Derived by the Research Lead from the operator FX breadth 11 swing-pullback brief: test continuation after a confirmed pullback in AUDCAD and NZDCAD, independently long or short. The brief supplies the pair set, 4h decisions and 1–7 day horizon; the Research Lead supplies the EMA trend, pullback confirmation and risk rules. Baselines: trend_bars=60, pullback_bars=12, holding_days=5. Execute market decisions at the next eligible hourly open, using hourly replay for protective stops and 4h signals only. Build 4h OHLC from four complete close-stamped hourly bars in UTC bins ending 00/04/08/12/16/20 UTC; discard incomplete bins, including bins crossing market closures. Use the approved FX calendar and zoneinfo DST; never fill missing observations or closures. Prefer the registered per-pair source; record the cutover manifest, source receipt/checksum, timestamp normalization, first/last usable dates and coverage before evaluation. Native 4h bars may be used only if their alignment matches this definition. No historical coverage or deployed source flag was verified in this research step. Freeze the run’s actual FxCostContext and release identifier; require RELEASED spread and financing accounting for the campaign’s economic test. A HELD context is not a financing-inclusive result. Paper admission requires all normal evidence and risk gates, plus explicit disclosure that current FX paper accounting omits carry; reconcile modeled financing separately. Live testing remains demo-only pending account/risk binding and paper acceptance. No orders or research jobs are launched by this proposal.