Abstract This study presents a comprehensive risk management analytical suite for interdependent assets, applying production engineering methodologies to cryptocurrency portfolio optimization. The proposal integrates principal component analysis, K-means clustering, hidden Markov model regime detection, structural shock decomposition, network causality analysis, GARCH volatility modeling, and stationarity testing to provide a multifaceted approach to risk assessment and decision support. Analysis of fourteen cryptocurrency assets over a multi-year period reveals extreme risk concentration with 67.89% of portfolio variance explained by the first systematic factor, six distinct operational clusters, and five market regimes with volatility ratios reaching 6.91-fold. The study identifies supply chain disruption events as the primary source of mean cumulative abnormal return of negative 10.27%, while network analysis reveals hierarchical information transmission structures with Ethereum and Bitcoin as dominant hubs. GARCH modeling demonstrates mean volatility persistence of 0.938 with half-lives averaging 20.8 days. These findings validate return-based methodologies and provide actionable insights for resource allocation, position sizing, and dynamic hedging strategies in high-volatility environments. The proposal establishes a guide for financial risk management under extreme uncertainty.
Fonte et al. (Mon,) studied this question.