Do Cryptocurrency Markets Differentiate Infrastructure from Regulatory Shocks? A Multi-Moment Event Study with Dependence-Robust Inference has been accepted for publication in Digital Finance (Springer). Editor: Wolfgang Härdle. Production and proofs are next; the version of record DOI will follow.
The paper started as a master’s thesis that was very sure of itself. Under curated events the variance contrast still looks large (about 3.5×). Under a Student-t copula bootstrap that respects cross-asset dependence and heavy tails, that contrast is not distinguishable from zero. Returns tell the same story. Directional, selection-conditional, unresolved on this sample.
An earlier cut reported an apparently decisive multi-asset volatility effect. That significance was mostly inference theatre: correlated assets treated as independent draws, Gaussian bootstrap on heavy-tailed innovations. The portable piece is an inference ladder and a size study for how multi-asset crypto event studies manufacture precision, plus treating event selection as something you measure rather than assert away.
Read the current package: paper page · Zenodo. An updated arXiv version will follow post-accept corrections.