Generated 2026-10-04 | PIT state quintiles (trailing 3y rank, no tuning) | block-bootstrap high-vs-low diffs (block 21d, 200 reps) | BH-FDR across conditional tests.
A. Conditional distributions (Q1=low state .. Q5=high state)
Forward sleeve vol/drawdown (Q5-Q1>0 means high state = more risk) and forward sleeve return. Bold = BH q<0.05.
| feature | target | Q1 | Q3 | Q5 | Q5-Q1 | boot95 | q |
|---|---|---|---|---|---|---|---|
| avg_corr20 | eqmom_fwd21 | +0.011 | +0.024 | +0.016 | +0.004 | [-0.013,+0.023] | 0.322 |
| breadth50_chg20 | eqmom_fwd21 | +0.027 | +0.016 | +0.014 | -0.013 | [-0.027,+0.004] | 0.0996 |
| breadth_sma200 | eqmom_fwd21 | +0.025 | +0.006 | +0.020 | -0.005 | [-0.025,+0.014] | 0.253 |
| breadth_sma50 | eqmom_fwd21 | +0.028 | +0.012 | +0.019 | -0.009 | [-0.023,+0.005] | 0.131 |
| corr_spy60 | eqmom_fwd21 | +0.014 | +0.017 | +0.018 | +0.003 | [-0.011,+0.018] | 0.337 |
| disp1 | eqmom_fwd21 | +0.015 | +0.019 | +0.015 | -0.001 | [-0.014,+0.016] | 0.57 |
| disp20 | eqmom_fwd21 | +0.015 | +0.024 | +0.010 | -0.005 | [-0.019,+0.013] | 0.349 |
| hhi_dv | eqmom_fwd21 | +0.020 | +0.017 | +0.014 | -0.006 | [-0.019,+0.007] | 0.236 |
| hl_spread_52w | eqmom_fwd21 | +0.022 | +0.018 | +0.011 | -0.011 | [-0.027,+0.004] | 0.142 |
| top10_dvwt | eqmom_fwd21 | +0.018 | +0.020 | +0.013 | -0.005 | [-0.020,+0.008] | 0.266 |
| top_sector_excess | eqmom_fwd21 | +0.019 | +0.015 | +0.017 | -0.002 | [-0.013,+0.011] | 0.455 |
| avg_corr20 | eqmom_mdd_fwd21 | -0.052 | -0.048 | -0.055 | -0.002 | [-0.017,+0.010] | 0.399 |
| breadth50_chg20 | eqmom_mdd_fwd21 | -0.058 | -0.049 | -0.045 | **+0.013 | [+0.003,+0.024] | 0.0377 |
| breadth_sma200 | eqmom_mdd_fwd21 | -0.053 | -0.053 | -0.055 | -0.002 | [-0.015,+0.011] | 0.337 |
| breadth_sma50 | eqmom_mdd_fwd21 | -0.055 | -0.050 | -0.044 | +0.010 | [+0.000,+0.021] | 0.0512 |
| corr_spy60 | eqmom_mdd_fwd21 | -0.053 | -0.054 | -0.044 | +0.009 | [-0.003,+0.019] | 0.154 |
| disp1 | eqmom_mdd_fwd21 | -0.047 | -0.050 | -0.054 | -0.007 | [-0.018,+0.005] | 0.207 |
| disp20 | eqmom_mdd_fwd21 | -0.049 | -0.046 | -0.056 | -0.007 | [-0.018,+0.006] | 0.242 |
| hhi_dv | eqmom_mdd_fwd21 | -0.043 | -0.047 | -0.058 | **-0.015 | [-0.023,-0.006] | 0 |
