Harwinder Sekhon
042b079861
Gate 6h (big-candle entry gate): blocks all XAU entries (core, XAU_PB, XAU_CP) when the primary TF's last 24h contains a bar with range >= K x prior-bar ATR. Ships default-OFF: validation rejected every threshold (K 3-6 on M5/M15 12mo) — removed trades were net-POSITIVE at every cell; M5 quiet-window trades are the only losing bucket. Config knobs big_candle_atr_multiple/lookback_hours, --big-candle backtest flag, Module 17 tests. M5 relaxation SHIPPED (validated 2026-07-30, Oanda bars, $0.30 spread): m5_min_abs_score 30->25, m5_min_confirmation_score 10->5, hour-21 UTC unblocked M5-only. Beats control in ALL 6 calendar years of the 5yr replay (+230.9R/2319 PF 1.26 DD 21.3 vs +168.0R/2048 PF 1.21 DD 27.0), 12mo +107.4R PF 1.36 vs +84.6R PF 1.33, OOS +26.1R vs +17.2R, $0.70 spread +86.9R vs +66.4R. M15: every relaxation retested negative — config unchanged. Declined-but-open: M5 hour-22 block, M15 hour-16 unblock (both fail 5yr/OOS). Evidence in the factory docstring. Tooling: generic --set PATH=JSON research override in gpu_backtest (typo-proof, logged); diagnose_filters filter_tags refreshed for gates 6d-6i with a staleness-guard test (Module 18); research scripts hour_analysis/burst_profile/book_diff. Suite: 268 tests. Claude-Session: https://claude.ai/code/session_014jqHpmNYmwYMowF6bPZyeD
169 lines
9.2 KiB
Python
169 lines
9.2 KiB
Python
"""XAU/USD signal producer config.
|
|
|
|
PlatformConfig defaults ARE the tuned XAU/USD configuration, so this
|
|
factory sets the timeframes plus documented per-timeframe overrides.
|
|
It must stay equivalent to the inline PlatformConfig(symbol=XAUUSD,
|
|
...) construction (plus those overrides) used by the backtester and
|
|
Telegram bot (enforced by a test in test_all.py).
|
|
|
|
M15 override (validated 2026-07-11, window 2025-03-30 -> 2026-05-31,
|
|
$0.30 spread, trailing exits): tp_floor_rr 1.0 scores +116.6R/460
|
|
trades PF 1.83 vs +105.4R/424 PF 1.82 at the 0.5 default — better in
|
|
both train/val halves and on the 2024-06 -> 2025-03 out-of-sample
|
|
window (+30.3R vs +10.6R; the 1R floor lifts targets past the min-TP
|
|
spread gate, roughly doubling trades there). The same knob HURTS M5
|
|
(pinned 12mo book: +131.3R/699 vs +139.9R/547, PF 1.55 vs 1.78), so it
|
|
is per-timeframe, exactly as PENDING_STEPS' decision rule anticipated.
|
|
|
|
M15 override 2 (validated 2026-07-11, 12mo window 2025-07-11 ->
|
|
2026-07-11 + OOS 2024-06-01 -> 2025-03-30, $0.30 spread): block the
|
|
mixed "trending" regime (ADX strong but EMAs unaligned — momentum
|
|
without directional structure). Those trades lost -6.8R/24 on 12mo,
|
|
negative in BOTH halves; blocking scores +65.7R/499 PF 1.38 DD 10.4R
|
|
vs +59.0R/521 PF 1.32 DD 12.6R baseline, better in both halves. OOS
|
|
is neutral (-1.4R = the 4 such trades that existed there; no power to
|
|
confirm or refute). M5 never trades this regime (its gates already
|
|
exclude it), so the block is M15-only. A second candidate — blocking
|
|
transitioning-regime SELLs (-4.0R/47 on 12mo) — was REJECTED: those
|
|
sells made +5.7R/32 in the 2024 OOS window (bull-market artifact).
|
|
|
|
M15 override 3 (validated 2026-07-16, 5-year replay 2021-07-17 ->
|
|
2026-07-16 under the audit-fixed backtester, $0.30 spread, trailing
|
|
exits): block SELL entries whose bar CLOSES at 15:00 or 17:00 UTC
|
|
(sell_blocked_hours_utc=[15, 17], gate 6g3 — BUYs unaffected; those
|
|
hours flank the already-blocked hour 16). The pocket lost -33.5R/165
|
|
trades, negative in 6/6 calendar years and in BOTH halves of the last
|
|
12 months (H1 -6.8R, H2 -4.2R); each hour independently is 5/6 years
|
|
negative with both halves negative (hours 14/18 fail that screen and
|
|
stay open). Cross-TF the M5 book CONFIRMS (same pocket -14.1R/59,
|
|
5/6 years negative) rather than contradicts — unlike the rejected
|
|
trending_down block. Full replay with the gate: +176.7R/1627 PF 1.31
|
|
DD 29.1R vs +144.9R/1781 PF 1.22 DD 34.4R control; better in ALL SIX
|
|
years; last-12mo halves both improve (H1 +36.3 vs +29.5, H2 +46.9 vs
|
|
+43.2; 12mo PF 1.55 vs 1.43); OOS 2024-06-01 -> 2025-03-30 improves
|
|
(+45.4R PF 1.47 DD 17.8 vs +41.9R PF 1.39 DD 21.7). SELL side goes
|
|
-1.2R -> +30.6R over 5 years. BUYs at those hours made +34.3R, hence
|
|
SELL-only rather than a blocked_hours_utc entry. Mechanism: NY-session
|
|
shorts against gold's macro uptrend (15:00 = post-US-data reversal
|
|
window, 17:00 = London close unwind).
