BB Lower + RSI Oversold in Ranging Markets

We tested a mean reversion strategy that buys when price closes below the lower Bollinger Band and RSI(14) drops below 30, filtered to low-ADX (below 50) regimes to avoid strong trends. It passed all six gates with a 1.4% average profit per trade, a 90% win rate in both training and validation, and a deflated Sharpe ratio of 27.81. The probability this is a fluke, after correcting for multiple testing, rounds to zero.

The strategy runs on 15-minute bars and uses a net profit exit, meaning it closes positions when a fixed profit target is reached. The regime filter is the key to the design: by only entering when ADX is below 50, it sidesteps choppy breakouts and focuses on genuine oversold bounces in range-bound conditions. We tested it across ten coins (BTC, ETH, SOL, XRP, DOGE, BNB, ADA, AVAX, LINK, LTC) with consistent results in validation.

Where to stay skeptical: the forward test count is zero. This strategy just graduated and has not yet recorded a single live trade in our paper trading environment. The 90% win rate and 1.4% per-trade profit are entirely in-sample, even though they held across training and validation folds. Until we see live executions, treat the numbers as a ceiling, not a floor. The 15-minute timeframe also means this strategy demands low-latency execution and tight fills. Slippage on a market order during a volatile RSI spike could easily cut into that 1.4% edge.

The deflated Sharpe ratio of 27.81 is unusually high, which either signals a robust edge or a regime that rewarded this exact setup during the test window. We will be watching the forward test closely to see whether the edge persists or decays when the market structure shifts.

If you have your own mean reversion idea or want to test a different regime filter, head to /prove and run it through the same six-gate gauntlet. To see how this strategy compares to the rest of the graduate class, check the full survivor registry at /survivors.