Mean Reversion Strategies: RSI-2 and Beyond

Mean reversion strategies bet that extreme moves will reverse. RSI-2 is the classic — buy when the 2-period RSI drops below some threshold (typically 5-10), sell when it climbs back above another threshold (typically 70-90). It looks simple in backtests. Trading it live reveals every subtle mistake in the backtest.

The first mistake is testing on the wrong universe. RSI-2 works well on stable, mean-reverting instruments (SPY, QQQ, blue chips) and works terribly on trending momentum stocks (small caps, biotech, crypto). Testing on the wrong universe gives you false confidence — and losses in production.

The second mistake is skipping the trend filter. Raw RSI-2 buys every dip, including the dip that becomes a crash. Adding a simple trend filter (e.g., only buy when price is above the 200-day moving average) improves risk-adjusted returns dramatically. Our backtests show this filter alone can turn a losing strategy into a winning one.

Exit rules matter enormously. The classic RSI-2 exit is when RSI climbs above 70. But many variants exist: fixed holding period (5 days), price target (2% profit), or opposite RSI signal. Each produces different results. We've backtested them all — the best exits are typically fast (1-5 days) with a small profit target.

The most important thing about mean reversion: the strategy needs to lose sometimes in exchange for the wins. If your backtest shows 95% win rate with tiny wins, that's not a great strategy — it's a strategy waiting to blow up on the 5% that go wrong. Look for 60-75% win rate with balanced win/loss ratios. That's more sustainable.