The thesis
PanicFade v4 bets that negative headline shocks are overreacted to and revert toward fundamentals. The entry is deliberately narrow: a name must show sentiment below -0.4 and already be technically oversold — RSI14 under 35 and trading below its SMA20 — before the strategy sizes into the panic. Exits are symmetric and mechanical: it takes profit on reversion (RSI14 back above 50, or price reclaiming the SMA20) and hard-stops fresh lows on continued negativity, treating that as a fundamental break rather than noise.
The design leans on recent sentiment-feedback literature (arXiv:2509.11970, 2025, plus corroborating MDPI/JRFM work). Crucially, it fades price-defined panic in liquid large caps — the 24-name universe is all mega-cap (AAPL, MSFT, JPM, UNH, XOM and peers) — which keeps the contrarian bet away from the names most likely to be falling for real, structural reasons.
Recent activity
The live book is quiet. Across the six scheduled runs from 1–8 October, most days executed zero trades; only 1 and 7 October fired, with two executions each. Recent fills show the fade in action: accumulating BAC (23 shares at $53.40), trimming UNH, and rotating through MCD, NKE, HD and WMT. As of 8 October the paper account held $5,238.84 cash against a total of $9,836.38 — a book that has drifted sideways-to-slightly-soft in recent sessions despite the strong backtest.
Validation and performance
This is where PanicFade v4 earns attention. The backtest spans 1,233 days for a 37.93% total return (6.79% CAGR), Sharpe 0.96, 62.88% win rate over 530 trades, and a contained 9.3% max drawdown. More importantly, the walk-forward validation passed cleanly: 4 of 4 folds positive, with returns that rise over time (1.6% → 3.88% → 10.14% → 21.7%) and an out-of-sample Sharpe of 2.44 in the most recent fold. A Probabilistic Sharpe Ratio of 0.985 across 27 trials suggests the edge is unlikely to be pure luck.
The risks
Balance demands scepticism. First, the Deflated Sharpe Ratio sits at 0.56 — respectable once you account for 27 trials, but a reminder that multiple-testing shaved real confidence off the headline number. Second, turnover is enormous at 5,807%, so the strategy is only viable where frictions stay low; the backtest's $530 of fees on 530 trades flatters that assumption. Third, and most tellingly, the live paper book has not yet mirrored the backtest's compounding — a near-flat equity curve against a 37.93% historical return is exactly the gap every validated strategy must eventually close or explain.
Verdict
PanicFade v4 is one of the better-validated strategies on the platform: a clear, literature-grounded thesis, disciplined exits, and folds that improve rather than decay. The open question is whether that edge survives live frictions and a quiet tape. The backtest says yes; the October book says not yet. Worth watching closely.