|
|
|
|
M5 override 2 (2026-07-16, chart patterns): enable BUY-only CP on M5
|
|
(inv_hs, double_bottom, bull_flag — drop all bearish patterns). SELL CP
|
|
is consistently negative (OOS -3.21R, 5yr +3.76R PF 1.07) while BUY CP
|
|
is positive across windows: 12mo +41.8R PF 1.60 (train +9.66 PF 1.30,
|
|
val +32.17 PF 1.86), 5yr +65.2R PF 1.48, OOS +1.20R (aggregate
|
|
positive, val half -1.89R — marginal but sign-correct in aggregate).
|
|
Gold's structural uptrend makes bullish patterns fundamentally more
|
|
reliable. M15/H1 fail on all windows; M5-only enablement.
|
|
|
|
M5 override 3 (validated 2026-07-21, 5-year replay 2021-07-21 ->
|
|
2026-07-21, $0.30 spread, next-open fills): trailing distance widened
|
|
to 0.85R (trail_dist_r 0.85, arm unchanged 0.60). Same 30-cell grid
|
|
methodology as the M15 campaign, opposite finding — on M5 the arm
|
|
point barely matters; the winners need a looser leash. Plateau arm
|
|
0.5-0.7 x dist 0.85-1.0 (+227..+234R vs +218.0R baseline). Chosen
|
|
0.60/0.85 (OOS-preferred end of the plateau): +231.2R/2028 PF 1.30 WR
|
|
57.1% vs +218.0R PF 1.29 WR 58.3%; both last-12mo halves better
|
|
(+56.8/+115.1 vs +54.0/+104.6); 5yr max DD 19.2R vs 19.1R (noise);
|
|
OOS 2024-06-01 -> 2025-03-30 +14.5R DD 15.1 vs +11.1R DD 15.4.
|
|
dist 1.0 was REJECTED at the OOS gate (+10.9R < baseline +11.1R)
|
|
despite the best in-sample total (+233.3R) — the OOS ordering picks
|
|
the conservative plateau edge. XAU_PB trades are unaffected (they
|
|
carry their own validated pb_trail 0.8/0.75 per-trade overrides);
|
|
XAU_CP trades share the global params and were validated inside this
|
|
book. M15 arms at 0.70 (its own campaign, same day); the divergence
|
|
is the per-TF pattern yet again.
|
|
|
|
M15 override 4 (validated 2026-07-21, 5-year replay 2021-07-21 ->
|
|
2026-07-21, $0.30 spread, next-open fills): trailing-stop arm raised to
|
|
0.70R (trail_arm_r 0.70, dist unchanged 0.75). Full grid arm 0.3-0.9 x
|
|
dist 0.5-1.0 (30 cells): every earlier-arm cell (0.3/0.45) is WORSE —
|
|
the "arm earlier to save fading trades" hypothesis is refuted; the
|
|
plateau is arm 0.70-0.80 x dist 0.70-0.85 (all >= +190R). Chosen
|
|
plateau center 0.70/0.75: +210.7R/1629 PF 1.35 WR 57.8% vs baseline
|
|
0.60/0.75 +175.7R PF 1.30 WR 53.9%; both last-12mo halves improve
|
|
(+49.7/+50.0 vs +36.4/+46.0); per-year better 4/6 (2022 -0.9R, 2023
|
|
-1.4R, both within tolerance); OOS 2024-06-01 -> 2025-03-30 +57.4R
|
|
PF 1.57 DD 17.0R vs +45.4R PF 1.47 DD 17.8R — better on every OOS
|
|
metric. One clause waived BY OWNER DECISION 2026-07-21: 5yr max DD is
|
|
33.9R vs 29.1R baseline (+4.8R deeper worst episode, 2022-23 chop),
|
|
accepted in exchange for +20% total R. M5 keeps the global 0.6 arm
|
|
(same per-TF divergence pattern as tp_floor_rr and the hour-9 block).
