Do Cryptocurrency Markets Differentiate Infrastructure from Regulatory Shocks? A Multi-Moment Event Study with Dependence-Robust Inference

Battle Count: 72

Directly relevant to event-driven trading strategies in cryptocurrency markets. The paper demonstrates that naive cross-asset inference severely over-rejects (35% at 5% nominal level), which is critical for any quant developing event-based signals. The finding that infrastructure events carry larger conditional variance responses (directional, 3.49x) could inform volatility-targeting or risk-parity adjustments around known event types. The weekly sentiment-leads-volatility finding (litigation-concentrated) could inform regulatory-event trading strategies. However, the core empirical result is a failure to reject, limiting direct actionable signal generation. The methodological toolkit (inference ladder, dependence-robust bootstrap) is broadly applicable to any multi-asset event study in crypto.

High complexity: requires fitting GJR-GARCH-X models with Student-t innovations on multiple assets, implementing a Student-t-copula parametric bootstrap with multivariate-t draws and per-asset marginal mapping, event-level block bootstrap, PELT change-point detection, Toda-Yamamoto VAR with HC1-robust Wald tests, Benjamini-Hochberg FDR correction, and a 1000-panel Monte-Carlo calibration study. The inference ladder spans seven distinct procedures. Requires careful handling of near-integrated persistence, positivity constraints, and multistart optimization. The replication package with 21 scripts indicates substantial implementation effort.

Complete replication package with scripts c1-c21, estimators, data, outputs, dependency specification, and release verifier. Public GitHub repository and Zenodo record (DOI: 10.5281/zenodo.18099608). Estimators available as standalone open-source Python packages (gjr-garch-x on PyPI, robust-eventstudy on PyPI). All results reproduce from repository alone without API access. Author explicitly reports self-correction of earlier significance claim with full transparency.