Essays On The Applications Of Network Analysis To The Reinsurance Market

Abstract

This dissertation consists of two topics. Chapter 1 The Microstructure of the Reinsurance Network among US Property-Casualty Insurers and Its Effect on Insurers' Performance models the connectivity within the US property-casualty (P/C) reinsurance market as a network. It provides the first detailed empirical analysis of the microstructure of the reinsurance network including both affiliated and unaffiliated insurers. I find that reinsurance networks are highly sparse and yet largely connected, and exhibit hierarchical core-periphery structure. Moreover, an insurer's network position, measured by its network centrality, has economically significant implications for its loss experience and performance. Particularly, I find that there is an inverse U-shaped relationship between an insurer's network position and its combined ratio, and a U-shaped relationship between an insurer's network position and its performance measured by risk adjusted return on assets and risk adjusted return on equity. I also analyze the resilience of the reinsurance network against possible contagion risk by simulating economic impacts resulting from failures of one or more strategically networked reinsurers. The simulation results suggest that US Property-Casualty insurance industry is resilient to the failure of one or more top reinsurers. Chapter 2 Tail Risk Spillover and Its Contribution to Systemic Risk: A Network Analysis for Global Reinsurers analyzes the dynamic short-run tail risk dependence among global reinsurers and studies its contributions to global reinsurers' systemic risk, where a reinsurer's tail risk is measured by the Value-at-Risk. The tail risk dependence or tail risk spillover among global reinsurers is modeled as networks based on Granger Causality test. The results show that the tail risk interconnectedness among global reinsurers is subject to the impacts of both the insurance industry-wide shock and economy-wide shocks, where the former seems to have a larger effect than the latter. Moreover, I find that a reinsurer's role in the tail risk network as measured by degree/eigenvector centrality contributes significantly to its systemic risk, i.e., a more central tail risk network position will cause a higher level of systemic risk. I also find that there is a threshold effect of tail risk connectedness to systemic risk. That is, when the tail risk connectedness, as measured by daily network density, is below its median state, an increase in a reinsurer's tail risk network centrality will result in a decrease in its systemic risk possibly through risk diversification. In contrast, when the tail risk connectedness is above such threshold, an increase in the reinsurer's tail risk network centrality will lead to an increase in its systemic risk.