Feeling the blues Moral hazard and debt dilution in Eurobonds before 1914
Introduction
The European sovereign debt crisis, which has been affecting some of the heavily indebted member states of the European Union since 2008, has casted doubts on the long-term sustainability of the European Monetary Union (EMU). Different proposals to tackle the crisis include introducing structural reforms in programme countries, higher inflation targets for the ECB and restructuring EU institutions along the lines of stronger federalism. An important line of this debate focuses on the debt mutualisation through the issue of “Eurobonds.” Having appeared under different labels and guises, the Eurobonds proposal, in essence, recommends that EMU countries pool all or a fraction of their debts. This would presumably reduce borrowing costs for member states and stabilise the European debt market. Although this proposal found some support, it also attracted strong criticisms for its possible caveats, particularly for its potential for generating moral hazard and diluting the outstanding debt stock.
Compared to other tried solutions for fiscal crises, such as debt restructuring or default, inflation and devaluation, and stabilisation plans from multilaterals, there is precious little evidence on the potential effectiveness of the Eurobonds proposal. The bailout and recapitalisation programmes organised since 2010 offer some suggestions of how Eurobonds could fare in the market, particularly after the reduction in mid-2011 of the interest paid by programme countries to close to the effective cost of funding of multilaterals. However, it is questionable whether the relatively short experience with these loans is a good estimate of the consequences of debt mutualisation for the future of European financial stability. The debate has therefore been mainly informed by untested hypotheses about the behaviour of financial markets and sovereigns after the issue of Eurobonds (Claessens et al., 2012).
This paper aims to contribute to this debate by drawing on the history of five guaranteed bonds issued before 1914. Somewhat ironically, these operations started with a Greek loan in 1833, which can arguably be considered the first Eurobond in history. The bonds we study were issued with the guaranty of other sovereigns, usually a combination of the great powers of the time (Britain, France, Germany and Russia) and were perceived by the market as instances of debt mutualisation, unlike the current European programme bonds (EFSF, EFSM and ESM). Perhaps because of the risks involved for the guarantors, these loans were seldom raised and often only after overcoming considerable political opposition within the guarantor countries themselves. After the first Greek loan, there were only four other guaranteed loans issued in our period of study – for Turkey in 1855, Egypt in 1885, China in 1895, and Greece again in 1898. Despite the significant differences between current and pre-1914 international financial architecture, we argue that the guaranteed bonds constitute the most relevant historical examples of debt mutualisation. Consequently, we borrow the current terminology in distinguishing between ‘blue bonds’ (mutualised debt) and ‘red bonds’, the exclusive responsibility of each sovereign (Delpla and von Weizsäcker, 2010).
We use the long historical record of these five loans to address three main questions – how the introduction of guaranteed bonds impacted existing creditors, how they were initially received by the markets, and how markets priced guaranteed debt relative to the other bonds of the assisted countries. The first two questions focus on the short-time horizon around the announcement and flotation of guaranteed bonds. By using an original dataset of daily market prices, we investigate the direct dilution of previous claims on the sovereign, which depended on the relative shares of ‘blue’ (guaranteed) and ‘red’ (non-guaranteed) bonds, seniority dispositions, possible write-downs of existing debts, liquidity of the new issues, and, when included, the benefits from foreign-imposed conditionality. We also compare the yields of guaranteed bonds to the marginal costs of funding of the guarantors. The third question has to be addressed over longer time horizons. In this part of the paper, we quantify the evolution of the spreads of guaranteed bonds relative to other non-guaranteed domestic bonds, and investigate the relation between the yields of guaranteed bonds, their non-guaranteed cousins, and their avuncular guarantors. These spreads are informative of the structural impact of these debt relief operations on the fiscal positions of the recipient nations and should help with addressing the contemporary criticisms about debtors' moral hazard from the issue of Eurobonds.
Our paper also relates to the literature on non-sovereign borrowing and foreign financial intervention in the pre-1914 period. Two of the countries under study here (China and Egypt) were under a status of informal colonial dependency from outside powers. Nevertheless, we show that guaranteed bonds of these countries were perceived differently in the market from other colonial issues, in that they were not priced entirely on the fiat of the colonial power. Colonial issues were more than guaranteed by the coloniser, as the latter reserved complete control over colonial finances (Accominotti et al, 2010, Chavaz, Flandreau, 2015). A different category of bonds was those issued under the financial (and sometimes political) control of foreign creditors, where the latter controlled the domestic sources of revenue set aside to service the foreign debt (Mitchener, Weidenmier, 2010, Tuncer, 2015). Two of the case studies in this paper (Egypt in 1885 and Greece in 1898) fall in this category. Despite the existence of international financial control in Egypt and Greece, their guaranteed bond issues were treated differently in the market. International financial control increased the credibility of each sovereign through the direct transfer of revenues to foreign creditors; however, it did not guarantee any repayments as in the case of guaranteed bonds. Furthermore, there was not necessarily an overlap between the list of guarantors and the powers involved in financial control. In the paper, we elaborate on these two different cases of infringement of sovereignty through colonial issues and international financial control.
