The thesis
PanicFade v4 is a contrarian bet that the market oversells bad news. The rule is deliberately narrow: enter a large cap only when headline sentiment drops below -0.4 and the name is already technically oversold (RSI14 under 35, trading below its SMA20), then add into the decline. Exits are equally mechanical — reversion back above RSI 50 or a reclaim of the SMA20 — with a hard stop on fresh lows that treats continued negativity as a genuine fundamental break rather than noise. The design leans on 2025 sentiment-feedback research (arXiv:2509.11970 and corroborating work in JRFM) arguing that negative-media overreaction reliably reverts toward fundamentals.
The universe is intentionally conservative: 24 blue chips across tech, financials, healthcare, staples and energy. This is panic-fading on names unlikely to actually go to zero, which is the right instinct for a strategy whose core risk is catching a falling knife.
Recent activity
The last two weeks have been mostly exits. The cleanest example is McDonald's — bought at $231.79 on 1 October and sold at $242.65 on the 15th, a textbook oversold-entry-to-reversion round trip. Not every trade cooperated: Nike (bought $36.10, sold $35.53) and Home Depot (bought $299.12, sold $293.95) were closed at small losses, likely stop-driven rather than reversion wins. The scheduled runs show discipline — many days execute zero trades and reject none, which is what you want from a strategy that only fires on a specific confluence.
Worth flagging honestly: the live paper book totals roughly $9,833 against what looks like a $10,000 start. The strategy is marginally underwater in live trading, even as its backtest shines.
Validation and performance
The numbers that matter are the walk-forward results, and they are strong. All four out-of-sample folds were positive, spanning the 2022 drawdown through 2026, with the most recent fold returning 21.7% at a 2.44 Sharpe. The deflated metrics hold up: a 0.985 probabilistic Sharpe ratio and a 0.56 deflated Sharpe across 27 trials suggest the edge is not purely a product of overfitting — a meaningful bar given how easy sentiment strategies are to curve-fit.
The full backtest returns 37.93% over 1,233 days, a 6.79% CAGR, 0.96 Sharpe, 62.88% win rate and a contained 9.3% max drawdown across 530 trades.
The risks
Two caveats temper the optimism. First, turnover is extreme at 5,807% — this strategy trades constantly, and in a real account frictions and slippage would bite far harder than the flat fee assumed here. Second, the live-versus-backtest gap deserves watching: a slightly negative live book is early and small-sample, but it is the only money-on-the-line evidence we have. PanicFade v4 is well-validated and well-reasoned; it now needs live results to confirm the paper promise.