What happened to India-specific relative returns after past coverage episodes: the full historical distribution, not just the average. Pick a channel, a market pair, and a horizon. Every number is an association from history; nothing here predicts anything, and past episodes do not constrain future ones.
| Episode start | Return over horizon |
|---|
Bars: count of past episodes whose cumulative relative return over the chosen horizon fell in each bin. The dashed line marks zero. Summary shows the bootstrap mean, its 95% interval, and whether it survives the multiple-testing control (Benjamini-Hochberg at 10%); most cells do not, and that absence of separable effects is a published finding, not a failure.