| hl_spread_52w | eqmom_mdd_fwd21 | -0.058 | -0.047 | -0.051 | +0.007 | [-0.004,+0.017] | 0.175 |
| top10_dvwt | eqmom_mdd_fwd21 | -0.047 | -0.045 | -0.057 | **-0.011 | [-0.019,-0.002] | 0.0147 |
| top_sector_excess | eqmom_mdd_fwd21 | -0.047 | -0.052 | -0.052 | -0.005 | [-0.015,+0.004] | 0.244 |
| avg_corr20 | eqmom_vol_fwd21 | +0.013 | +0.012 | +0.014 | +0.001 | [-0.001,+0.004] | 0.258 |
| breadth50_chg20 | eqmom_vol_fwd21 | +0.015 | +0.012 | +0.011 | **-0.003 | [-0.006,-0.001] | 0 |
| breadth_sma200 | eqmom_vol_fwd21 | +0.014 | +0.012 | +0.014 | +0.000 | [-0.002,+0.003] | 0.45 |
| breadth_sma50 | eqmom_vol_fwd21 | +0.014 | +0.012 | +0.012 | **-0.003 | [-0.005,-0.001] | 0.0275 |
| corr_spy60 | eqmom_vol_fwd21 | +0.013 | +0.013 | +0.012 | -0.001 | [-0.003,+0.002] | 0.355 |
| disp1 | eqmom_vol_fwd21 | +0.011 | +0.013 | +0.014 | +0.002 | [+0.000,+0.004] | 0.0512 |
| disp20 | eqmom_vol_fwd21 | +0.012 | +0.012 | +0.014 | +0.002 | [+0.000,+0.004] | 0.0512 |
| hhi_dv | eqmom_vol_fwd21 | +0.011 | +0.012 | +0.014 | **+0.002 | [+0.001,+0.004] | 0 |
| hl_spread_52w | eqmom_vol_fwd21 | +0.015 | +0.012 | +0.012 | **-0.002 | [-0.005,-0.001] | 0 |
| top10_dvwt | eqmom_vol_fwd21 | +0.012 | +0.012 | +0.014 | **+0.002 | [+0.000,+0.003] | 0.0147 |
| top_sector_excess | eqmom_vol_fwd21 | +0.012 | +0.013 | +0.013 | +0.002 | [-0.000,+0.004] | 0.106 |
| avg_corr20 | flagship_fwd21 | +0.008 | +0.019 | +0.024 | **+0.016 | [+0.003,+0.033] | 0 |
| breadth50_chg20 | flagship_fwd21 | +0.027 | +0.013 | +0.015 | -0.012 | [-0.024,+0.003] | 0.069 |
| breadth_sma200 | flagship_fwd21 | +0.029 | +0.008 | +0.014 | -0.015 | [-0.035,+0.001] | 0.067 |
| breadth_sma50 | flagship_fwd21 | +0.029 | +0.013 | +0.010 | **-0.019 | [-0.033,-0.006] | 0 |
| corr_spy60 | flagship_fwd21 | +0.013 | +0.017 | +0.026 | +0.013 | [+0.000,+0.026] | 0.0512 |
| disp1 | flagship_fwd21 | +0.012 | +0.017 | +0.022 | +0.010 | [-0.001,+0.022] | 0.069 |
| disp20 | flagship_fwd21 | +0.013 | +0.019 | +0.019 | +0.006 | [-0.007,+0.020] | 0.256 |
| hhi_dv | flagship_fwd21 | +0.018 | +0.019 | +0.012 | -0.006 | [-0.019,+0.005] | 0.196 |
| hl_spread_52w | flagship_fwd21 | +0.028 | +0.019 | +0.009 | **-0.019 | [-0.032,-0.009] | 0 |
| top10_dvwt | flagship_fwd21 | +0.016 | +0.021 | +0.011 | -0.005 | [-0.016,+0.007] | 0.204 |
