The Thesis
RSI Snap-Back is a classic contrarian bet dressed in modern tech clothing. It trades only the Mag-7 — AAPL, MSFT, NVDA, GOOGL, AMZN, META, and TSLA — on the premise that large-cap tech names overshoot in the short term and revert. The rules are refreshingly simple: buy when the 14-period RSI drops below 35, exit when it climbs above 70, and never hold more than four names at once. That hard 4-slot book is the strategy's spine, forcing rotation into the most oversold candidates and capping how much concurrent drawdown the portfolio can absorb.
There is intellectual honesty in the constraint. Mean-reversion strategies fail most spectacularly when they average down into a name that keeps falling; a bounded book is a crude but effective brake on that failure mode.
Recent Activity: Patiently Flat
The live tape tells a quiet story. Across six scheduled runs between 25 September and 5 October, the strategy executed zero trades and rejected zero orders, sitting on a flat $10,000 of cash each session. No name in the universe dipped below the RSI 35 entry threshold — the Mag-7 simply hasn't been oversold enough to trigger an entry.
This is the strategy working as designed, not a malfunction. A mean-reversion system should do nothing in a grinding or rising tape. But it also means the model has contributed no fresh P&L in recent weeks, and an investor paying attention sees a strategy whose edge is entirely theoretical until the next dislocation arrives.
Backtest and Validation
Over 451 days the backtest returned 20.95% (final equity $12,095), an 11.21% CAGR, with a 66.67% win rate across 37 trades. The win rate is the headline strength — two-thirds of trades closed profitably, consistent with a well-tuned reversion signal. Fees were negligible at $37 total and FX cost was zero.
The risks are equally plain. The Sharpe ratio of 0.61 is modest — this is not a smooth ride for the return it delivers. The 23.73% maximum drawdown is steep for a strategy pitching discipline and controlled exposure; nearly a quarter of capital was underwater at the worst point. Turnover of 773% signals heavy churn, and while low fees cushion that today, it leaves the strategy sensitive to any rise in trading costs or slippage.
Most importantly, the validation field is null. There is no out-of-sample or walk-forward check on record. A 20.95% in-sample return with a 66.67% win rate is exactly the kind of result that can flatter an over-fit rule set, and on a universe of just seven names the risk of curve-fitting to a few lucky reversals is real.
Verdict
RSI Snap-Back is a coherent, well-disciplined idea with a respectable backtest and a sensible risk cap. But the modest Sharpe, the deep drawdown, and above all the absence of any validation pass mean it should be read as promising rather than proven. Until it trades through an out-of-sample window — and actually fires in live conditions — the jury stays out.