Cross-market scoreboard, 2026-10-22T13:48:50.563Z Inventory es-fx (tier 1) Six ES strategies on four FX pairs: markets ES 24, EURUSD 24, GBPUSD 24, USDJPY 24, AUDUSD 24; keys on every market 24 keys: A A2 B B2 BB-L-F BB-L-S BB-LS-F BB-LS-S DC-L-F DC-L-S DC-LS-F DC-LS-S DV-L-F DV-L-S DV-LS-F DV-LS-S MC-L-F MC-L-S MC-LS-F MC-LS-S ST60-L ST60-L-T ST60-LS ST60-LS-T [ok] fx_spec.md says: "with the same rules and the same default parameters" [ok] every member has versions gc-daily (tier 1) Golden cross 50/200, daily: markets ES 4, Gold 4; keys on every market 4 keys: GC-L-EMA GC-L-SMA GC-LS-EMA GC-LS-SMA [ok] ma_cross_spec.md says: "Golden cross (50 / 200) on ES and gold" [ok] every member has versions gc-1h (tier 1) Golden cross 50/200, 1-hour: markets ES 12, Gold 12, EURUSD 12; keys on every market 12 keys: GH-L-N GH-L-S GH-L-T1 GH-L-T123 GH-L-T2 GH-L-T3 GH-LS-N GH-LS-S GH-LS-T1 GH-LS-T123 GH-LS-T2 GH-LS-T3 [ok] ge_cross_spec.md says: "nothing changed except the market and its costs" [ok] every member has versions ema921 (tier 1) EMA 9/21 crossover, 1-hour: markets ES 4, NQ 4, GC 4; keys on every market 4 keys: EM-L-N EM-L-T2 EM-LS-N EM-LS-T2 [ok] em_spec.md says: "before any EMA 9/21 value, signal or trade was computed on any of the three markets" [ok] every member has versions turtle (tier 1) Turtle trading, daily: markets Gold 4, EURUSD 4, BTC 4; keys on every market 4 keys: TT-S1-1U TT-S1-4U TT-S2-1U TT-S2-4U [ok] tt_btc_spec.md says: "Identical to `tt_spec.md`" [ok] every member has versions noise (tier 1) Noise area breakout, exchange sessions: markets ES 6, NQ 6, Gold 6; keys on every market 6 keys: AS-L AS-LS LD-L LD-LS NY-L NY-LS [ok] na_spec.md says: "before any noise-area band, signal or trade was computed for this test" [ok] every member has versions im (tier 1) Intraday momentum: markets ES 4, NQ 4; keys on every market 4 keys: OV-L OV-LS RTH-L RTH-LS [ok] ih_spec.md says: "before any first- or last-half-hour return was computed for this test" [ok] every member has versions ib (tier 1) Initial balance breakout: markets NQ 3, ES 3; keys on every market 3 keys: IB-CLOSE IB-MID IB-X1 [ok] ibbes_spec.md says: "**unchanged**" [ok] every member has versions adx (tier 1) ADX + DI, 1-hour: markets EURUSD 2, GBPUSD 2, USDJPY 2, AUDUSD 2; keys on every market 2 keys: ADX-20 ADX-25 [ok] adx_spec.md says: "as in the FX study" [ok] every member has versions noise-btc (tier 2) Noise area breakout: New York session versions and Bitcoin's UTC day: markets ES 2, NQ 2, Gold 2, BTC 2; keys on every market 2 keys: NY-L NY-LS [ok] na_btc_spec.md says: "Rules (as in the earlier noise-area tests)" [ok] every member has versions ad (tier 2) Accumulation/Distribution: Bitcoin 4-hour and SPY daily: markets BTC 3, SPY 3; keys on every market 3 keys: CHO DIV MA [ok] adspy_spec.md says: "The same three readings as on Bitcoin" [ok] every member has versions Check 2. t recomputed from every trade log against the published summaries (difference < 0.01) [ok] 240 versions match their summary t; 2 without a summary file (t from the log) [ok] every version has its six published gates All checks passed. Pairs: tier 1 682 (version, market A -> market B); tier 2 30 Question 1. Transfer of a positive result (tier 1) net > 0 on A -> net > 0 on B: family-weighted transfer 65% vs base 65%, lift -0 pts (8 families); plain 50% vs 44% (301 of 682 pairs) shuffle test: 16% of 10,000 shuffles gave a family-weighted lift at least as large es-fx: 36% vs base 