The Thesis
PanicFade v4 is a contrarian mean-reversion strategy built on a well-documented market asymmetry: negative headline shocks tend to be overreacted to and then revert toward fundamentals. The rules are refreshingly concrete. It buys names hit by sentiment below -0.4 that are already technically oversold — RSI14 under 35 and trading below their SMA20 — and sizes into the panic rather than flinching from it. Exits are equally mechanical: reversion (RSI14 back above 50, or price reclaiming the SMA20) takes profit, while a hard stop on fresh lows amid continued negativity treats persistent selling as a genuine fundamental break rather than a fade opportunity.
The approach is grounded in recent literature (arXiv:2509.11970, 2025, corroborated by JRFM/MDPI work on stock-level investor sentiment), which lends the design more credibility than a purely curve-fit rule set. The universe is a sober 24-name roster of mega-cap liquidity — AAPL, MSFT, NVDA, JPM, JNJ, WMT and peers — where mean reversion is more reliable and slippage is manageable.
Validation
This is where PanicFade earns attention. The strategy passed a four-fold walk-forward test with all four folds positive, and — importantly — performance improved across the sequence: 1.6%, 3.9%, 10.1%, and 21.7% per fold. The most recent out-of-sample window (May 2025–Aug 2026) posted a 21.7% return at a 2.44 Sharpe with only a 2.6% max drawdown. Headline robustness metrics are strong: a Probabilistic Sharpe Ratio of 0.985 and 27 trials logged for multiple-testing honesty.
The Risks
Balance is warranted. The full-sample Sharpe is a middling 0.96, and the backtest CAGR is just 6.79% over 1,233 days despite a 37.9% cumulative return — respectable, not spectacular. The Deflated Sharpe Ratio of 0.56 is the number that keeps me honest: after adjusting for those 27 trials, confidence in genuine skill is only moderate. Turnover is extreme at 5,807%, meaning the edge must survive real-world frictions the backtest may underweight, even if modelled fees were a token $530.
Live Activity
Right now, PanicFade is waiting. The last six scheduled runs (Sep 1–8) executed zero trades and rejected zero — no names cleared the panic-plus-oversold bar. The only recent fill was a 10-share WMT buy at $103.12 on Aug 28. Live equity sits near flat at roughly $10,020 against $8,958 in cash, so the book is barely deployed.
That idleness cuts both ways. It reflects genuine discipline — the strategy refuses to manufacture signals in a calm tape — but it also underscores regime dependence: PanicFade needs volatility and negative shocks to work, and a quiet market leaves it earning nothing. It is a well-validated specialist, not an all-weather engine, and should be judged as such.