| top_sector_excess | flagship_fwd21 | +0.015 | +0.013 | +0.023 | +0.008 | [-0.000,+0.016] | 0.06 |
| avg_corr20 | flagship_mdd_fwd21 | -0.038 | -0.035 | -0.050 | **-0.011 | [-0.023,-0.001] | 0.0463 |
| breadth50_chg20 | flagship_mdd_fwd21 | -0.045 | -0.038 | -0.041 | +0.005 | [-0.005,+0.012] | 0.239 |
| breadth_sma200 | flagship_mdd_fwd21 | -0.057 | -0.038 | -0.039 | **+0.018 | [+0.006,+0.028] | 0 |
| breadth_sma50 | flagship_mdd_fwd21 | -0.046 | -0.037 | -0.040 | +0.007 | [-0.002,+0.016] | 0.17 |
| corr_spy60 | flagship_mdd_fwd21 | -0.037 | -0.039 | -0.051 | **-0.014 | [-0.026,-0.005] | 0 |
| disp1 | flagship_mdd_fwd21 | -0.038 | -0.041 | -0.042 | -0.005 | [-0.013,+0.004] | 0.204 |
| disp20 | flagship_mdd_fwd21 | -0.037 | -0.038 | -0.043 | -0.006 | [-0.016,+0.004] | 0.131 |
| hhi_dv | flagship_mdd_fwd21 | -0.037 | -0.038 | -0.045 | **-0.008 | [-0.015,-0.001] | 0.0377 |
| hl_spread_52w | flagship_mdd_fwd21 | -0.046 | -0.039 | -0.037 | +0.009 | [-0.001,+0.016] | 0.069 |
| top10_dvwt | flagship_mdd_fwd21 | -0.037 | -0.038 | -0.044 | **-0.007 | [-0.014,-0.000] | 0.0463 |
| top_sector_excess | flagship_mdd_fwd21 | -0.037 | -0.041 | -0.042 | -0.005 | [-0.012,+0.002] | 0.131 |
| avg_corr20 | flagship_vol_fwd21 | +0.009 | +0.010 | +0.013 | **+0.004 | [+0.002,+0.007] | 0 |
| breadth50_chg20 | flagship_vol_fwd21 | +0.012 | +0.010 | +0.011 | -0.002 | [-0.004,+0.001] | 0.179 |
| breadth_sma200 | flagship_vol_fwd21 | +0.016 | +0.009 | +0.010 | **-0.006 | [-0.009,-0.003] | 0 |
| breadth_sma50 | flagship_vol_fwd21 | +0.013 | +0.009 | +0.010 | **-0.003 | [-0.005,-0.001] | 0.0147 |
| corr_spy60 | flagship_vol_fwd21 | +0.009 | +0.010 | +0.014 | **+0.005 | [+0.003,+0.008] | 0 |
| disp1 | flagship_vol_fwd21 | +0.009 | +0.010 | +0.012 | **+0.002 | [+0.000,+0.004] | 0.0147 |
| disp20 | flagship_vol_fwd21 | +0.009 | +0.010 | +0.012 | **+0.003 | [+0.000,+0.005] | 0.0377 |
| hhi_dv | flagship_vol_fwd21 | +0.010 | +0.010 | +0.011 | +0.001 | [-0.000,+0.003] | 0.0762 |
| hl_spread_52w | flagship_vol_fwd21 | +0.013 | +0.010 | +0.009 | **-0.004 | [-0.006,-0.001] | 0 |
| top10_dvwt | flagship_vol_fwd21 | +0.010 | +0.010 | +0.011 | +0.001 | [-0.000,+0.003] | 0.067 |
| top_sector_excess | flagship_vol_fwd21 | +0.009 | +0.010 | +0.012 | **+0.002 | [+0.000,+0.004] | 0 |