38% (lift -2 pts; 180 of 480 pairs) gc-daily: 100% vs base 100% (lift +0 pts; 8 of 8 pairs) gc-1h: 67% vs base 67% (lift +0 pts; 48 of 72 pairs) ema921: 63% vs base 67% (lift -4 pts; 16 of 24 pairs) turtle: 100% vs base 100% (lift +0 pts; 24 of 24 pairs) noise: 71% vs base 39% (lift +33 pts; 14 of 36 pairs) ib: 80% vs base 83% (lift -3 pts; 5 of 6 pairs) adx: 0% vs base 25% (lift -25 pts; 6 of 24 pairs) gross > 0 on A -> gross > 0 on B: family-weighted transfer 84% vs base 83%, lift +1 pts (8 families); plain 81% vs 74% (503 of 682 pairs) 4+ gates on A -> net > 0 on B: family-weighted transfer 64% vs base 66%, lift -2 pts (6 families); plain 44% vs 44% (54 of 682 pairs) Question 2. Do the rankings agree? Spearman of t across versions (families with 4+ versions) 24 market pairs: median rho 0.25; positive in 18 es-fx: 10 pairs, median rho 0.23 gc-daily: 1 pairs, median rho 0.80 gc-1h: 3 pairs, median rho -0.03 ema921: 3 pairs, median rho 0.40 turtle: 3 pairs, median rho 0.20 noise: 3 pairs, median rho 0.77 im: 1 pairs, median rho 0.80 Question 3. Close and distant pairs close: family-weighted transfer 52% vs base 51%, lift +1 pts (5 families); plain 38% vs 32% (112 of 346 pairs) distant: family-weighted transfer 72% vs base 69%, lift +3 pts (6 families); plain 58% vs 56% (189 of 336 pairs) median rank correlation: close 0.40, distant 0.20 Question 4. Long-only and long-and-short versions long only: family-weighted transfer 75% vs base 72%, lift +3 pts (5 families); plain 57% vs 55% (150 of 274 pairs) long and short: family-weighted transfer 56% vs base 59%, lift -3 pts (8 families); plain 44% vs 37% (151 of 408 pairs) shuffle test (long and short): 45% long and short, close: family-weighted transfer 40% vs base 43%, lift -3 pts (5 families); plain 26% vs 25% (54 of 212 pairs) long and short, distant: family-weighted transfer 56% vs base 61%, lift -5 pts (6 families); plain 54% vs 49% (97 of 196 pairs) long only, close: family-weighted transfer 70% vs base 64%, lift +6 pts (3 families); plain 48% vs 43% (58 of 134 pairs) long only, distant: family-weighted transfer 75% vs base 73%, lift +2 pts (5 families); plain 63% vs 66% (92 of 140 pairs) Question 5. Versions that passed every gate, and their twins Gold GH-L-N (gc-1h, tier 1, t 3.57): ES net + gross + t 0.32 gates 1/6; EURUSD net + gross + t 0.14 gates 1/6 NQ NY-LS (noise, tier 1, t 2.88): ES net + gross + t 0.92 gates 2/6; Gold net - gross + t -0.14 gates 2/6 NQ NY-L (noise, tier 1, t 2.70): ES net + gross + t 1.26 gates 2/6; Gold net + gross + t 0.77 gates 3/6 NQ NY-LS (noise-btc, tier 2, t 2.88): ES net + gross + t 0.92 gates 2/6; Gold net - gross + t -0.14 gates 2/6; BTC net - gross - t -7.95 gates 1/6 NQ NY-L (noise-btc, tier 2, t 2.70): ES net + gross + t 1.26 gates 2/6; Gold net + gross + t 0.77 gates 3/6; BTC net - gross + t -3.24 gates 1/6 Tier 2 (adapted replications, not pooled) net: family-weighted transfer 77% vs base 81%, lift -5 pts (2 families); plain 67% vs 70% (21 of 30 pairs) gross: family-weighted transfer 93% vs base 94%, lift -1 pts (2 families); plain 89% vs 90% (27 of 30 pairs) noise-btc NY-L: ES + t 1.26 2/6; NQ + t 2.70 6/6; Gold + t 0.77 3/6; BTC - t -3.24 1/6 noise-btc NY-LS: ES + t 0.92 2/6; NQ + t 2.88 6/6; Gold - t -0.14 2/6; BTC - t -7.95 1/6 ad CHO: BTC + t 0.72 1/6; SPY + t 0.74 2/6 ad DIV: BTC + t 0.98 2/6; SPY + t 0.31 1/6 ad MA: BTC + t 0.64 1/6; SPY + t 0.91 2/6 Cross-market scoreboard, independent second computation IDENTICAL (112 numbers)