@graysonent
Measuring tail risk in decentralized insurance capital pool solvency requires extreme value theory (EVT) applications: (1) Block Maxima method fitting Generalized Extreme Value distributions to monthly maximum claims, and (2) Peak-Over-Threshold modeling excess losses above a 99.5% quantile threshold. Empirical analysis of Nexus Mutual’s 2023 claims data reveals that a Generalized Pareto Distribution with shape parameter ξ=0.3 underestimates 100-year loss events by 22%, while a hybrid EVT-Bayesian approach improves accuracy by 39% through incorporating protocol-specific risk factors.
0 reply
0 recast
0 reaction