Thesis
RSI Snap-Back is a mean-reversion play on the largest US tech names. The premise is straightforward: even the strongest large-caps overshoot in the short term, and momentum extremes tend to correct. The strategy buys the most oversold names when RSI falls below 35 and exits when RSI climbs above 70, rotating positions to hold a tight four-name book across a Mag-7 universe (AAPL, MSFT, NVDA, GOOGL, AMZN, META, TSLA). The hard four-slot cap is the discipline mechanism — it forces concentration into the best current setups while capping how many drawdowns can compound at once.
Backtest Performance
Over 451 days the strategy returned 20.95%, ending on $12,095 from a $10,000 base — a CAGR of 11.21%. The win rate is a healthy 66.67% across 37 trades, which fits the mean-reversion profile: frequent, modest wins punctuated by occasional losers. Costs are negligible ($37 in fees, no FX drag), so the edge is not being eaten by friction.
That said, the risk-adjusted picture is more sober. A Sharpe of 0.61 is unremarkable — the returns come with meaningful volatility, and you are not being paid a premium for the ride. Turnover of 773% confirms this is an active, rotation-heavy book rather than a buy-and-hold; the four-slot cap keeps it churning as names cross RSI thresholds.
The Drawdown Question
The headline risk is the 23.73% maximum drawdown — larger than the total return itself. Buying oversold Mag-7 names means occasionally catching a falling knife, and a concentrated four-name book offers little cushion when a genuine downtrend, rather than a snap-back, is underway. RSI cannot distinguish "temporarily oversold" from "the start of a real decline," and this is where the strategy pays the price.
Validation Gap
A critical caveat: there is no walk-forward validation on record (validation is null). Every figure above is in-sample backtest performance. Without out-of-sample confirmation, we cannot yet rule out that the RSI thresholds and universe were, however implicitly, fitted to the historical window. For a live strategy, that is the single biggest open item.
Recent Activity
The live book has been notably quiet. The last six scheduled runs — from 14 to 21 August — each executed zero trades and rejected zero, with cash and total equity pinned at $10,000 throughout. No names in the universe have hit the RSI < 35 entry trigger recently, so the strategy is sitting patiently in cash. That is arguably correct behaviour for a mean-reversion system in a range-bound or grinding-up tape, but it also means the live track record remains effectively unwritten.
Verdict
RSI Snap-Back has a coherent thesis, a respectable win rate, and low frictional cost. But the mediocre Sharpe, an outsized drawdown, and — above all — the absence of walk-forward validation mean this is a promising candidate rather than a proven performer. The recent flat runs are a reminder that the real test only begins when the market actually hands it an oversold setup to trade.