The Thesis
The news-sentiment strategy runs on a simple, intuitive premise: buy when recent news sentiment turns positive, exit when it turns negative. It hunts across a 24-name universe of U.S. large-caps spanning tech (AAPL, MSFT, GOOGL, NVDA), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE, ABBV), and consumer staples (PG, KO, WMT, COST). It's currently live as a paper-traded agent.
The logic is clean and easy to reason about. The problem, as the data shows, is that a clean thesis and a tested thesis are not the same thing.
Recent Activity
The strategy built its book in early summer, executing buys in AAPL (6 @ $312.06), MSFT (4 @ $428.23), BAC (35 @ $55.93), UNH (4 @ $402.85), and GOOGL (5 @ $360.40) between late May and early July. Since then it has gone quiet.
The scheduled runs from Aug 13–20 tell the story: mostly 0 executed, with 1–2 orders rejected on several days. Cash has sat pinned at $990.06 — the book is nearly fully invested, so new signals have little room to act. Total account value drifted between roughly $10,239 and $10,403 over that window, so the agent is modestly in the green on a ~$10,000 base, but it's coasting on existing holdings rather than trading its edge.
Backtest & Validation
Here is where enthusiasm should cool. Over a 451-day backtest the strategy took just 2 trades, returning 0.19% (CAGR ~0.1%) with a Sharpe of 0.74 and a tiny max drawdown of 0.05%. Turnover was 36.5% and fees a trivial $2. A near-flat curve with two trades isn't a signal — it's an absence of one. Reported win rate was 0.
Walk-forward validation did not pass. Of 4 folds, only 1 was positive — and folds 1 through 3 recorded zero trades apiece. All activity landed in fold 4 (Dec 2025 → May 2026): 2 trades, 0.19% return, Sharpe 1.5. The deflated Sharpe ratio of 0.551 (across 6 trials) and probabilistic Sharpe of 0.923 look supportive on paper, but they're computed over a sample so sparse it can't carry weight.
The Read
Strengths: the thesis is legible, the drawdown is minimal, and the live book is currently profitable. There's nothing broken here.
Risks: the strategy barely trades. Three of four validation folds produced no signals at all, which points to a sentiment trigger that fires too rarely to be evaluated — or to sit fully invested and stall, as the recent rejected runs suggest. With only two trades in the backtest, every performance number is statistical noise, and the failed validation is the honest verdict.
This is a strategy to watch, not fund. It needs either a more sensitive signal or a longer live track record before its edge — if any — can be distinguished from a flat line. Right now, the most accurate description is untested.