The next section describes and compares the current proposals for the issuance of Eurobonds with the corresponding debate in the nineteenth century, emphasising the similarities between the two. Section 3 provides a simple framework to think about the relation between debt dilution and fiscal discipline. Section 4 contains the empirical analysis split into two parts. In the first part we use structural breakpoint methods to identify the short-run impact of the introduction of guaranteed bonds on the previous debt stock. In the second part we investigate the pricing relations between guaranteed and non-guaranteed bonds with the help of factor analysis and dynamic panel VARs. The conclusion section follows.
Section snippets
Varieties
Since September 2008 a half-dozen plans to implement some version of Eurobonds emerged in the literature from academics, trade associations and official organisations. These plans diverge on several levels, especially on the questions of coverage and guaranties. The majority of plans recommend that Eurobonds cover only partially the funding requirements of participating countries.1
Debt dilution, fiscal discipline and spreads
In this section we lay down a simple framework to think about the consequences of issuing guaranteed bonds. Let sg be the fraction of total debt guaranteed ex-post, which is a multiple of the original B' = (1 + δ)B, where δ can be positive or negative. If multilateral debt is simply added to the debt stock then δ > 0. However, if there is a simultaneous debt write-down, it is possible that δ < 0. As the new guaranteed bonds are issued with seniority and the explicit guaranty, their yield must
Data
In our empirical analysis we use 17 bond series, comprising 5 guaranteed bonds, 7 non-guaranteed bonds, and 5 bonds from the main guarantors (the UK, France, and Russia). Our first empirical exercise focuses on the short-term dynamics of adjustment to the issue of guaranteed bonds. For that purpose we use daily prices of the bonds (Table 2). The second exercise studies the long-term relation between guaranteed and non-guaranteed bonds. Because of the length of the periods considered, we use
Conclusions
This paper contributes to the debate on debt mutualisation through the issue of Eurobonds as a solution to the fiscal crisis in the Eurozone by providing the first empirical analysis of the closest available historical parallels: the pre-1914 guaranteed bonds. We started by uncovering a nineteenth-century debate, which shares uncanny similarities with the current arguments pro and against Eurobonds. Despite domestic political opposition, the European powers got involved in these operations to
References (0)
Cited by (12)
Macroeconomic and policy implications of eurobonds
2021, International Review of Law and EconomicsCitation Excerpt :There is certainly political disagreement between member states in the matter of Eurobonds. Yet, academic research has provided some strong arguments in support of joint liability in sovereign borrowing (Basu and Stiglitz, 2015; Tirole, 2015; Baglioni and Cherubini, 2016; Basu, 2016; Favero and Missale, 2016), albeit with some reservations (Beetsma and Mavromatis, 2014; Esteves and Tunçer, 2016; Hatchondo et al., 2017).5 So, what is it all about?
Sovereign debt guarantees and default: Lessons from the UK and Ireland, 1920-1938
2016, European Economic Review'Propitious contrast': Romanian borrowing in a Balkan mirror, 1878-1913
2023, Financial History ReviewGlobal directed technical change model with fiscal and monetary policies, and public debt
2024, Economic Change and RestructuringIn Defense of Public Debt
2022, In Defense of Public DebtEUROPEAN INTEGRATION AS A CHALLENGE FOR THE IMPLEMENTATION OF ECONOMIC STATE SOVEREIGNTY
2022, Access to Justice in Eastern Europe
The authors thank Marc Flandreau, Kris Mitchener, Kim Oosterlinck and Vincent Bignon for detailed comments on previous drafts. The paper has also benefited from discussion with the participants in the First CEPR Economic History Symposium in Perugia, the 10th EHES Conference in London, the 8th Conference of the SEEMHN in Amiens, the 33rd Conference of the APHES in Braga, the 2015 Meeting of the American Economic Association in Boston and seminars at the Banque de France, the University of Geneva and the New University of Lisbon. The usual disclaimer applies. The research leading to these results has received funding from the People Programme (Marie Curie Actions) of the European Union's Seventh Framework Programme FP7/2007-2013/ under REA grant agreement n° 608129.