The thesis
News-sentiment runs a deliberately simple premise: buy names on positive recent news sentiment, exit on negative. It operates over a 24-name large-cap universe spanning tech (AAPL, MSFT, GOOGL, NVDA), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE, ABBV), and staples/industrials (PG, KO, WMT, COST, CAT, HON, XOM, CVX). It is currently flagged live, trading a paper book worth roughly $10,283 as of the latest scheduled run, with only about $990 in cash.
Sentiment-driven entry is intuitive and cheap to reason about, which is a genuine strength. But it lives or dies on signal quality and timing, and the record so far gives us very little to judge it on.
Recent activity
The last executed trade was a buy of 5 shares of GOOGL at $360.40 on 2026-07-01, preceded by buys in UNH, BAC, MSFT, and AAPL through June. Since then the strategy has gone quiet. Every scheduled run from 2026-09-08 through 2026-09-15 reported 0 executed orders, with a handful rejected each session (one to two candidates turned away per run). Portfolio total value drifted between roughly $10,277 and $10,377 across that window — motion driven by mark-to-market on existing holdings, not new decisions.
In other words: the model is holding, and its signal is not currently clearing whatever entry bar it sets.
Backtest and validation
The headline backtest covers 451 days and returns +0.19% (final equity $10,018.55, ~0.1% CAGR), with a Sharpe of 0.74, a max drawdown of ~0.05%, turnover of 36.5%, and just $2 in fees. The catch is the trade count: the entire result rests on 2 trades with a stated 0% win rate. That is not a track record; it is an anecdote with a Sharpe attached.
Walk-forward validation makes this concrete and did not pass. Of 4 folds, 3 executed zero trades — the strategy simply found nothing to do across most of 2024–2025 — and only 1 of 4 folds was positive. All of the return comes from fold 4 (2025-12 to 2026-05), which posted the same +0.19% and an out-of-sample Sharpe of 1.5 on those two trades. The probabilistic Sharpe (PSR 0.92) looks reassuring, but the deflated Sharpe (DSR 0.55), which accounts for the 6 trials run, is much more sober.
The verdict
Strengths: low turnover, negligible costs, a clean and explainable thesis, and no ugly drawdowns to date. Risks: the sample is dangerously small, three of four validation windows produced no signal at all, and the one profitable fold does all the lifting — a textbook setup for overfitting a single lucky window. Its recent silence may be honest discipline or a sign the signal rarely fires.
Until news-sentiment accumulates trades across multiple regimes, treat its metrics as provisional. It has earned a live slot to gather evidence, not yet our confidence.