The Holdout Problem in Sovereign Debt Markets
I develop a sovereign debt model with endogenous re-entry to international financial markets via debt renegotiation and a possibility for lenders to holdout and litigate. This renders the outcome of a renegotiation process to be characterized by both a haircut and a lenders’ participation rate. I use this model to show that the lenders’ threat to litigate buys commitment to the sovereign. Precisely, in order to increase the lender’s participation rate and hence reduce subsequent litigation, governments in default negotiate lower haircuts; as a result, lenders charge lower spreads ex-ante, during the periods in which the country has access to international financial markets. I use this model to evaluate the role of collective action clauses and find that the optimal threshold for the economy of Argentina during the 1990s was 80%, which is only 5pp above the most widespread threshold used in sovereign debt contracts under NY law since the 2001 Argentine default.
In this paper we develop a sovereign default model with endogenous re-entry to financial markets via debt renegotiation. We use this model to evaluate how shocks to risk-free interest rates trigger default episodes through two channels: borrowing costs and expected renegotiation terms after default. The first channel makes repayment less attractive when risk-free interest rates are high due to higher borrowing costs. The second channel works through the expected subsequent renegotiation process: when risk-free rates are high, lenders are willing to accept a higher haircut in exchange for resuming payments. Thus, high risk-free rates imply better renegotiation terms for a borrower, making default more attractive ex-ante. We calibrate the model to study the 1982 Mexican default, which was preceded by a drastic increase in federal funds rates in the US. We find that the renegotiation process is key for reconciling the model to the widespread narrative that the increase in US interest rates triggered the 1982 default episode.
Reserve Management During Partial Default Episodes
I study the optimal accumulation of international reserves in a sovereign debt model with endogenous re-entry to international financial markets via debt renegotiation. Partial default is what distinguishes my model from a simple off-the-shelf model of renegotiation in which income shocks are the only source of uncertainty. This gives an extra motive for governments to accumulate reserves in good times as a way to hedge against default risk. I find reserves are quantitatively relevant for hedging against income shocks while in financial autarky and for reducing lenders' bargaining power in debt restructuring episodes. Thus, my model accounts for an empirical regularity in emerging markets: governments generally deplete their stocks of reserves during the first stages of a default episode and re-accumulate them in the last stages preceding restructuring.
IMF Program Design and Risk Management: An Event Study Analysis
With Joel Okwuokei; “Does Weak Implementation of Conditionality Increase Financial Risk to the IMF? An Empirical Investigation”
We investigate the factors behind the interruptions in IMF-supported programs under the General Resources Account (GRA) and examine the extent to which a weak implementation of conditionality increases financial risks to the Fund. We use the MONA database to characterize the features of virtually all Fund arrangements since 2002 and the JP Morgan’s EMBIG spreads to proxy for the countries’ capacity to repay. We find that the Board’s report of weak implementation following a program review generally elevates spreads. We also identify a negative correlation between the number of conditionalities evaluated in a program review and its underlying degree of implementation, but this effect is highly dependent on the schedule of future IMF loans.
Preferential Credit and Productivity in Brazil
With Pedro Tanure; previously presented as “Earmarked Loans and Economic Performance in Brazil”
We develop a general equilibrium model with sectoral linkages in which firms face borrowing constraints that can be alleviated by government subsidies. Subsidies are financed by labor taxes. We use this model to evaluate how the Brazilian government’s policy of directing subsidized credit to specific sectors, called earmarked loans, impacts output per worker through two channels. The first one is the general equilibrium effect of alleviating the borrowing constraints of firms in the targeted sector (which also alleviates the constraints of firms in other sectors, given the sectoral linkages), which increases output. The second channel works in the opposite direction. In order to raise funds to subsidize private loans, the government needs to tax labor and hence distorts households’ consumption–labor supply decisions. In this setting, the production network structure is key to determine whether subsidies to a given sector are beneficial or detrimental to the economy. We calibrate the model using Brazilian data and find the observed increase in earmarked credit volume from one third in 2008 of total credit outstanding to one half in 2013 reduced the economy’s welfare. We also compare the optimal to the realized policy and identify the sectors that were over and under subsidized.
AddressDepartment of Economics4-101 Hanson Hall1925 Fourth Street SouthMinneapolis, MN 55455 - USA