| avg_corr20 | momls_fwd21 | +0.000 | +0.004 | -0.022 | **-0.023 | [-0.041,-0.008] | 0 |
| breadth50_chg20 | momls_fwd21 | -0.003 | -0.001 | -0.007 | -0.004 | [-0.024,+0.012] | 0.337 |
| breadth_sma200 | momls_fwd21 | -0.012 | +0.002 | +0.003 | +0.015 | [-0.002,+0.034] | 0.069 |
| breadth_sma50 | momls_fwd21 | -0.013 | -0.004 | +0.004 | **+0.017 | [+0.005,+0.032] | 0 |
| corr_spy60 | momls_fwd21 | +0.003 | +0.000 | -0.026 | **-0.029 | [-0.051,-0.011] | 0 |
| disp1 | momls_fwd21 | -0.003 | +0.001 | -0.015 | -0.012 | [-0.028,+0.002] | 0.137 |
| disp20 | momls_fwd21 | -0.005 | +0.003 | -0.015 | -0.010 | [-0.027,+0.005] | 0.159 |
| hhi_dv | momls_fwd21 | -0.005 | -0.007 | -0.001 | +0.004 | [-0.007,+0.016] | 0.309 |
| hl_spread_52w | momls_fwd21 | -0.012 | -0.005 | -0.001 | +0.011 | [+0.001,+0.023] | 0.0512 |
| top10_dvwt | momls_fwd21 | -0.001 | -0.009 | -0.002 | -0.002 | [-0.012,+0.007] | 0.393 |
| top_sector_excess | momls_fwd21 | -0.000 | +0.001 | -0.012 | -0.012 | [-0.026,+0.000] | 0.069 |
| avg_corr20 | spy_vol_fwd21 | +0.008 | +0.009 | +0.016 | **+0.008 | [+0.006,+0.012] | 0 |
| breadth50_chg20 | spy_vol_fwd21 | +0.013 | +0.009 | +0.010 | **-0.003 | [-0.006,-0.001] | 0.0275 |
| breadth_sma200 | spy_vol_fwd21 | +0.015 | +0.009 | +0.009 | **-0.007 | [-0.010,-0.004] | 0 |
| breadth_sma50 | spy_vol_fwd21 | +0.015 | +0.009 | +0.009 | **-0.006 | [-0.009,-0.003] | 0 |
| corr_spy60 | spy_vol_fwd21 | +0.008 | +0.010 | +0.016 | **+0.008 | [+0.005,+0.011] | 0 |
| disp1 | spy_vol_fwd21 | +0.008 | +0.009 | +0.014 | **+0.006 | [+0.004,+0.009] | 0 |
| disp20 | spy_vol_fwd21 | +0.008 | +0.009 | +0.014 | **+0.006 | [+0.004,+0.009] | 0 |
| hhi_dv | spy_vol_fwd21 | +0.009 | +0.011 | +0.010 | **+0.001 | [+0.000,+0.003] | 0.0463 |
| hl_spread_52w | spy_vol_fwd21 | +0.015 | +0.009 | +0.008 | **-0.007 | [-0.010,-0.004] | 0 |
| top10_dvwt | spy_vol_fwd21 | +0.009 | +0.011 | +0.011 | **+0.002 | [+0.001,+0.003] | 0.0147 |
| top_sector_excess | spy_vol_fwd21 | +0.008 | +0.010 | +0.014 | **+0.005 | [+0.003,+0.008] | 0 |
B. Interaction cells (PIT terciles, a=feature1, b=feature2)
Mean forward target per cell. worst = lowest-mean cell; worst_minus_rest is the gap to the rest of the sample.
