Inconclusive · Exits · 2026-09-26; timeframe gap fill 2026-10-11
Are the 1.5R / 2R targets too far? Does breakeven help?
What this means
Moving the stop to breakeven after TP1 gives no measurable edge on this 15m test of 30 stocks: mean setup R was +0.107 with breakeven and +0.099 without, well within one standard error. Matching the same setups in both runs, breakeven saved 61 runners from a full −1R stop but scratched 28 runners at 0R that would otherwise have gone on to TP2 or TP3, so the gain (+61R on runners) and the cost (−57R) almost cancel. Changing the targets did no better: closer targets raise the win rate (59% at 0.75R/1.25R) but lower the average (+0.06R), and wider ones do the reverse. The breakeven choice is therefore a matter of preference about variance rather than an edge. Matching the same setups target by target on 1h and daily shows what the small 15m sample could not: closer targets cost real expectancy, about 0.03R for 1.25/1.75, 0.07-0.08R for 1.0/1.5 and 0.10-0.12R for 0.75/1.25, even though they win more often. Further targets (2.0/3.0) added about 0.05R on daily only. Breakeven still makes no difference. These are simulations on our own price data, not TradingView fills, and 1h/4h cover only 2023-26.
Explanation checked against the data.
By timeframe
| Timeframe | Verdict | n | Result |
|---|---|---|---|
| 15-minute | Not supported | 997 | all target pairs and BE off within ~1 SE of +0.107R |
| 1-hour | Supported | 3,688 | simulated, paired vs 1.5/2.0: 1.0/1.5 -0.066R, 0.75/1.25 -0.095, 1.25/1.75 -0.031 (all q<0.001); 2.0/3.0 +0.019 (q=0.20); single TP2 +0.009 (q=0.12) gap fill G10, 2026-10-11 (price simulation; 1h/4h prices cover Nov 2023-Sep 2026 only) |
| 4-hour | Inconclusive | 1,073 | simulated: 0.75/1.25 -0.049 (q=0.11), 1.0/1.5 -0.033 (q=0.12), 2.0/3.0 +0.008 (q=0.81) gap fill G10, 2026-10-11 (price simulation; 1h/4h prices cover Nov 2023-Sep 2026 only) |
| Daily | Supported | 2,041 | simulated: 1.0/1.5 -0.079R, 0.75/1.25 -0.116, 1.25/1.75 -0.037 (q<0.001); 2.0/3.0 +0.047 (q=0.016); single TP2 +0.019 (q=0.007) gap fill G10, 2026-10-11 (price data) |
- Hypothesis
- Closer or farther targets, or turning breakeven off, change expectancy.
- Method
- TradingView 15m, 30 instruments (2023-08 to 2026-09), one factor at a time via the study API; baseline reproduces the main sweep.
- Result
- Mean setup R (SE): baseline 1.5/2.0 +0.107 (0.041), 43.8% win; 1.0/1.5 +0.099 (0.032), 53.6%; 0.75/1.25 +0.060 (0.028), 59.1%; 1.25/1.75 +0.099 (0.037); 2.0/3.0 +0.129 (0.050), 37.4%; BE off +0.099 (0.042); single TP1 (19 instruments) +0.041 (0.049). Gap fill (2026-10-11): price simulation of the same configs on TradingView setups, paired with a simulated 1.5/2.0 baseline that matches TradingView's outcome on 95-96% of setups (1h/4h Nov 2023-Sep 2026, daily 2016-26). Difference vs baseline: 1h 1.0/1.5 -0.066R, 0.75/1.25 -0.095, 1.25/1.75 -0.031 (each q < 0.001), 2.0/3.0 +0.019 (q = 0.20); daily -0.079, -0.116, -0.037 (q < 0.001) and 2.0/3.0 +0.047 (q = 0.016); 4h same direction, none significant (0.75/1.25 -0.049, q = 0.11). Single TP2 vs dual: +0.009 (1h), +0.006 (4h), +0.019 (daily, q = 0.007). Breakeven off: +0.002 / +0.001 / -0.001.
- How much to trust it
- Medium: 15m on 30 instruments with real fills; 1h/4h/1d from a price simulation that reproduces TradingView's outcome on 95-96% of setups (1h/4h only 2023-26)
- Did it change the live rule?
- No.
- Notes
- Closer targets trade expectancy for win rate; all differences within ~1 SE. Gap-fill script gaps/scripts/b_exits.py (G10); a full-history TradingView run of these configs on 1h/4h is listed in gaps/tradingview_runs.md.
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).