The thesis
RSI Snap-Back is a mean-reversion strategy that trades the seven largest-cap tech names — AAPL, MSFT, NVDA, GOOGL, AMZN, META, and TSLA. The logic is textbook contrarian: large-cap tech tends to overshoot on short-term momentum extremes, so the strategy buys the most oversold names when RSI falls below 35 and exits when RSI climbs above 70, aiming to pocket the snap-back. A hard four-slot book caps how many positions it can hold at once, which enforces discipline and limits concurrent drawdown exposure. It is a clean, explainable idea with a clear entry and exit rule — no black boxes.
Recent activity
Here is the honest headline: the strategy is currently doing nothing. Across six scheduled runs between October 2 and October 9, every single session reported 0 executed, 0 rejected, with cash and total equity both parked at exactly $10,000. There are no recent trades on record at all. That is not a malfunction — it is the rule working as designed. None of the seven names has dipped to an RSI below 35, so there is simply nothing oversold enough to buy. The flat cash position is a feature of the discipline, but it also means the live track record is, for now, a straight line.
Backtest and validation
Over a 451-day backtest the strategy returned 20.95%, lifting a $10,000 book to $12,095.14, for a CAGR of 11.21%. The win rate is a healthy 66.67% across 37 trades, and trading costs were negligible — $37 in fees and zero FX cost. Those are the strengths.
The risks sit right next to them. The Sharpe ratio is only 0.61, which tells us the returns came with meaningful volatility rather than a smooth ride. The maximum drawdown of 23.73% is steep for a $10k book and nearly swallows a full year's worth of gains at the stated CAGR. Turnover of 773% is very high — the strategy churns its capital roughly eight times over — and while fees stayed low in this test, that pace leaves it exposed to slippage and cost drag in live conditions. Most important, the validation field is null: there is no out-of-sample or walk-forward check here. A 20.95% return drawn purely from an in-sample backtest should be read as a hypothesis, not a proven edge.
Verdict
RSI Snap-Back is a sensible, well-constrained idea that backtests respectably and keeps its rules simple enough to trust. But two things temper the enthusiasm. First, the headline return rests on a single unvalidated backtest with a chunky drawdown and a mediocre risk-adjusted score — exactly the profile where overfitting hides. Second, the live book has yet to take a single position, so we have no real-world evidence either way. The patient, cash-heavy posture is the right behaviour for a mean-reverter waiting for genuine extremes; the task now is to see whether a real oversold signal, when it finally arrives, delivers the snap-back the backtest promises.