In this paper, we explore how an extremely large and persistent catastrophic shock is shared between two countries under the assumption that markets are complete, but each country is subject to endogenous solvency constraints. Catastrophic shocks we consider here constitute of exogenous country-specific events, and have serious repercussions on national outputs; examples include natural disasters and disastrous epidemics.
We introduce solvency constraints into a complete markets setup following the theoretical framework proposed by Lustig (2007). On the one hand, a reason for adopting a complete markets setup is that catastrophic shocks are publicly observable, and markets are therefore likely to exist for such states. On the other hand, a reason for assuming solvency constraints is that it is not easy to force agents to fully recognize outstanding liabilities in a cross-border financial arrangement. In Lustig’s (2007) framework, Lucas trees and all contingent claims are traded among agents, however agents’ solvency is limited in that human capital cannot be used as collateral. An essential aspect of this characterization of solvency constraints is that the current size of short positions in a financial instrument is endogenously determined by the future value of collateral assets (long positions in other financial instruments).
The existing literature intensively explored to which extent the transaction of non-contingent bonds among agents can compensate for the absence of insurance markets in the presence of exogenous borrowing constraints. In this case, the upper limit is exogenously imposed on the current outstanding debts or short positions. The most important result from this setup is that because borrowing constraints keep agents from fully financing uncovered losses by non-contingent bonds, either large or persistent shocks are not shared effectively among agents.
The major results from our setup differ substantially from those from the above combination of incomplete markets and borrowing constraints in the following respects. First, the constrained efficient allocation is fairly close to the corresonding Preto optimal allocation. That is, most of large and persistent catastrophic shocks are shared effectively between the two countries.
Second, however, solvency constraints affect profoundly how the two countries trade financial instruments to share catastrophic shocks. As for the ex ante financial arrangement or the insurance contract made prior to a catastrophic event, the insurance payment to a damaged country is severely constrained by the limited solvency of a nondamaged country as an insurer, given the expectation that Lucas trees will be heavily discounted in the aftermath.
