HYPE vs the Assistance Fund's cost basis

Every bucket is positive because the asset went up. This sample is of a near-continuous uptrend. A table where the worst bucket still prints a positive mean is describing the asset, not an edge. Only the ordering across buckets carries information, and one asset over this period is thin evidence for it.

The windows overlap. A 180-day forward return on daily bars reuses each day 180 times. The whole -day sample is only non-overlapping 180-day windows, so a bucket of ~88 rows is closer to one independent observation than to 88. Treat the hit rates as descriptive, not as trials: "100%" means no 180-day window starting in that bucket happened to end lower, which across a single bull run is close to a tautology.

The basis is observed; the relationship is not. The AF's position and spend are facts on-chain. That price mean-reverts toward them is a hypothesis this chart does not establish — a slow moving average of price would track a similar line, and an earlier version of this work using a 180-day VWAP proxy correlated 0.997 with a plain SMA ratio. Knowing the fund's true cost removes the proxy problem, not the causal one.