R0712: PASS

Descriptive statistics, no trading costs in the probabilities (the backtest columns alone use estimated costs). Rules were discovered and tuned on 2022-01 → 2024-12 only, frozen, then tested once on 2025-01 → 2026-10. Millions of hypotheses were searched, so read the corrected p-values and the out-of-sample columns, not the training hit rates. Not investment advice; not a trading system.

Rule: When (price is below the CPR) AND prior-day range ≤ 1.21% (≈302 pts at 25,000) at 11:00

Target: price reaches the prior-day low after 11:00 (days not yet at PDL) · Decision time: 11:00 IST (uses bars up to 11:00 only) · Method: rulefit · Family: level / structure

71%train hit (n=42) vs base 18% · lift 4.02

58%OOS hit (n=33) vs base 16% · lift 3.71

[41%–73%]OOS Wilson 95% CI

8.3e-07OOS BH q (all) · family BH q 5.0e-07 · Holm 3.5e-05

5.0e-04OOS matched-random placebo p

Verdict criteria

✔ Out-of-sample lift in the same direction as train
✔ OOS Wilson 95% CI excludes the OOS base rate
✔ OOS BH q ≤ 0.05 across all candidates tested OOS AND within its target family
✔ Train BH q ≤ 0.05 (across all hypotheses searched)
✔ Stable by year (same direction in ≥80% of years with ≥10 days, incl. every OOS year with ≥10)
✔ OOS matched-random placebo p ≤ 0.05
✔ OOS sample ≥ 20 days

Train p 7.1e-14; train BH q 5.4e-11; train Holm 6.6e-07 (corrections over all 9,242,869 hypotheses); whole-pipeline shuffled-label placebo family-wise p 0.010 (100 permutations). Extra conditions beat the best single condition OOS: True.

By year

year	rule days	hit	base (all days)	lift
2022	11	82%	19%	4.29
2023	19	63%	20%	3.14
2024	12	75%	14%	5.46
2025 (OOS)	20	55%	15%	3.71
2026 (OOS)	13	62%	16%	3.73

Spot-points backtest (estimated costs)

enter short at 11:00 price, take profit at PDL; stop 0.4% of entry; exit at close. Costs: Nifty futures round-trip model from nifty-setups-backtest/src/costs.py (brokerage, exchange, SEBI, stamp, GST, STT by date: 0.0125% → 0.02% from Oct-2024 → 0.05% from Apr-2026) + 0.5 pt slippage per side; spot used as a futures proxy (basis ignored); stop-first on ambiguous bars.

window	trades	win rate	avg net pts	avg gross pts	total net	max DD pts	PF
train 2022–24 (in-sample)	42	74%	+13.5	+19.6	+566.6	+139.2	+1.9
OOS 2025–26	33	61%	+11.7	+21.9	+385.8	+223.8	+1.6

equity curve

trades.csv · Generated 2026-10-10 00:44 IST.