In this paper, we propose a model of endogenous partial insurance and we investigate its implications for macroeconomic outcomes, such as wealth inequality, asset accumulation, interest rate, and consumption smoothing. To this end, we include participation costs to state-contingent asset markets into an otherwise standard Aiyagari (1994) model. We highlight the resulting non-monotonic relationship between wealth and insurance-market participation when insurance is costly. Poor households remain uninsured, middle-class households participate in the insurance market, while rich households decide to self-insure by only purchasing risk-free assets. After theoretically characterizing the endogenous partial equilibrium, we quantify its effect, emphasizing the roles of a participation channel and an interest rate channel.

We propose a rationale for the existence of lending mechanisms, such as the IMF and the ESM, which provide lending to distress sovereigns. First, we show that perfectly competitive markets for sovereign bonds are characterized by the interest overhang externality: when the ownership of debt is anonymous and dispersed, the market price of newly issued bond might be too low to avoid default, even though preventing default would be in the interest of existing creditors. Then, we show that a policy maker/institution can address this externality by facilitating lending to the small open economy government at more favourable terms. This policy is ex-post Pareto improving: the borrower can enjoy credit at lower interest rates, while investors gain from the delay in default even though they are directly financing the policy. Finally, the ex-ante gains are tightly related to the fiscal policy used to finance the intervention. We prove the existence of a Pareto set of fiscal policies that makes the intervention beneficial for all agents.

  • Sovereign Default and Information Frictions [Work in Progress]
    with  Christian Hellwig  (Toulouse School of Economics) and Constance de Soyres (Toulouse School of Economics).

We develop a model of sovereign bond pricing based on dispersed information that can account quantitatively for the sovereign bond spread puzzle. The first contribution is to characterize empirically the very high level of sovereign bond CDS spreads relative to historical default data, and we show that this discrepancy is more severe for bonds with higher ratings and shorter maturities. Second, we build a model with the new ingredient of information frictions and show that it explains a significant fraction of the spreads unexplained by default risk.