The thesis
RSI Snap-Back is a mean-reversion play on the seven largest US tech names — AAPL, MSFT, NVDA, GOOGL, AMZN, META, and TSLA. The logic is deliberately mechanical: buy the most oversold names when their RSI drops below 35, exit when RSI climbs above 70, and never hold more than four positions at once. That hard 4-slot book is the discipline mechanism — it caps how much concurrent reversion risk the strategy can take on at any moment and forces rotation rather than sprawl.
It is a clean, legible idea. Large caps that overshoot to the downside on short-term momentum often snap back, and constraining the book to four names keeps the strategy from averaging into an entire sector drawdown.
Recent activity
Here the picture is quiet — pointedly so. Across six scheduled runs between 29 September and 7 October, the strategy executed zero trades and rejected zero orders. Cash and total equity have both sat flat at $10,000 the entire window. No name in the universe has been oversold enough to trip the RSI < 35 entry, so the book is empty and the strategy is simply waiting.
That is arguably the system working as designed: a reversion strategy with no signal should do nothing rather than force a position. But it also means recent live performance is uninformative — there is no fresh evidence, good or bad, to update on.
Backtest and validation
The backtest covers 451 days and 37 trades, ending at $12,095 of equity — a 20.95% total return, or roughly 11.21% CAGR. The win rate is a healthy 66.67%, and fees are negligible at $37 total with no FX cost.
The risk side is where I'd slow down. The maximum drawdown is 23.73% — larger than the full-period return, which tells you the equity curve was not a smooth ride. A Sharpe of 0.61 is modest; the returns came with meaningful volatility, not quiet compounding. Turnover of 773% also signals an active, rotation-heavy book, which is fine on a zero-friction backtest but deserves scrutiny once real spreads and slippage enter the picture.
Most importantly, the validation field is null. There is no out-of-sample or holdout result reported here, so the 20.95% is in-sample performance only. For a rules-based strategy with tunable thresholds (35 and 70 are choices), that is the single biggest caveat: we cannot yet distinguish genuine edge from curve-fit.
Verdict
RSI Snap-Back is a coherent, disciplined idea with a respectable backtest win rate and trivial costs. But a drawdown exceeding its return, a sub-1 Sharpe, and — crucially — no validation data mean it should be read as a promising hypothesis, not a proven edge. Its current dormancy is a feature, not a flaw; the real test comes when a Mag-7 name finally sells off and the book has to act.