The Thesis
RSI Snap-Back is a mean-reversion play on the Magnificent Seven. The premise is clean: large-cap tech names tend to rebound sharply after short-term momentum extremes, so the strategy buys the most oversold names (RSI below 35) and exits the overbought ones (RSI above 70). A hard four-slot book enforces discipline and caps concurrent drawdown exposure. It trades a tight universe — AAPL, MSFT, NVDA, GOOGL, AMZN, META, and TSLA — and rotates to keep that book lean.
As a design, it is coherent. Mean reversion is a well-documented effect in liquid large-caps, and the fixed-slot cap is a sensible structural guardrail against over-concentration.
Recent Activity
The last two executed trades were entries: 10 shares of NVDA at $204.04 on 2 November, and 7 shares of GOOGL at $280.28 on 30 October. Since then, the strategy has gone quiet. Its six most recent scheduled runs — from 5 through 12 November — each report 0 executed, 0 rejected, with cash parked at a steady $5,956.92.
This inactivity is a feature, not a bug: with no name hitting the RSI < 35 trigger, the strategy correctly sits on its hands rather than forcing trades. That said, the idle cash is notable. At roughly $5,957 in cash against portfolio totals hovering between $9,637 and $9,837, more than half the book is uninvested and only two of four slots are filled. The strategy is waiting patiently — but it is also currently running below its $10,000 starting line, down a few percent on the live mark.
Backtest and Validation
On paper, the numbers are respectable. Over 451 days the backtest returned 20.95% (11.21% CAGR), finishing at $12,095 on a win rate of 66.67% across 37 trades. Costs were negligible — $37 in fees, zero FX — and the two-in-three win rate suggests the entry signal has genuine edge.
The risks sit right next to the rewards. The Sharpe of 0.61 is modest, and a maximum drawdown of 23.73% is steep for a strategy whose whole pitch is limiting drawdown exposure; a quarter of equity can evaporate before the snap-back arrives. Turnover of 773% is also high, which in a live book with real spreads and slippage can quietly erode the thin fee advantage shown here.
Most importantly: validation is null. There is no out-of-sample or walk-forward result on record. A single in-sample backtest, however pretty, is the easiest number in quant to overfit — and with only 37 trades, the sample is thin enough that the win rate could shift meaningfully under different market regimes.
Verdict
RSI Snap-Back has a sound thesis, disciplined structure, and a live record of not over-trading. But the headline return rests on an unvalidated backtest, the drawdown undercuts the strategy's own risk narrative, and it is currently underwater with half its capital idle. Treat the 20.95% as a hypothesis, not a track record, until out-of-sample validation confirms the edge survives outside the data it was fit on.