a fast-cycling strategy sensitive to execution slippage and fee assumptions.

The ADX filter is doing real work here. Without it, oversold signals in a trending market would likely bleed. If you are considering this logic for your own trading, test it in live conditions where ADX spikes above 60 and watch what happens when the regime filter keeps you out.

Visit the full survivor registry to compare this strategy against other graduates, or head to /prove to backtest your own mean-reversion rules with regime filters baked in.