The thesis
SilentDip rests on a tidy empirical asymmetry: a sharp price drop without a salient negative headline tends to be noise that reverts, while the same drop carrying genuine bad news keeps drifting lower. The strategy therefore demands two conditions before it acts — oversold technicals (a stretched RSI, price below SMA20) and the absence of meaningful negative sentiment. It exits as names mean-revert toward SMA20 or as RSI normalizes.
The idea is not folklore. It extends Chan's 2003 work on price reactions to news versus no-news, by way of a November 2025 SSRN working paper (Clinch, He, Landsman & Li) on the temporal shift from drift to reversal. Grounding a signal in peer-reviewed microstructure research is a point in its favour.
Validation
This is where SilentDip looks strongest. Across four walk-forward folds spanning September 2021 to October 2026, all four were positive — including the brutal 2021–2022 drawdown window (fold 1, +4.96%) and the choppy 2022–2024 stretch (fold 2, +1.91%). The more recent folds are where it shines: +26.26% (Sharpe 1.84) and the latest out-of-sample +16.75% at a Sharpe of 1.91, with a trivial 2.39% max drawdown.
The statistical hygiene is credible too. A Probabilistic Sharpe Ratio of 0.991 says the positive Sharpe is almost certainly real and not a fluke of the sample. Full-period numbers — 52.62% total return, 70.55% win rate across 294 trades, a 10.38% worst drawdown — paint a steady, high-hit-rate profile rather than a lottery.
The risks
Two cautions deserve equal billing. First, the Deflated Sharpe Ratio is only 0.563 against 42 trials. After penalising for how many variants were tested, confidence in genuine edge drops from near-certain (PSR) to roughly a coin-flip-plus. That gap is the honest cost of search. Second, the headline 52.62% translates to a CAGR of just 8.69% over 1,279 days — respectable, not spectacular — earned with eye-watering 4,192% turnover. The approach is fee- and slippage-sensitive by construction, even if modelled fees stayed low here.
There is also a structural vulnerability: any no-news mean-reversion system is implicitly short tail risk. It buys falling knives it believes are blunt, and a mislabelled 'quiet' dip that keeps falling is exactly its failure mode.
Live so far
The live record is brief and, candidly, quiet. The only recent execution was a 2-share CAT buy at $622.54 on 30 October. Every scheduled run since — 3 through 10 November — reported 0 executed, 0 rejected, with cash parked at $8,740.24 and total equity drifting between roughly $9,859 and $9,984. On a $10k Season 2 base, that is marginally underwater and mostly idle.
That inactivity is arguably the strategy behaving correctly: no qualifying setups means no trades. But it also means the live sample is far too thin to confirm the backtest. SilentDip has earned a look through strong, well-validated research — now it needs qualifying dips to actually show up.