We develop a multi-agent model of the insurance and reinsurance sectors and study systemic risk arising from risk model homogeneity. Risk models are accurate only up to a certain degree. In particular, they may err with respect to the correlation of risk events. With a small but significant probability, they would therefore lead to the insolvency and bankruptcy of the company. Would all insurance and reinsurance companies use the same risk model, bankruptcies in the insurance sector would happen in clusters. This would result in structural problems for the entire sector. Evidently, such a result is dangerous on a systemic level and should be avoided. A diversified insurance and reinsurance sector with multiple risk models capturing different parts of the variance is desirable both from the point of view of the individual company, and from that of society and public regulators.
Nevertheless, the number of risk models employed in the insurance sector remains very small. Risk models are research-intensive and must be carefully maintained. The risk modeling sector is internationally dominated by just three risk modeling companies, RMS (Risk Management Solutions), EQECAT, and AIR (Applied Insurance Research); most insurance and reinsurance firms employ combinations of estimates of models provided by these. What is more, the models in question will also not be substantially different, will be built on similar premises, and will be prone to similar types of errors. One reason for this is that risk models are research-intensive and must be carefully maintained. Official accreditation, a densely connected professional network and cautious attitudes in the face of considerable potential losses add to the entry barriers in this field.
While the insurance sector has been extensively studied from game theory and risk modeling perspectives, systemic risk in the insurance sector has not been sufficiently investigated. Risk models are accurate only up to a certain degree. In particular, they may err with respect to the correlation of risk events. With a small but significant probability, they would therefore lead to the insolvency and bankruptcy of the company.
We characterize systemic risk as it results from the number and diversity of risk models, from the statistical properties of damages and claims, and from the somposition of the insurance and reinsurance sectors. We investigate the contributions of proportional reinsurance, excess of loss reinsurance, and of special purpose vehicles such as cat bonds in alleviating systemic risk. Firm level and financial data is used for calibrating the model.
We consider current and proposed insurance regulation and evaluate it in the light of the results of the model. To complement this, we provide some insights from historical cases of catastrophe insurance (hurricane, maritime, earthquake, flood, etc.).