The thesis
PanicFade v4 is a contrarian mean-reversion strategy with a specific trigger: it buys large caps that have been hit by a negative headline shock (sentiment below -0.4) and are already technically oversold — RSI14 under 35 and trading below their 20-day moving average. The wager is that negative media sentiment is systematically overreacted to and reverts toward fundamentals, so the strategy sizes into the panic rather than away from it. It exits on reversion (RSI14 above 50 or a reclaimed SMA20) and hard-stops on fresh lows under continued negativity, treating that as a genuine fundamental break rather than noise.
The approach is grounded in recent sentiment-feedback literature (arXiv:2509.11970, 2025, plus corroborating work in JRFM/MDPI), which is a point in its favour: the edge has a documented rationale, not just a curve fit. The universe is 24 liquid mega-caps across tech, financials, healthcare, staples and industrials — names where a reversion thesis is most defensible.
Recent activity
The strategy reset to $10,000 on 2026-10-29 (Season 2) and has traded steadily since: three executions on 30 Oct and 2 Nov, two on 3 Nov, one on 4 Nov, with zero rejections throughout. Equity has drifted down modestly — from $10,000 to $9,871.17 by 4 Nov, roughly -1.3% in its first live week.
The trade log shows the whipsaw risk up close. CAT was bought at 622.54, 613.49-level churn, and ultimately sold at 585.79 — a losing round-trip. GOOGL was bought at 280.28, sold at 282.94, then re-bought at 278.48. NVDA (204.04) and a 20-share BAC position at 50.04 round out recent entries. This is a high-frequency fader, and short holding periods mean individual names can bleed before reverting.
Backtest and validation
Over 1,233 days the backtest returned 37.93% (6.79% CAGR, final equity $13,793), with a 62.88% win rate across 530 trades, a 0.96 Sharpe and a contained 9.3% max drawdown. The headline strength is the validation: four walk-forward folds, all four positive (1.6%, 3.88%, 10.14%, 21.7%), with out-of-sample return of 21.7% at a 2.44 Sharpe. The probabilistic Sharpe ratio is 0.985 and the deflated Sharpe — which penalises the 27 trials run — is 0.56. A reviewer approved it at risk 0.35.
Strengths and risks
The strengths are real: consistent sign across folds, a defensible literature-backed thesis, and a deflated Sharpe that survives the multiple-testing haircut. The risks deserve equal weight. Turnover is extreme at 5,807%, so results are sensitive to fees and slippage; the 6.79% CAGR is modest for that activity. The full-sample 0.96 Sharpe is average — the flattering 2.44 comes from the most recent fold, so recency may be doing heavy lifting. And the live tape, while young, is slightly red. Promising, validated, but unproven in production.