Thirty years of research in international finance comes to a puzzling conclusion: A country that defaults on its debt does not seem to face serious penalties in credit markets in the medium and long run. The effects of defaults on borrowing costs are small or short-lived, and defaulters often regain access to new capital just one year after the crisis (see for example Panizza et al 2009 and Gelos et al 2011). These results stand in sharp contrast to economic theory, which suggests reputational punishment and market exclusion.
This column reassesses the empirical findings on the costs of sovereign default and questions the consensus view that credit markets have short memories. In contrast to earlier literature, we focus on actual default outcomes (ie investor losses), not only on the occurrence of default per se (ie any missed payment). Our intuition is simple: A haircut of 77%, as in Argentina in 2005, is likely to have different implications to a haircut of 10%, as in the restructuring of Uruguay in 2003. This basic idea, however, could not be tested before, primarily due to a lack of sufficient data on restructurings and haircuts.
The study makes two contributions. We first construct a comprehensive archive of all sovereign debt restructuring cases vis-à-vis foreign banks and bondholders between 1970 and 2010. For this purpose we gathered and cross-checked data from more than 200 sources, including the IMF archives, books, policy reports, offering memoranda, private sector research and articles in the financial press. Like in Sturzenegger and Zettelmeyer (2006) we compute haircuts based on the difference between the present values of old and new instruments, discounted at market rates prevailing immediately after the exchange. In a second step, we use the haircut data to study the relationship between restructuring outcomes and subsequent borrowing conditions for debtor governments.
A new debt restructuring archive: Computing haircuts 1970-2010
The dataset, part of which is downloadable here, reveals a number of previously unknown facts on sovereign debt and restructurings:
We find that restructurings are ‘routine affairs’ in low- and middle-income countries. Between 1970 and 2010 there were as many as 180 sovereign debt restructurings in 68 countries. Some countries, like Argentina, Brazil, or Nigeria completed more than six restructurings in the last decades.
The average haircut across all debt restructurings is 37%, meaning that creditors lost, on average, 37 cents on the dollar in present value terms. Haircut size has increased over time, from a mean value of just 25% in the 1980s, to about 50% in the 1990s and 2000s. Figure 1 plots our haircut estimates over time and the respective, inflation-adjusted amounts restructured, as represented by the size of the circles (US dollars of 1980).
Figure 1. Haircuts and restructuring amounts over time
The data show a large variation in haircut size – some restructurings imply investor losses of only about 10 or 20%, while other cases show haircuts of 50%, or even 90%. Table 1 reports summary statistics across all 180 cases, while Table 2 shows an overview of recent deals.
Table 1. Haircut estimates by type of restructuring and era
Table 2. Haircuts in selected recent restructurings (1999-2010)
Post-crisis access to capital: The price of haircuts
The main result from our econometric analysis reveals a new stylised fact: higher haircuts are strongly associated with higher borrowing costs after debt crises, and longer periods of market exclusion.
For illustration, Figure 2 shows that high-haircut countries experience post-restructuring spreads that are about 200 basis points higher than low-haircut countries, especially in years three to seven after the restructuring.
Figure 2. Haircut size and post-restructuring bond spreads
While suggestive, this Figure could potentially be misleading. Perhaps high-haircut countries are materially different from low-haircut countries, and that is what is causing the difference in spreads, not the haircut itself. So, in order to isolate the partial correlation between haircuts and subsequent spreads, we need to control for other potential determinants of spreads.
The econometric analysis confirms that haircut size is a main predictor of post-crisis bond spreads of the sovereign, even after controlling for country and year fixed effects and a number of time-varying determinants of sovereign credit risk identified in the received literature. In the benchmark specification, a 22 percentage point increase in haircuts (one standard deviation) is associated with spreads that are 150 basis points higher (1.5 percentage points) compared to the baseline. While the coefficient decreases over time, it is still 70 basis points in years four and five after the restructuring. In contrast, when using a binary default variable to proxy for past credit history, we find significant effects only one year after the crisis. So while defaults may seem costless when ignoring haircuts, larger haircuts can be clearly associated with larger subsequent spreads.
We next address whether higher haircuts are coupled with longer times of exclusion from capital markets. For illustration, we plot three survival functions for the duration of market exclusion in Figure 3 differentiating by the size of haircut. More simply, each line reports the cumulative probability of remaining excluded from capital markets (y-axis) in the year after the restructuring indicated by the x-axis. The Figure suggests a strong negative correlation between haircut size and the probability of reaccessing international capital markets. Countries imposing HSZ>60% are very likely to remain excluded even after 10 years, with an unconditional probability exceeding 50%.
Figure 3. Probability of reaccessing capital markets after restructurings, depending on haircut size
The econometric analysis provides further indication that the size of haircuts is closely linked to exclusion duration. When estimating a Cox-type survival model, and after controlling for time and region effects, a 30 percentage point increase in haircuts (one standard deviation in this sample) is associated with a 50% lower likelihood of reaccessing international capital markets in any year after the restructuring. Here, reaccess is defined as the first year with a loan or bond placement in international markets and/or in which net debt inflows turn positive.
This column presents new data on the universe of sovereign debt restructurings since 1970. The data and stylised facts are relevant from a policy perspective, as they enable more informed judgements on debt crises outcomes and private creditor burden sharing in the past decades. In addition, the dataset sheds new light on sovereign debt as an asset class. In particular, it provides, for the first time, representative estimates of sovereign debt recovery rates, which were unavailable before.
Our analysis casts doubt on the widespread belief that “debts which are forgiven will be forgotten” (Bulow and Rogoff 1989b). Instead, our findings suggest that defaults can have substantial adverse consequences for governments in the medium run. Once we control for the size of haircuts, defaulting may be more costly than previously believed.
Despite this, our results should not be misinterpreted. We did not identify a direct channel linking haircuts and sovereign borrowing conditions. Thus, we cannot be sure whether we observe punishment effects, reputational effects or neither of the two. The results also do not imply that countries in default should try to minimise their haircut.
In a nutshell, however, we provide indicative evidence for the existence of a trade-off– achieving a high degree of debt relief now can have benefits in the short-run, but may also imply worse borrowing conditions in the future.
Bulow, Jeremy, and Kenneth S Rogoff (1989), “Sovereign Debt: Is to Forgive to Forget?”, American Economic Review, 79(1):43-50.
Gelos, Gaston R, Ratna Sahay, and Guido Sandleris (2011), "Sovereign Borrowing by Developing Countries: What Determines Market Access?", Journal of International Economics, 83(2):243-254.
Panizza, Ugo, Sturzenegger, Federico, and Jeromin Zettelmeyer (2009), “The Economics and Law of Sovereign Debt and Default”, Journal of Economic Literature, 47(3) 651-98.
Sturzenegger, Federico, and Jeromin Zettelmeyer (2006), Debt Defaults and Lessons from a Decade of Crises, MIT Press.