Sensitivity analysis of entropy-resilience estimates across rolling windows and histogram bins
| Rolling window (weeks) | Histogram bins | Mean entropy | Std entropy | Mean resilience | Std resilience | Min resilience | Max resilience | Best lead-lag (weeks) | Best lead-lag correlation |
|---|---|---|---|---|---|---|---|---|---|
| 8 | 5 | 0.808 | 0.163 | 0.192 | 0.163 | 0.031 | 0.766 | 9 | −0.145 |
| 8 | 8 | 0.736 | 0.128 | 0.264 | 0.128 | 0.083 | 0.819 | 11 | −0.179 |
| 8 | 10 | 0.709 | 0.124 | 0.291 | 0.124 | 0.097 | 0.836 | 9 | −0.152 |
| 8 | 12 | 0.669 | 0.112 | 0.331 | 0.112 | 0.163 | 0.848 | 9 | −0.190 |
| 12 | 5 | 0.806 | 0.200 | 0.194 | 0.200 | 0.013 | 0.822 | 2 | −0.111 |
| 12 | 8 | 0.776 | 0.155 | 0.224 | 0.155 | 0.027 | 0.862 | 3 | −0.103 |
| 12 | 10 | 0.757 | 0.152 | 0.243 | 0.152 | 0.021 | 0.875 | 9 | −0.114 |
| 12 | 12 | 0.735 | 0.143 | 0.265 | 0.143 | 0.093 | 0.885 | 7 | −0.140 |
| 16 | 5 | 0.782 | 0.244 | 0.218 | 0.244 | 0.005 | 0.855 | 2 | −0.114 |
| 16 | 8 | 0.773 | 0.194 | 0.227 | 0.194 | 0.016 | 0.888 | 10 | −0.125 |
| 16 | 10 | 0.764 | 0.186 | 0.236 | 0.186 | 0.022 | 0.898 | 11 | −0.143 |
| 16 | 12 | 0.753 | 0.177 | 0.247 | 0.177 | 0.024 | 0.906 | 10 | −0.155 |
| 20 | 5 | 0.744 | 0.274 | 0.256 | 0.274 | 0.008 | 0.877 | 10 | −0.127 |
| 20 | 8 | 0.756 | 0.216 | 0.244 | 0.216 | 0.022 | 0.905 | 2 | −0.120 |
| 20 | 10 | 0.753 | 0.212 | 0.247 | 0.212 | 0.023 | 0.914 | 10 | −0.145 |
| 20 | 12 | 0.749 | 0.208 | 0.251 | 0.208 | 0.018 | 0.920 | 9 | −0.146 |
| Rolling window (weeks) | Histogram bins | Mean entropy | Std entropy | Mean resilience | Std resilience | Min resilience | Max resilience | Best lead-lag (weeks) | Best lead-lag correlation |
|---|---|---|---|---|---|---|---|---|---|
| 8 | 5 | 0.808 | 0.163 | 0.192 | 0.163 | 0.031 | 0.766 | 9 | −0.145 |
| 8 | 8 | 0.736 | 0.128 | 0.264 | 0.128 | 0.083 | 0.819 | 11 | −0.179 |
| 8 | 10 | 0.709 | 0.124 | 0.291 | 0.124 | 0.097 | 0.836 | 9 | −0.152 |
| 8 | 12 | 0.669 | 0.112 | 0.331 | 0.112 | 0.163 | 0.848 | 9 | −0.190 |
| 12 | 5 | 0.806 | 0.200 | 0.194 | 0.200 | 0.013 | 0.822 | 2 | −0.111 |
| 12 | 8 | 0.776 | 0.155 | 0.224 | 0.155 | 0.027 | 0.862 | 3 | −0.103 |
| 12 | 10 | 0.757 | 0.152 | 0.243 | 0.152 | 0.021 | 0.875 | 9 | −0.114 |
| 12 | 12 | 0.735 | 0.143 | 0.265 | 0.143 | 0.093 | 0.885 | 7 | −0.140 |
| 16 | 5 | 0.782 | 0.244 | 0.218 | 0.244 | 0.005 | 0.855 | 2 | −0.114 |
| 16 | 8 | 0.773 | 0.194 | 0.227 | 0.194 | 0.016 | 0.888 | 10 | −0.125 |
| 16 | 10 | 0.764 | 0.186 | 0.236 | 0.186 | 0.022 | 0.898 | 11 | −0.143 |
| 16 | 12 | 0.753 | 0.177 | 0.247 | 0.177 | 0.024 | 0.906 | 10 | −0.155 |
| 20 | 5 | 0.744 | 0.274 | 0.256 | 0.274 | 0.008 | 0.877 | 10 | −0.127 |
| 20 | 8 | 0.756 | 0.216 | 0.244 | 0.216 | 0.022 | 0.905 | 2 | −0.120 |
| 20 | 10 | 0.753 | 0.212 | 0.247 | 0.212 | 0.023 | 0.914 | 10 | −0.145 |
| 20 | 12 | 0.749 | 0.208 | 0.251 | 0.208 | 0.018 | 0.920 | 9 | −0.146 |
Note(s): The table reports sensitivity results for alternative rolling-window lengths and histogram-bin counts . For each specification, normalized Shannon entropy , resilience , and the strongest negative lead-lag correlation between and are computed. The lead-lag horizon is measured in weeks
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.