The thesis
RSI Snap-Back bets on a well-worn market reflex: large-cap tech names that sell off hard tend to rebound just as sharply. The rule set is refreshingly simple — buy when the 14-day RSI drops below 35, exit when it pushes above 70, and never hold more than four names at once. That hard 4-slot book is the strategy's main risk control, capping how much concurrent drawdown exposure it can accumulate. The universe is the Mag-7: AAPL, MSFT, NVDA, GOOGL, AMZN, META, and TSLA — liquid, heavily traded, and exactly the kind of momentum-extreme candidates the thesis needs.
Recent live activity
The live book has been quiet. Its two most recent executed trades were a buy of 7 GOOGL at $280.28 on 30 October and 10 NVDA at $204.04 on 2 November. Since then, six consecutive scheduled runs (4–11 November) have all logged 0 executed, 0 rejected — RSI simply hasn't printed an oversold signal worth acting on. Cash has sat unchanged at $5,956.92 throughout, meaning roughly 60% of the portfolio is uninvested and only two of the four slots are filled.
That patience hasn't been rewarded yet. Portfolio total value has drifted from $10,022.55 on 4 November down to $9,772.57 by 11 November — a modest loss over the week, driven entirely by mark-to-market on the two open positions rather than any trading activity.
Backtest and validation
Over 451 days and 37 trades, the backtest returned 20.95% (final equity $12,095.14, CAGR 11.21%), with a 66.67% win rate and negligible costs — $37 in fees and zero FX. Two-of-three trades closing green is genuinely strong for a reversion system.
But the risk ledger tempers the enthusiasm. The 23.73% max drawdown exceeds the entire total return, and a Sharpe of 0.61 confirms the ride was bumpy relative to the reward. Turnover of 773% over the period signals a lot of churn for the eventual gain. Most importantly, the validation field is null — there is no out-of-sample or walk-forward check on record. A high win rate on in-sample data alone is exactly where overfitting hides.
Strengths and risks
Strengths: a clear, falsifiable thesis; tight position limits; cheap execution; and a win rate that suggests the edge is real on historical data.
Risks: an unvalidated backtest, a drawdown larger than the return, and a live account that is currently cash-heavy, half-deployed, and slightly below its starting mark. The six idle runs could reflect admirable discipline — or a strategy waiting on signals that the current regime isn't generating.
Verdict
RSI Snap-Back is a tidy, honest mean-reversion engine with a plausible edge, but it is still in the show-me phase. Until an out-of-sample validation lands and the live book either deploys its idle cash or demonstrates the snap-back it's named for, treat the 20.95% backtest as a hypothesis, not a track record.