77%win rate
Best tested configuration by win rate
Win = a trade that banked profit on the way up under our exit rules — not held to the close. Backtested on recorded market data, June–July 2026; fills modeled at mid. Past performance is not indicative of future results.
The result above is backtested, not a live-capital track record. The separate live journal records each signal from the current signal-only stream as it is published. Usually 1–4 signals per trading day on SPX, averaging about 2.5 on the July 2026 signal-only stream. Some sessions produce none.
One definition, no close-only hindsight.
The 77% figure is the highest win rate among our tested configurations — 684 graded signals across 29 trading days of recorded market data, June 3 through July 16, 2026. The sample below is graded under our production exit rule, a different tested configuration from the 77% headline. Sample-row option entries on this page are modeled at the quoted ask — the price a buyer actually pays at the time of the signal; exits are modeled at mid.
A trade counts as a win only when the managed exit rules banked profit on the way up. A contract that later rose after the position was closed does not retroactively become a win. A large premium peak is not presented as the achieved result unless the exit rules actually captured it.
A systematic model reads the market all session. Every signal is graded before it fires. No discretionary picks.
Backtested results and the live signal-only journal are labeled separately throughout this page.
- Test period
- June–July 2026
- Sample fills
- Entries at ask · exits at mid
- Win definition
- Profit banked under the exit rules
- Close treatment
- Not assumed held to close
- Live-capital claim
- None
- Past performance
- Not indicative of future results
A chronological sample, losses included.
These rows come from one recent, continuous window of the backtested record — not a hand-picked winner list. Window law: the most recent contiguous window containing 8–14 graded signals, every signal included, in signal order, wins and losses. At 10:11:04 and 10:15:20 the model fired twice in the same second on the same contract — two internal setups agreeing. A subscriber would receive one alert; both fires stay in the graded record.
Window: July 14, 2026 · all 10 signals · backtested on recorded market data
| Date · time (ET) | Direction | SPX contract | Modeled entry | Result under exit rules | Outcome |
|---|---|---|---|---|---|
| Jul 14 · 10:03:32 ET | SHORT | SPX 7495 PUT | $5.80 | +37.1% | WIN — profit banked |
| Jul 14 · 10:11:04 ET | SHORT | SPX 7490 PUT | $5.50 | +24.5% | WIN — profit banked |
| Jul 14 · 10:11:04 ET | SHORT | SPX 7490 PUT | $5.50 | +24.5% | WIN — profit banked |
| Jul 14 · 10:15:20 ET | SHORT | SPX 7495 PUT | $6.20 | +28.2% | WIN — profit banked |
| Jul 14 · 10:15:20 ET | SHORT | SPX 7495 PUT | $6.20 | +28.2% | WIN — profit banked |
| Jul 14 · 10:16:06 ET | SHORT | SPX 7485 PUT | $6.10 | −99.6% | LOSS |
| Jul 14 · 10:16:17 ET | SHORT | SPX 7485 PUT | $7.60 | −99.7% | LOSS |
| Jul 14 · 10:28:20 ET | SHORT | SPX 7500 PUT | $5.90 | −99.6% | LOSS |
| Jul 14 · 10:29:06 ET | SHORT | SPX 7495 PUT | $5.00 | −99.5% | LOSS |
| Jul 14 · 11:51:13 ET | SHORT | SPX 7515 PUT | $4.20 | +32.1% | WIN — profit banked |
A small banked win remains a small win. It is not replaced by a larger premium peak that occurred later. A loss remains visible under the same rules used for every other row.
Losses belong in the record. They show where the signal failed under the same exit rules used to grade every win.
Empty hours stay empty.
The same window, laid out across the trading day. Signals sit at their signal times; the rest of the session is quiet because nothing met the bar. Zero-signal sessions are part of the record too.
Timeline of July 14, 2026: ten signals fired between 10:03 AM and 11:51 AM Eastern; no signals for the rest of the session.
A morning cluster, one late-morning signal, then four hours of nothing — the bar was not met again.
No signal was published. The model did not find a trade that met the quality bar.
See the signal. See what happened next.
Backtested performance is hypothetical and was not achieved in live accounts. Past performance is not indicative of future results. Sentinel Alpha LLC is not a registered investment adviser. See full disclaimer.