| pair | target | a1b1 | a1b3 | a3b1 | a3b3 | worst | worst-rest |
|---|---|---|---|---|---|---|---|
| breadth_sma200 x avg_corr20 | spy_vol_fwd21 | +0.013 | +0.016 | +0.007 | +0.010 | a3b1 | -0.004 |
| breadth_sma200 x avg_corr20 | eqmom_vol_fwd21 | +0.015 | +0.014 | +0.012 | +0.013 | a2b2 | -0.001 |
| breadth_sma200 x avg_corr20 | flagship_vol_fwd21 | +0.014 | +0.016 | +0.008 | +0.012 | a3b1 | -0.002 |
| breadth_sma200 x avg_corr20 | eqmom_mdd_fwd21 | -0.066 | -0.052 | -0.048 | -0.053 | a2b1 | -0.009 |
| breadth_sma200 x avg_corr20 | flagship_mdd_fwd21 | -0.060 | -0.058 | -0.035 | -0.046 | a1b3 | -0.020 |
| breadth_sma200 x avg_corr20 | momls_fwd21 | +0.030 | -0.021 | +0.004 | -0.003 | a1b3 | -0.020 |
| breadth_sma200 x disp20 | spy_vol_fwd21 | +0.010 | +0.017 | +0.007 | +0.009 | a3b1 | -0.003 |
| breadth_sma200 x disp20 | eqmom_vol_fwd21 | +0.012 | +0.014 | +0.012 | +0.013 | a1b1 | -0.001 |
| breadth_sma200 x disp20 | flagship_vol_fwd21 | +0.012 | +0.017 | +0.008 | +0.010 | a3b1 | -0.002 |
| breadth_sma200 x disp20 | eqmom_mdd_fwd21 | -0.055 | -0.053 | -0.048 | -0.051 | a1b1 | -0.004 |
| breadth_sma200 x disp20 | flagship_mdd_fwd21 | -0.051 | -0.059 | -0.033 | -0.039 | a1b3 | -0.021 |
| breadth_sma200 x disp20 | momls_fwd21 | -0.005 | -0.017 | -0.004 | +0.009 | a1b3 | -0.014 |
| avg_corr20 x disp20 | spy_vol_fwd21 | +0.008 | +0.008 | +0.009 | +0.016 | a1b2 | -0.003 |
| avg_corr20 x disp20 | eqmom_vol_fwd21 | +0.012 | +0.014 | +0.012 | +0.014 | a2b1 | -0.002 |
| avg_corr20 x disp20 | flagship_vol_fwd21 | +0.009 | +0.010 | +0.011 | +0.013 | a2b1 | -0.002 |
| avg_corr20 x disp20 | eqmom_mdd_fwd21 | -0.055 | -0.054 | -0.049 | -0.054 | a1b1 | -0.005 |
| avg_corr20 x disp20 | flagship_mdd_fwd21 | -0.039 | -0.038 | -0.041 | -0.047 | a3b2 | -0.008 |
| avg_corr20 x disp20 | momls_fwd21 | -0.005 | -0.002 | -0.010 | -0.021 | a3b3 | -0.022 |
| breadth_sma200 x top10_dvwt | spy_vol_fwd21 | +0.011 | +0.017 | +0.007 | +0.009 | a3b1 | -0.003 |
| breadth_sma200 x top10_dvwt | eqmom_vol_fwd21 | +0.012 | +0.015 | +0.012 | +0.014 | a3b1 | -0.001 |
| breadth_sma200 x top10_dvwt | flagship_vol_fwd21 | +0.013 | +0.016 | +0.009 | +0.010 | a3b1 | -0.002 |
| breadth_sma200 x top10_dvwt | eqmom_mdd_fwd21 | -0.044 | -0.058 | -0.044 | -0.056 | a1b3 | -0.009 |
| breadth_sma200 x top10_dvwt | flagship_mdd_fwd21 | -0.039 | -0.061 | -0.036 | -0.039 | a1b3 | -0.023 |
| breadth_sma200 x top10_dvwt | momls_fwd21 | -0.005 | -0.012 | +0.003 | +0.003 | a1b2 | -0.012 |
| breadth50_chg20 x avg_corr20 | spy_vol_fwd21 | +0.008 | +0.016 | +0.007 | +0.012 | a3b1 | -0.003 |
| breadth50_chg20 x avg_corr20 | eqmom_vol_fwd21 | +0.013 | +0.016 | +0.012 | +0.011 | a3b3 | -0.002 |
| breadth50_chg20 x avg_corr20 | flagship_vol_fwd21 | +0.010 | +0.014 | +0.009 | +0.012 | a3b1 | -0.002 |
| breadth50_chg20 x avg_corr20 | eqmom_mdd_fwd21 | -0.050 | -0.064 | -0.053 | -0.045 | a1b3 | -0.015 |
| breadth50_chg20 x avg_corr20 | flagship_mdd_fwd21 | -0.043 | -0.052 | -0.038 | -0.044 | a1b3 | -0.013 |
| breadth50_chg20 x avg_corr20 | momls_fwd21 | +0.020 | -0.010 | -0.004 | -0.019 | a3b3 | -0.018 |
C. Incremental information over trailing SPY ret/vol
Rank-OLS R^2 on the target; delta_R2 is the increment from adding the state feature to the SPY ret/vol baseline.
