Supported · Timeframes · 2026-09-26 (15m) to 2026-09-29 (1d)
Does the unfiltered strategy make money on every timeframe, and which earns most per setup?
What this means
With every filter off, the average setup made money on all four timeframes, and the 95% ranges stay above zero on each. The edge per setup is thin and grows with timeframe: +0.08R on 15m, +0.19R on 1h, +0.23R on 4h and +0.27R on daily, with roughly half of setups losing a full 1R. Per stock per year the order reverses among the live timeframes: 1h has about six times as many setups as daily (2.5 vs 0.4 a year), so it earns about 0.48R a stock-year against 0.11R. These results use today's index members and zero slippage, so real-world figures would be lower.
Explanation checked against the data.
By timeframe
| Timeframe | Verdict | n | Result |
|---|---|---|---|
| 15-minute | Supported | 16,496 | +0.081R (+0.062, +0.101), 42.2% win |
| 1-hour | Supported | 13,208 | +0.193R (+0.170, +0.216), 46.6% win |
| 4-hour | Supported | 3,990 | +0.226R (+0.188, +0.264), 47.2% win |
| Daily | Supported | 2,098 | +0.273R (+0.220, +0.327), 49.3% win |
- Hypothesis
- Harmonic Master with all entry filters off has a positive expectancy on 15m, 1h, 4h and daily across a broad stock universe.
- Method
- TradingView Strategy Tester, driven over CDP; 503-symbol NasdaqAndSP watchlist (today's Nasdaq-100/S&P 500 + BTCUSDT); 15m 2023-08 to 2026-09, 1h/4h/1d 2016-01 to 2026-09; 2% risk sizing, 0.01% commission, 0 slippage; stop min(D,X)x0.99, 50% at 1.5R + runner to 2R (3R on monthly S3), BE after TP1. Setup R recomputed from per-tranche trade logs; 95% CI = bootstrap clustered by instrument.
- Result
- Mean setup R (95% CI), setup win rate, setups: 15m +0.081 (+0.062, +0.101), 42.2%, 16,496; 1h +0.193 (+0.170, +0.216), 46.6%, 13,208; 4h +0.226 (+0.188, +0.264), 47.2%, 3,990; 1d +0.273 (+0.220, +0.327), 49.3%, 2,098. PF (R) 1.14 / 1.37 / 1.43 / 1.54. Every one of the 7 patterns positive on 1h.
- How much to trust it
- High: real TradingView fills, large n, reconciled to TV counters; but survivorship universe and zero slippage
- Did it change the live rule?
- confirmed the baseline; live alerts run on 1h/4h/1d
- Notes
- No 1h config file was saved (same config inferred). Cypher was ON in all sweeps although the Pine default is OFF. Per setup, longer timeframes earn more; per instrument-year, 1h earns most because it has ~6x more setups than daily.
Scripts and data for this study will be linked here when the research code is published on GitHub.
Education and research only. Not financial advice. Backtests use today's index members (survivorship bias).