Stop at 11:00, or skip the noon exit?

No validated edgeNo improvement established

Research completed Oct 9, 2026 · Published Oct 15, 2026 · Retrospective test on previously studied history

The idea.

Same gold rule, two timing changes, alone and together: stop taking new signals at 11:00 New York, or stop forcing the exit at noon. Each version was compared with the rule it came from, after costs.

EARLY rejects signals that close at or after 11:00 New York and keeps the noon (12:00) exit. HOLD keeps every signal but drops the noon exit, so a position runs until its stop, its target or the 16-bar cap. COMBINED does both. Parents: Wave 3's COST and BOTH rules, unchanged, with stop and target at 1R and 2R from the signal close.

The test.

Gold CFD · Dukascopy gold bid prices, 15-minute candles, January 2018 to September 2026; paired test January 2023 to September 2026 · Assumed round-trip cost: gold 0.40 points; stress test at 2×.

The result.

Six timing changes against their own parents.

Each row is a timing version minus its parent: the mean daily difference in summed R over 967 observed New York dates, Jan 2023 to Sep 2026, days without trades included. Dot: base cost. Ring: double costs. Bar: descriptive 95% block CI at base cost. Every row is below zero at base cost, and every interval includes zero.

What each timing change did to the BOTH rule.

Stop at 11:00.

BOTH + stop at 11:00 minus BOTH · summed R

Mean R per trade: +0.179 → +0.203

Jan 2023 to Sep 2026, base cost. It removed 43 late trades that had summed +4.498 R: 25 winners (+14.289 R) and 18 losers (−9.791 R). The trades it kept were unchanged; the total fell from +31.269 to +26.771 R.

No noon exit.

BOTH + no noon exit minus BOTH · summed R

Mean R per trade: +0.179 → +0.148

Jan 2023 to Sep 2026, base cost. Same 175 trades; 71 of them now ran past noon to the stop, the target or the bar cap. The total moved by −5.455 R.

Both changes.

BOTH + both changes minus BOTH · summed R

Mean R per trade: +0.179 → +0.162

Jan 2023 to Sep 2026, base cost. It removed the same 43 late trades (+4.498 R) and changed 35 exits (−5.415 R). The total fell from +31.269 to +21.357 R.

Without the noon exit, period by period.

HOLD minus its parent: the change in summed base-cost R in each period, not account return. HOLD keeps the same trades, so each bar shows one trade count. For both parents it helped in 2023–2024 and hurt more in 2025–Sep 2026.

What it taught us.

No tested timing change established an improvement over its parent. In the BOTH rule, stopping at 11:00 removed 43 trades that had summed +4.498 R; dropping the noon exit gained 1.992 R in 2023–2024, then lost 7.447 R in 2025–Sep 2026.

Wave 05: Trade only when the evidence is strong? →

Discuss on Reddit

Evidence and limits

Retrospective: every date here had been studied before, and this gold rule was picked after seeing earlier results, so this is not an untouched out-of-sample test.
Costs are assumptions, not measured broker quotes. No swap or financing.
R is the planned signal-to-stop risk, not an account return.
The paired 95% block CIs use 2,000 moving-block replicates of 20 observed dates. They describe the mean daily difference, not R per trade, and do not adjust for the six comparisons, the two interaction estimates or earlier searches.
The total fell in every comparison. A smaller drawdown or a higher average per remaining trade does not count as an improvement here.
Without the noon exit, one gap in the observed data stretched a COST position to about 56 hours of clock time. The cap counts observed bars, not hours, and financing is not modeled.

Provenance

Research journal record RVD-W4-RESULT-001
wave4/results/study.json · sha256 030f51b3468a…
wave4/results/claim-ledger.json · sha256 ebeb12c16f94…
wave4/results/report-receipt.json · sha256 efd0099d238a…
wave4/REPORT.md · sha256 e9e4e3170e40…
Claim review record RVD-W4-WEB-EDGE-LOG-NOTE-001