|
|
|
|
M5 override (validated 2026-07-11, full-history replay 2020-03-17 ->
|
|
2026-07-11 + 12mo book, $0.30 spread): block hour 9 UTC (post-London-
|
|
open whipsaw). Hour-9 entries lost -28.9R/143 over 6 years, negative
|
|
in 5/7 calendar years including winning years (2022 -9.2R). Blocking:
|
|
+207.1R vs +178.3R baseline on the 6-year replay, max DD 28.7R vs
|
|
35.5R, improves 5/7 years (2023 goes -13.0R -> breakeven); the tuned
|
|
12mo book is unchanged (+139.0R vs +139.3R). M15-and-up keep hour 9:
|
|
it is POSITIVE there (M15 6yr +1.7R, M15 OOS +8.6R), so the block is
|
|
M5-only, same per-timeframe pattern as the M15 overrides.
|
|
|
|
M5 override 2 — relaxation campaign (validated 2026-07-30, Oanda
|
|
bars, $0.30 spread, trailing exits): UNBLOCK hour 21 UTC on M5 only,
|
|
combined with the config.py M5 gate relaxations (m5_min_abs_score
|
|
30->25, m5_min_confirmation_score 10->5). The combined config beat
|
|
the pre-relaxation control in ALL 6 calendar years of the 5yr replay
|
|
2021-07-30 -> 2026-07-30 (+230.9R/2319 PF 1.26 DD 21.3 vs
|
|
+168.0R/2048 PF 1.21 DD 27.0), on 12mo (+107.4R PF 1.36, halves
|
|
+68.6/+38.8 vs +84.6R PF 1.33), on OOS 2024-06-01 -> 2025-03-30
|
|
(+26.1R vs +17.2R), and at $0.70 spread (+86.9R vs +66.4R). Hour-21
|
|
trades alone: +28.0R/22, positive every year — overturning the old
|
|
block per the full-evidence rule. Hour 21 stays BLOCKED on M15+
|
|
(12mo M15 unblock tested -0.3R). Declined-but-open from the same
|
|
campaign: an M5 hour-22 block (+4.8R on 12mo — the burst-heavy
|
|
"institutional" hour — but -3.2R on 5yr and -1.1R on OOS: regime-
|
|
specific); M15 hour-16 unblock (+0.8R/+57 trades 12mo but -18.3R
|
|
DD +34% on 5yr); every SELL-gate/vol/TP-size relaxation retested
|
|
negative (see memory notes for the full grid).
|
|
"""
|
|
|
|
from typing import Optional, List
|
|
from config import PlatformConfig, Timeframe
|
|
from app_config import apply_pair_overrides
|
|
|
|
_CONFIRMATIONS = {
|
|
Timeframe.M5: [Timeframe.M15, Timeframe.H1],
|
|
Timeframe.M15: [Timeframe.M5, Timeframe.H1],
|
|
Timeframe.H1: [Timeframe.M15, Timeframe.H4],
|
|
Timeframe.H4: [Timeframe.H1, Timeframe.D1],
|
|
}
|
|
|
|
|
|
def make_config(
|
|
primary_tf: Timeframe,
|
|
confirmation_tfs: Optional[List[Timeframe]] = None,
|
|
) -> PlatformConfig:
|
|
config = PlatformConfig()
|
|
# Deployment overlay (config.json) BEFORE the validated per-TF
|
|
# deltas below, so e.g. the M5 hour-9 append lands on the JSON base.
|
|
apply_pair_overrides(config)
|
|
config.primary_timeframe = primary_tf
|
|
config.confirmation_timeframes = (
|
|
confirmation_tfs or _CONFIRMATIONS.get(primary_tf, [Timeframe.M15, Timeframe.H4])
|
|
)
|
|
if primary_tf == Timeframe.M5:
|
|
# Block hour 9 UTC on M5 only; UNBLOCK hour 21 on M5 only
|
|
# (positive every year of the 5yr replay) — see docstring.
|
|
config.xau_strategy.blocked_hours_utc = [
|
|
h for h in config.xau_strategy.blocked_hours_utc if h != 21
|
|
] + [9]
|
|
# BUY-only chart patterns on M5 — see module docstring.
|
|
config.xau_strategy.cp_enabled = True
|
|
config.xau_strategy.cp_patterns = ["inv_hs", "double_bottom", "bull_flag"]
|
|
# Wider trailing distance on M5 only — see module docstring.
|
|
config.signals.trail_arm_r = 0.60
|
|
config.signals.trail_dist_r = 0.85
|
|
if primary_tf == Timeframe.M15:
|
|
# 1R minimum target on M15 only — see module docstring.
|
|
config.signals.tp_floor_rr = 1.0
|
|
# Block the mixed "trending" regime on M15 only — see docstring.
|
|
config.xau_strategy.blocked_regimes = ["trending"]
|
|
# Block SELL entries at 15:00/17:00 UTC on M15 only — see docstring.
|
|
config.xau_strategy.sell_blocked_hours_utc = [15, 17]
|
|
# Later trailing-stop arm on M15 only — see module docstring.
|
|
config.signals.trail_arm_r = 0.70
|
|
config.signals.trail_dist_r = 0.75
|
|
return config
|