The thesis
News-sentiment runs a simple, legible rule across a 24-name large-cap universe (AAPL, MSFT, NVDA, JPM, V, KO, PFE and friends): buy when recent news sentiment turns positive, exit when it turns negative. It's the kind of strategy that's easy to reason about and easy to audit — which is exactly why its activity record deserves a close look.
Recent activity
The strategy is live, but "live" currently means mostly idle. Its scheduled runs from 2 to 9 November executed almost nothing: a single trade on 3 November (21 shares of KO at $95.13) and zeros everywhere else. The only other executions on record are an opening batch on 29 October — V, NVDA and PFE. Since then the book has largely sat still.
That stillness isn't costless. Total account value has drifted from roughly $9,987 at the start of November down to $9,634 by the 9th, with cash parked at $2,387.18 throughout. In other words, the strategy is holding a handful of positions that have bled a little, and its sentiment signal hasn't fired strongly enough to rotate out of them.
Backtest and validation
The headline backtest looks benign and almost inert: +0.19% total return over 451 days, final equity $10,018.55, a 0.1% CAGR, Sharpe 0.74, and a trivially small 0.05% max drawdown. Turnover was 36.53%, fees just $2. The problem is the sample: two trades. A strategy that transacts twice in a year-and-a-half isn't really being tested — it's being observed doing nothing.
Cross-validation makes this concrete. Of four walk-forward folds, three produced zero trades and zero return; only fold 4 (Dec 2025–May 2026) did anything, posting that same +0.19% and a 1.5 out-of-sample Sharpe. So the flattering out-of-sample numbers — and the high PSR of 0.923 — all rest on one active window. The deflated Sharpe ratio (DSR 0.551, across six trials) is more honest, and validation correctly returns passed: false with only one of four folds positive.
One figure needs scrutiny: win rate is recorded as 0 despite a marginally positive return on two trades. That inconsistency suggests either an accounting quirk or an exit that booked losses while a holding carried the return — worth a data check before anyone leans on this strategy.
Verdict
The strength here is discipline: low turnover, negligible fees, tiny drawdown, and a thesis you can explain in a sentence. The risk is that there is almost no evidence it works. A signal that triggers in one fold out of four, on a two-trade sample, is untested rather than proven. Before promoting it, the sentiment threshold likely needs loosening so it actually fires — and the zero win rate needs explaining.