The Overhang Hangover

Abstract

We revisit the debt overhang question. We first use non-parametric techniques to isolate a panel of countries on the downward sloping section of a debt Laffer curve. In particular, overhang countries are ones where a threshold level of debt is reached in sample, beyond which (initial) debt ends up lowering (subsequent) growth. On average, significantly negative coefficients appear when debt face value reaches 60% of GDP or 200% of exports, and when its present value reaches 40% of GDP or 140% of exports. Second, we depart from reduced form growth regressions and perform direct tests of the theory on the thus selected sample of overhang countries. In the spirit of event studies, we ask whether, as the overhang level of debt is reached: (i) investment falls precipitously as it should when it becomes optimal to default; (ii) economic policy deteriorates observably, as it should when debt contracts become unable to elicit effort on the part of the debtor; and (iii) the terms of borrowing worsen noticeably, as they should when it becomes optimal for creditors to preempt default and exact punitive interest rates. We find a systematic response of investment, particularly when property rights are weakly enforced, some worsening of the policy environment, and a fall in interest rates. This easing of borrowing conditions happens because lending by the private sector virtually disappears in overhang situations, and multilateral agencies step in with concessional rates. Thus, while debt relief is likely to improve economic policy (and especially investment) in overhang countries, it is doubtful that it would ease their terms of borrowing, or the burden of debt.

Keywords: Debt overhang, kernel estimation, debt contracts, investment, debt relief

JEL Classification: E62, F34, F43, H63