| target | feature | R2 base | R2 +state | dR2 | partial IC |
|---|---|---|---|---|---|
| eqmom_vol_fwd21 | top10_dvwt | 0.049 | 0.148 | +0.100 | +0.324 |
| eqmom_vol_fwd21 | hhi_dv | 0.049 | 0.131 | +0.082 | +0.294 |
| eqmom_vol_fwd21 | corr_spy60 | 0.042 | 0.103 | +0.061 | -0.251 |
| eqmom_mdd_fwd21 | top10_dvwt | 0.008 | 0.058 | +0.050 | -0.224 |
| eqmom_vol_fwd21 | disp1 | 0.049 | 0.097 | +0.048 | +0.226 |
| eqmom_vol_fwd21 | disp20 | 0.049 | 0.096 | +0.048 | +0.224 |
| eqmom_vol_fwd21 | avg_corr20 | 0.049 | 0.095 | +0.047 | -0.221 |
| eqmom_mdd_fwd21 | hhi_dv | 0.008 | 0.055 | +0.047 | -0.217 |
| spy_vol_fwd21 | disp20 | 0.316 | 0.357 | +0.041 | +0.245 |
| spy_vol_fwd21 | disp1 | 0.316 | 0.356 | +0.040 | +0.243 |
| eqmom_mdd_fwd21 | corr_spy60 | 0.006 | 0.045 | +0.039 | +0.197 |
| spy_vol_fwd21 | hl_spread_52w | 0.316 | 0.351 | +0.036 | -0.228 |
| flagship_vol_fwd21 | top10_dvwt | 0.110 | 0.141 | +0.030 | +0.185 |
| spy_vol_fwd21 | top_sector_excess | 0.316 | 0.341 | +0.025 | +0.192 |
| eqmom_mdd_fwd21 | avg_corr20 | 0.008 | 0.029 | +0.021 | +0.145 |
| eqmom_mdd_fwd21 | breadth_sma200 | 0.008 | 0.028 | +0.019 | -0.139 |
| eqmom_vol_fwd21 | breadth50_chg20 | 0.049 | 0.067 | +0.019 | -0.141 |
| flagship_vol_fwd21 | hhi_dv | 0.110 | 0.129 | +0.019 | +0.146 |
| spy_vol_fwd21 | avg_corr20 | 0.316 | 0.333 | +0.017 | -0.158 |
| spy_vol_fwd21 | top10_dvwt | 0.316 | 0.332 | +0.016 | +0.153 |
| flagship_mdd_fwd21 | top10_dvwt | 0.026 | 0.042 | +0.016 | -0.127 |
| flagship_vol_fwd21 | disp1 | 0.110 | 0.125 | +0.015 | +0.128 |
| spy_vol_fwd21 | breadth_sma200 | 0.316 | 0.330 | +0.014 | -0.145 |
| momls_fwd21 | corr_spy60 | 0.010 | 0.023 | +0.012 | -0.112 |
| eqmom_vol_fwd21 | top_sector_excess | 0.049 | 0.061 | +0.012 | +0.115 |
D. Correlation -> momentum-spread
Forward 21d raw momentum long-short spread (decile long/short, hold 21). Sub-period rank-IC of correlation vs forward spread:
- avg_corr20:2006-2012: IC -0.049
- corr_spy60:2006-2012: IC -0.062
- avg_corr20:2012-2017: IC -0.169
- corr_spy60:2012-2017: IC -0.043
- avg_corr20:2017-2021: IC -0.275
- corr_spy60:2017-2021: IC -0.250
- avg_corr20:2021-2026: IC -0.068
- corr_spy60:2021-2026: IC -0.122
Parameter plateau (OOS rank-IC vs forward spread):
- avg_corr20: -0.144
- corr_spy60: -0.206
- disp1: +0.014
- disp20: +0.029
No-tuning conditional timing (long spread unless avg_corr20 in top PIT quintile, else flat; OOS >= 2016):
- spread B&H: -28.6%/-0.06/-53.6% (ret/sharpe/maxdd)
- state-gated: +2.5%/0.09/-35.9%
- deltaSharpe +0.16, deltaMaxDD +17.7%, block-placebo p 0.160
E. Simple market-state classifier
States from PIT quintiles of breadth200 / avg_corr20 / disp20 / breadth50_chg20. Mean forward target and gap vs NEUTRAL.
