The Thesis
RSI Snap-Back bets on a well-worn phenomenon in large-cap tech: momentum extremes don't last. When one of the Magnificent Seven — AAPL, MSFT, NVDA, GOOGL, AMZN, META, or TSLA — drops its 14-day RSI below 35, the strategy enters a position. It exits when RSI climbs above 70. A hard cap of four simultaneous positions enforces concentration discipline and limits how much concurrent drawdown the book can absorb at once.
The logic is intuitive. These are the most liquid, most-analyzed equities on the planet. Institutional desks actively fade extreme moves, and index rebalancing creates structural buying pressure on dips. Mean reversion in this universe isn't a contrarian bet so much as a structural one.
Backtest Performance
Over 451 days of backtesting, RSI Snap-Back compounded a starting $10,000 into $12,095 — a 20.95% total return and an annualized CAGR of 11.21%. The win rate sits at a respectable 66.67% across 37 trades, meaning two in three entries resolved in the strategy's favor.
The Sharpe ratio of 0.61 is honest rather than exciting. It reflects a strategy that earns its return but not without turbulence — the 23.73% maximum drawdown is the number that demands attention. At nearly a quarter of the portfolio, a bad stretch in tech (a sector-wide derating, a macro shock) could feel painful even if the long-run edge holds.
Turnover of 773% over the backtest period signals frequent rotation. Each trade incurred a flat $1 fee; total friction was $37 across 37 trades — negligible at this scale, but worth monitoring if position sizes grow.
Live Activity: Quiet Market, Patient Strategy
Since going live, RSI Snap-Back has run daily scheduled checks and executed zero trades across every session logged from June 16 through June 23. Cash remains fully deployed in reserve at $10,000.
This is not a malfunction — it is the strategy behaving exactly as designed. RSI < 35 on any Mag-7 name requires a meaningful sell-off, and in a rangebound or trending-upward market, those signals simply don't fire. A system that trades only when its conditions are met is preferable to one that forces activity.
That said, six consecutive no-trade days do surface a real question: how frequently does this edge appear in live markets? With only 37 trades across 451 backtest days, the average holding generates roughly one signal every 12 days. Extended quiet periods should be expected — and tolerated.
Strengths
- Clear, falsifiable rules. Entry and exit are fully mechanical; there is no discretion to second-guess.
- Concentrated universe. Limiting to seven names means the strategy knows its names deeply and avoids liquidity risk.
- Positive win rate. 66.67% is solid for a mean-reversion approach, where losers can be large if a name continues lower.
Risks to Watch
- Drawdown depth. A 23.73% max drawdown against a universe of mega-caps suggests the strategy can get caught in sustained downtrends — RSI can stay oversold for weeks when fundamentals shift.
- No formal validation. The
validationfield is currently null. Until out-of-sample or walk-forward results are available, the backtest return should be treated as an upper bound, not a forecast. - Trade sparsity. Thirty-seven trades is a thin sample for statistical confidence in the win rate. A few bad trades could shift the picture meaningfully.
Bottom Line
RSI Snap-Back is a disciplined, rules-driven system with a credible thesis and a clean backtest. Its current silence reflects market conditions, not strategy failure. The priority for the next evaluation cycle should be formal out-of-sample validation — that will determine whether the 11.21% CAGR is repeatable or a product of a favorable backtest window.