The Thesis
RSI Snap-Back operates on a straightforward premise: the Magnificent Seven — AAPL, MSFT, NVDA, GOOGL, AMZN, META, and TSLA — are too liquid and too closely watched to stay oversold for long. The strategy enters the most RSI-depressed names in that universe (RSI < 35) and exits once momentum recovers to overbought territory (RSI > 70). A hard cap of four simultaneous positions enforces concentration discipline and bounds drawdown exposure by design.
The logic is sound in principle. These names benefit from persistent analyst coverage, index rebalancing flows, and institutional accumulation on dips — all mechanical forces that tend to support snap-backs. The RSI thresholds are deliberately wide, filtering for genuine short-term extremes rather than routine fluctuations.
Backtest Performance
Across a 451-day backtest window, the strategy produced a 20.95% total return (CAGR ~11.2%), ending with $12,095 from a $10,000 starting book. The win rate of 66.7% across 37 trades is a meaningful edge — two-thirds of signals resolved in the strategy's favour.
The Sharpe ratio of 0.61 reflects modest risk-adjusted returns. Not exceptional, but not dismissible either, particularly for a rules-based system with no curve-fitting beyond two RSI thresholds. Total fees came to $37 (flat per trade), and there were no FX costs given the all-USD universe.
The figure that demands attention is maximum drawdown: 23.73%. For a strategy focused on blue-chip names with a tight book, that's a substantial peak-to-trough decline. It suggests that entering on RSI < 35 doesn't reliably catch the bottom — during prolonged tech sell-offs, positions can keep moving against the strategy before reverting. Turnover of 773% over the period confirms active rotation, which amplifies both the upside and the drawdown path.
Recent Activity: Patience or Paralysis?
The live paper-trading log for the past two weeks tells a uniform story: zero trades executed across every scheduled run from June 3 through June 10. The cash balance has sat at $10,000 throughout, fully uninvested.
This could reflect genuine discipline — if no Mag-7 name has touched RSI < 35 during a period of broad strength, the strategy is correctly standing aside. Mean-reversion systems are supposed to wait. The risk, however, is that prolonged inactivity in live mode diverges meaningfully from backtest assumptions about signal frequency. If the market regime has shifted — with large-cap tech grinding higher without sharp pullbacks — the RSI < 35 entry threshold may be triggered far less often going forward than it was over the backtest window.
Strengths and Risks
Strengths
- Clean, interpretable rules with no ambiguous discretionary inputs
- High win rate (66.7%) suggests the core signal has genuine predictive value
- Universe concentration in the most liquid equities limits slippage risk in live deployment
- Four-slot cap prevents runaway concentration in any single down-trending name
Risks
- 23.73% max drawdown is high relative to Sharpe; the strategy can endure extended pain before a snap-back materialises
- No validation metadata available yet — backtest and live metrics are currently the same dataset, so overfitting risk cannot be assessed independently
- A sustained low-volatility bull run in tech could leave the strategy idle for extended periods, with opportunity cost accumulating
- The strategy holds no explicit stop-loss; a name that enters a structural decline (not a cyclical dip) could breach the RSI < 35 threshold and keep falling
Outlook
RSI Snap-Back is a coherent, well-scoped strategy with a credible edge in the backtest record. The immediate priority should be independent out-of-sample validation to confirm the win rate and drawdown profile aren't artefacts of the test window. The quiet live stretch is not alarming on its own — but if no signals fire across the next several weeks, it's worth revisiting whether the entry threshold is calibrated for the current market regime.