The thesis
PanicFade v4 bets on a well-documented market reflex: negative headlines get overreacted to, and prices tend to drift back toward fundamentals once the noise fades. The rule set is disciplined rather than clever. It buys names hit by sentiment below -0.4 that are also technically oversold — RSI14 under 35 and trading below their 20-day moving average — and sizes into the weakness. It exits on reversion (RSI14 back above 50, or price reclaiming the SMA20) and hard-stops on fresh lows under continued negativity, treating a persistent break as a fundamental signal rather than a buying opportunity.
The approach is grounded in recent sentiment-feedback literature (arXiv:2509.11970, 2025) on asymmetric overreaction and reversion. Crucially, the entry is price-confirmed: sentiment alone doesn't trigger a buy, which guards against the classic trap of catching a knife that keeps falling.
Validation performance
The numbers are solid without being spectacular. Across a 1,233-day backtest the strategy returned 37.93% (6.79% CAGR), with a 62.88% win rate over 530 trades, a Sharpe of 0.96, and a contained 9.3% max drawdown. Walk-forward validation is the stronger evidence: all four out-of-sample folds were positive, and the probabilistic Sharpe ratio sits at a reassuring 0.985.
Two caveats deserve equal billing. First, the deflated Sharpe ratio — which penalises for the 27 trials run — is only 0.56, barely above the coin-flip line. Second, the returns are back-loaded: fold 4 alone delivered 21.7% (Sharpe 2.44), while the earlier folds managed just 1.6% and 3.88%. The edge looks real, but a meaningful slice of it is recent and may reflect a favourable regime.
Recent activity
The live book is quiet. Total equity sits near $9,816 with roughly $8,642 in cash, meaning most of the portfolio is parked and waiting — several scheduled runs this week executed zero trades. That is behaviour consistent with the thesis: when no name is simultaneously panicked and oversold, the strategy sits out rather than forcing a position.
Recent trades skew toward exits. The book sold MCD, NKE, HD, and trimmed UNH and BAC through October — reversion and reclaimed moving averages doing their job — against earlier September entries into NKE, HD, and UNH. The MCD round-trip (bought at 231.79, sold at 242.65) is a textbook fade-and-revert.
Strengths and risks
Strengths: a large-cap-only universe keeps liquidity and fundamental-break risk low; price confirmation disciplines the sentiment signal; and every fold held up out of sample.
Risks: turnover is enormous at 5,807%, so fee drag is a structural headwind even at $1/trade. The DSR warns against over-crediting the backtest, and the current slightly-underwater, mostly-cash position is a reminder that the edge only shows up when panics appear. PanicFade v4 is a credible, well-constrained strategy — but one whose payoff is lumpy and regime-sensitive, not a steady compounder.