Thesis
The news-sentiment strategy operates on a straightforward premise: buy when recent news sentiment for a ticker is positive, exit when it turns negative. The universe covers 24 large-cap U.S. names spanning technology, financials, healthcare, consumer staples, and energy — liquid names where sentiment signals have the best chance of being priced before they dissipate.
Deployment & Recent Activity
The strategy launched May 31, 2026 with a $10,000 paper allocation. Its opening move was a 6-share buy of AAPL at $312.06 (~$1,872 outlay), leaving roughly 81% of the book in cash. The two subsequent daily runs — June 1 and June 2 — generated no signals: 0 trades executed, 0 rejected. Portfolio NAV has drifted between $9,974 and $10,016 in that window, tracking almost entirely with the AAPL position.
The low trade frequency is itself a signal worth noting. With 24 tickers under coverage and daily sweeps, the strategy is either applying a tight sentiment threshold or finding very few qualifying setups in the current market environment.
Backtest Performance
Over 451 backtested days, the strategy produced:
| Metric | Value |
|---|---|
| Total return | 0.19% |
| CAGR | 0.10% |
| Sharpe ratio | 0.74 |
| Max drawdown | 0.05% |
| Trades | 2 |
| Win rate | — (no closed trades) |
The drawdown figure is impressively low, but the context undercuts it: two trades in 451 days means the portfolio spent the vast majority of the backtest period in cash. Capital preservation is easy when capital is rarely deployed. Annualised turnover of 36.5% looks active relative to the trade count, suggesting the position sizing is meaningful when it does fire — but the sample is too thin to read much into it.
Cross-Validation Verdict
Formal validation failed. The backtest was divided into four time-ordered folds:
- Folds 1–3 (Aug 2024 – Dec 2025): zero trades, zero return across all three periods.
- Fold 4 (Dec 2025 – May 2026): 2 trades, +0.19% return, Sharpe 1.5.
Only one of four folds was positive, and all of the strategy's lifetime return is concentrated in the most recent window. That alone would concern any risk framework — recency concentration is a classic overfitting marker.
The probabilistic Sharpe ratio (PSR) of 0.923 is encouraging: it implies a 92.3% probability that the true Sharpe is above zero. But the deflated Sharpe ratio (DSR) — which adjusts for the fact that 6 strategy variants were trialled — drops to 0.551. After correcting for multiple comparisons, the edge is roughly a coin flip.
Strengths
- Minimal drawdown: the strategy does not blow up; it simply stays flat or slightly positive.
- OOS Sharpe of 1.5: if fold 4 is representative of genuine regime fit, the risk-adjusted return is compelling.
- Clean risk score: the reviewer assessed risk at zero, consistent with the cautious sizing and liquid universe.
Risks
- Insufficient trade history: two trades do not constitute a statistically meaningful sample. Win rate cannot be evaluated; expected value is unknown.
- Single-fold concentration: all performance lives in one recent period. This could reflect a genuine sentiment-regime shift — or lucky timing.
- Signal sparsity: if the strategy continues generating zero signals daily, the question becomes whether the sentiment pipeline is calibrated appropriately or whether it is simply too conservative to be useful.
- DSR below 0.6: accounting for trial count, the probability of a real edge is not yet convincing enough to warrant significant live capital.
What to Watch
The next 30–60 days of paper trading will be more informative than the entire backtest. Watch for: whether the signal fires in a range of market conditions (not just low-volatility drift), whether exits execute cleanly, and whether the cash drag normalises as more names meet the sentiment threshold. A validation re-run after accumulating 10+ closed trades would provide a far cleaner read on edge reliability.