We investigate differential market responses to infrastructure versus regulatory events in cryptocurrency markets using event study methodology with 4-category event classification. From 50 candidate events (2019-2025), 31 meet our impact and estimation-data criteria across 4 cryptocurrencies: Bitcoin (BTC), Ethereum (ETH), Solana (SOL), and Cardano (ADA). We employ constant mean and market-adjusted models with event-level block bootstrap confidence intervals (CIs) that properly account for cross-sectional correlation. Our primary comparison focuses on negative-valence events: infrastructure failures (10 events identified; 8 with sufficient estimation data for analysis) versus regulatory enforcement (7 events). We find infrastructure failures produce mean Cumulative Abnormal Return (CAR) of -7.6% (bootstrap 95% CI: [-25.8%, +11.3%]) and regulatory enforcement produces mean CAR of -11.1% (CI: [-31.0%, +10.7%]). The difference in mean CARs of +3.6 percentage points (pp) has CI [-25.3%, +30.9%], p = 0.81. This is a null finding: markets respond similarly to both shock types when controlling for event valence. Robustness checks confirm: (1) consistent negative sign across all window specifications ([0, +1] to [-5, +30]), (2) results survive leave-one-out exclusion of FTX and Terra, (3) market model with BTC/equal-weighted (EW) proxy attenuates but does not flip results. The 4-category classification addresses prior conflation of upgrades with failures. Interpretation note: This exploratory analysis should be treated as hypothesis-generating; any post-hoc theoretical framing requires prospective testing with larger samples.

翻译：本研究采用事件研究法，结合四分类事件体系，探究加密货币市场对基础设施事件与监管事件的不同反应。从50个候选事件（2019-2025年）中，我们筛选出31个满足影响力和估计数据标准的事件，涵盖比特币（BTC）、以太坊（ETH）、Solana（SOL）和Cardano（ADA）四种加密货币。我们运用常均值模型和市场调整模型，并采用事件层面的区块自助法置信区间（CIs），以恰当处理横截面相关性。核心对比聚焦于负面事件：基础设施故障（识别出10起事件；其中8起具备足够估计数据用于分析）与监管执法（7起事件）。研究发现，基础设施故障产生的平均累积异常收益率（CAR）为-7.6%（自助法95% CI：[-25.8%, +11.3%]），而监管执法产生的平均CAR为-11.1%（CI：[-31.0%, +10.7%]）。两者平均CAR的差异为+3.6个百分点（pp），其CI为[-25.3%, +30.9%]，p = 0.81。这是一个零结果发现：在控制事件情感倾向后，市场对两类冲击的反应相似。稳健性检验证实：（1）所有事件窗口设定（[0, +1]至[-5, +30]）下均保持一致的负向符号；（2）排除FTX和Terra事件后结论依然成立；（3）采用BTC/等权重（EW）代理的市场模型会削弱但不会逆转结果。四分类体系解决了以往研究将系统升级与故障混为一谈的问题。解读说明：本探索性分析应被视为假设生成研究；任何事后理论框架均需通过更大样本的前瞻性检验加以验证。