Extreme weather events like tropical cyclones, floods, and heavy precipitation may negatively affect several macroeconomic variables such as income growth, total tax revenue, average inflation, and exchange rates. Coping with such impacts often requires countries to increase spending via short-term deficit financing. For those countries facing debt sustainability issues – however – this deficit increase can be very costly, and their recovery thus often relies on financial aid from international donors.

The extent of foreign financial aid is determined after the occurrence of a damaging event. Therefore, aid may take months to materialize and it is impossible to a priori assess what amount will be provided. For this reason, several international initiatives - such as the Sendai Framework for Disaster Risk Reduction, the Paris Agreement, and the InsuResilience Global Partnership - advocate for the adoption of financial instruments that can provide faster and more predictable flows of funding. Among such instruments are sovereign catastrophe pools.

Sovereign catastrophe pools allow different countries to pool their risk into a single portfolio and buy insurance as a group. The main benefit of risk pooling stems from risk diversification, namely the fact that large losses are not experienced simultaneously by all pool’s members. Thanks to risk diversification, insuring the pooled risk requires much less capital than insuring all individual risks separately. Therefore, the higher the pool’s diversification, the higher/lower the coverage/premium that countries obtain/pay for a given premium/coverage. Increasing pools’ risk diversification thus implies increasing countries’ financial resilience.

Currently, three sovereign catastrophe risk pools exist: the Caribbean Catastrophe Risk Insurance Facility (CCRIF), the African Risk Capacity (ARC), and the Pacific Catastrophe Risk Assessment and Financing Initiative (PCRAFI). These pools proved successful, but the risk diversification they provide might be limited for two reasons. First, they were simply not designed with the explicit goal of maximizing risk diversification. Second, they pool risk only regionally and do not benefit from possible diversification benefits obtainable with the inclusion of countries located elsewhere.

The paper addresses these two limitations by introducing a method that finds optimal sovereign risk pools, namely the smallest subgroup of countries - within a given group - having the group’s highest achievable risk diversification. By focusing on tropical cyclone risk, the method is used to answer the following questions: do CCRIF and PCRAFI exploit the full diversification potential of their respective region? If not, could an optimal regional extension of these pools close this diversification gap? To what extent would optimal global pooling increase CCRIF’s and PCRAFI’s diversification?

Results are shown in Figure 1. CCRIF’s and PCRAFI’s risk diversifications are lower than those obtained by the optimal regional pools in the Caribbean and the Pacific, and therefore they do not fully exploit their regional diversification potentials. While an optimal regional extension of PCRAFI would close this gap, this is not the case for CCRIF, where the achieved diversification of a CCRIF’s optimal regional extension reaches 89 % of the diversification capacity in the Caribbean. This possibly occurs because of an initial inefficient design of CCRIF due to the inclusion of correlated countries in the pool. Optimal global extensions of CCRIF and PCRAFI lead to a higher diversification than the original pools (+60 % for CCRIF and +65 % for PCRAFI) and their optimal regional extensions (+15 % for CCRIF and +23 % for PCRAFI). For PCRAFI, this also implies an improvement with respect to the highest achievable diversification in the Pacific. For CCRIF, instead, global pooling provides about the same diversification as the maximum achievable in the Caribbean.