The thesis
PanicFade v4 bets on a well-documented market reflex: negative headline shocks tend to be overreacted to, and prices drift back toward fundamentals once the noise clears. The rules are deliberately narrow. It only buys names where sentiment drops below -0.4 and the stock is already technically oversold — RSI14 under 35 and trading below its 20-day moving average — then sizes into the fear. Exits are symmetric and mechanical: take the reversion when RSI14 reclaims 50 or price recovers the SMA20, and hard-stop on fresh lows under continued negativity, treating that as a genuine fundamental break rather than a panic.
The approach is grounded in 2025 sentiment-feedback literature (arXiv:2509.11970 and a corroborating JRFM study), not folklore. The universe is a conservative 24 mega-cap names — AAPL, MSFT, JPM, UNH, XOM and peers — which keeps liquidity high and limits the chance that "panic" is really a distressed small-cap unraveling.
Recent activity
The strategy has been quiet, which is consistent with its design — it waits for panic rather than manufacturing trades. Most scheduled runs this week executed nothing (0 trades on 16–22 October). The exception was 23 October, when it added CAT (1 share @ $688.79) and GOOGL (3 @ $321.67). Earlier in the month it trimmed winners and stops — selling MCD, NKE, HD and UNH — while accumulating BAC around $53. That cadence, buying into weakness and rotating out on recovery, matches the stated playbook.
One honest note: the live paper account currently sits at a total of roughly $9,799 against cash of $6,978 — marginally below its $10,000 starting line. Live results are early and should not be confused with the backtest.
Backtest and validation
Over 1,233 days the backtest returned 37.93% (6.79% CAGR) with a 62.88% win rate across 530 trades and a contained 9.3% max drawdown. The headline weakness is a Sharpe of 0.96 — respectable but below the psychologically important 1.0, and the CAGR is modest for an actively traded book.
The validation is where PanicFade earns its keep. All four walk-forward folds were positive, and — notably — performance improved out of sample, from a barely-there 1.6% in the 2021–22 bear fold to 21.7% (Sharpe 2.44) in the most recent window. The Probabilistic Sharpe Ratio of 0.985 and a Deflated Sharpe of 0.56 across 27 trials suggest the edge is unlikely to be pure overfitting.
Strengths and risks
Strengths: clean out-of-sample behaviour, disciplined mechanical exits, and a liquid, defensible universe. Risks: turnover is extreme at 5,807%, the standalone Sharpe is sub-1, the DSR of 0.56 is only moderate confidence, and the thin 2022-fold return warns that this strategy struggles when panics don't revert. Promising and well-validated — but not yet proven live.