Strategy Thesis
News-sentiment operates on a straightforward premise: accumulate positions when recent news flow turns positive for a ticker, and exit when sentiment flips negative. The universe spans 24 large-cap names across tech, financials, healthcare, consumer staples, and energy — a deliberately diversified watch-list designed to surface sentiment-driven moves wherever they emerge.
Recent Activity
Live paper-trading began in late May and has produced five executed buys so far: AAPL (6 shares @ $312.06 on May 31), MSFT (4 @ $428.23 on June 3), BAC (35 @ $55.93 on June 12), UNH (4 @ $402.85 on June 18), and GOOGL (5 @ $360.40 on July 1). No sells have been triggered, suggesting the strategy has not yet encountered a negative-sentiment reversal severe enough to meet its exit threshold.
The past week tells a quieter story. Daily scheduled runs from July 3–10 generated zero executions in five of six sessions, with one rejection on July 7. Portfolio equity has oscillated between roughly $9,957 and $10,073, consistent with passive mark-to-market drift rather than active positioning. Cash sits at $990 — about 10% of the paper portfolio — indicating the bulk of capital is already deployed.
Backtest & Validation Performance
Over 451 days of backtesting, the strategy produced a +0.19% total return and a +0.10% CAGR, with a Sharpe ratio of 0.74 and a maximum drawdown of just 5%. Only two trades were completed in this window, and the win-rate stat is currently uninformative at 0 (too few completed round-trips to score). Total fees were $2 with no FX costs, reflecting the USD-denominated universe.
The validation picture is mixed but worth unpacking carefully:
- Folds 1–3 (Aug 2024 – Dec 2025): Zero trades, zero return. The strategy was effectively dormant for roughly 16 months of the backtest window.
- Fold 4 (Dec 2025 – May 2026): 2 trades, +0.19% return, Sharpe 1.5. All of the strategy's positive performance is concentrated here.
- PSR (Probabilistic Sharpe Ratio): 0.923 — there is roughly a 92% probability the true Sharpe exceeds a benchmark of zero, which is meaningful for a short live history.
- DSR (Deflated Sharpe Ratio): 0.551 — once adjusted for the fact that six parameter combinations were trialed, confidence drops to about 55%, reflecting the multiple-testing penalty.
The validation gate has not been passed (passed: false), and with only 1 of 4 folds showing positive results, that verdict is appropriate.
Strengths
- Low drawdown: A 5% maximum drawdown is genuinely conservative for an equity strategy, suggesting the exit logic is working as intended when it fires.
- Strong recent-fold Sharpe: The 1.5 out-of-sample Sharpe in Fold 4 is a constructive signal if the signal environment that drove it persists.
- Diversified universe: 24 names across seven sectors reduces idiosyncratic blow-up risk.
Risks & Open Questions
- Signal sparsity is the dominant concern. Two trades in 15 months of backtest history is not enough to draw statistical conclusions. The strategy may be too selective, or the sentiment threshold may need recalibration to fire more consistently without sacrificing edge.
- Recency concentration: All measured return came from a single fold. Whether Fold 4 represents genuine alpha or a favorable macro window (e.g., a news-driven rally) is unknowable with this sample size.
- No exits yet in live trading: The strategy holds five positions with no realized closes. Until the exit logic executes on live data, the full round-trip behavior remains untested in production.
- DSR discount: The 0.551 DSR is an honest reminder that six trials were run. Any claimed edge needs more out-of-sample data before capital allocation warrants scaling.
Outlook
News-sentiment is a live experiment still in early innings. The theoretical logic is sound and the risk controls appear conservative, but the strategy needs materially more trade history before its performance can be meaningfully evaluated. The focus for the next review period should be on whether signal frequency increases and whether the exit mechanism triggers cleanly on a negative-sentiment event.