The thesis
PanicFade v4 bets that negative headline shocks are, on average, overreactions. When a large cap gets hit by sentiment below -0.4 and is already technically oversold — RSI14 under 35 and trading below its SMA20 — the strategy sizes into the fear rather than away from it. Exits are equally mechanical: it takes reversion (RSI14 back above 50 or price reclaiming the SMA20) and hard-stops fresh lows on continued negativity, treating a persistent breakdown as a genuine fundamental break rather than a fadeable panic.
The design is grounded in recent literature on sentiment feedback and negative-media overreaction (arXiv:2509.11970, 2025; corroborated by JRFM/MDPI's stock-level investor-sentiment work). It runs on a defensive-tilted universe of 24 mega-caps — the AAPL/MSFT/JPM/JNJ/PG kind of names — with live-quote-only cash accounting and sanity-banded stops to keep the logic honest.
What the backtest says
Over 1,233 days the strategy returned 37.93% (final equity $13,793 on $10k), a 6.79% CAGR, with a Sharpe of 0.96 and a contained 9.3% max drawdown. Win rate is a healthy 62.88% across 530 trades — consistent with a mean-reversion profile that wins often but modestly. The standout caveat is turnover: 5,807%. This is a high-churn engine, and while modelled fees here are trivial ($530 total, no FX cost), real-world slippage on that much rotation deserves scrutiny before anyone extrapolates the paper Sharpe.
Validation holds up
The walk-forward evidence is the strongest part of the story. Across 4 folds spanning 2021–2026, all 4 were positive, and performance actually improved out-of-sample: 21.7% OOS return at a 2.44 OOS Sharpe in the most recent fold (2025-05 to 2026-08). The deflated statistics are reassuring given 27 trials were run — PSR 0.985 and DSR 0.56 suggest the edge survives multiple-testing correction rather than being a lucky draw. Fold 1, covering the 2022 bear, was the weakest (+1.6%, Sharpe 0.19), which is exactly where a buy-the-dip strategy should struggle — an honest sign the model isn't curve-fit to easy regimes.
The catch: it's not trading
Here's the balance. Over the last six scheduled runs (2026-08-20 through 08-27), PanicFade v4 executed zero trades every single day — cash and total portfolio both parked at a flat $10,000. That isn't a malfunction; it's the strategy's selectivity. Its entry gate (deep sentiment shock plus oversold plus sub-SMA20) is deliberately narrow, so in a calm, grinding-higher tape it simply has nothing to fade. The upside is discipline — no forcing trades. The downside is dependency: this is a strategy that earns its keep only when markets panic, and it will sit in cash, contributing nothing, through quiet stretches like the current one.
Verdict
PanicFade v4 is one of the better-validated strategies in the stable — real out-of-sample robustness, sensible risk controls, and a thesis with academic footing. The risks are equally clear: heavy turnover that could erode the live edge, and a signal so selective it can go idle for weeks. It's a specialist tool, not an all-weather one. Judge it by how it behaves in the next drawdown, not by a flat August.