| target | HEALTHY | FRAGILE | PANIC | RECOVERY | NEUTRAL | HEALTHY-N | FRAGILE-N | PANIC-N |
|---|---|---|---|---|---|---|---|---|
| spy_vol_fwd21 | +0.007 | +0.009 | +0.015 | +0.009 | +0.009 | -0.002 | +0.000 | +0.006 |
| eqmom_vol_fwd21 | +0.012 | +0.014 | +0.013 | +0.012 | +0.011 | +0.001 | +0.003 | +0.002 |
| flagship_vol_fwd21 | +0.008 | +0.010 | +0.013 | +0.010 | +0.009 | -0.000 | +0.001 | +0.005 |
| eqmom_mdd_fwd21 | -0.049 | -0.054 | -0.051 | -0.044 | -0.045 | -0.004 | -0.010 | -0.006 |
| flagship_mdd_fwd21 | -0.035 | -0.039 | -0.048 | -0.037 | -0.034 | -0.000 | -0.004 | -0.014 |
| eqmom_fwd21 | +0.012 | +0.026 | +0.019 | +0.021 | +0.013 | -0.001 | +0.014 | +0.006 |
| flagship_fwd21 | +0.009 | +0.021 | +0.028 | +0.011 | +0.011 | -0.002 | +0.010 | +0.017 |
| momls_fwd21 | +0.002 | +0.004 | -0.024 | +0.005 | +0.002 | -0.000 | +0.002 | -0.026 |
Reading
- 38/88 state-quintile forward-outcome differences are BH-significant.
- The risk-conditioning result (breadth/correlation/dispersion -> forward sleeve VOL) is the strongest, most incremental Phase-2 finding; it is a risk-forecast overlay, not a return edge.
- For the momentum spread, high correlation forecasts a weaker spread and (if the plateau + placebo hold) supports a simple flat-in-high-corr timing candidate.
- Sleeve RETURN conditioning remains weak: state tells you how much risk, not which way.
Phase-2 verdict
Conditioning the frozen sleeves on market state changes their forward RISK distribution materially and incrementally, but not their return direction. Specifically: (1) high concentration (top-10 dollar-volume weight / HHI) and high average cross-market correlation forecast deeper forward DRAWDOWNS and higher forward vol for eqmom and flagship, BH-significant and incremental over trailing SPY ret/vol (top10_dvwt -> eqmom_vol_fwd21 deltaR2 +0.10, partial IC +0.32; top10_dvwt -> eqmom_mdd_fwd21 deltaR2 +0.05); (2) the joint state low-breadth x high-correlation / high-concentration / high-dispersion is the worst forward-drawdown cell (flagship_mdd ~ -6% vs -4% elsewhere) - the hypothesised success condition, in risk form. The correlation -> momentum-spread effect is sign-stable across all four sub-periods and is specific to correlation (dispersion carries no spread plateau), BUT the raw decile spread has negative OOS Sharpe itself and the no-tuning gate's block-placebo p=0.16, so it is NOT promotable standalone.
Recommendation: MORE RESEARCH -> Phase 3 is justified on the RISK/DRAWDOWN side only. A return-timing overlay is not supported. Phase 3 must add a formal chronological walk-forward and an untouched final holdout before any promotion.