This Time Is Different: Eight Centuries of Financial Folly Read online

  In the modern era, however, the idea of using gunboat diplomacy to collect debts seems far-fetched (in most cases). The cost-benefit analysis simply does not warrant governments’ undertaking such huge expenses and risks, especially when borrowing is typically diversified across Europe, Japan, and the United States, making the incentives for an individual country to use military force even weaker.

  What carrots or sticks, then, can foreign creditors actually hold over sovereign borrowers? This question was first posed coherently in a classic paper by Jonathan Eaton and Mark Gersovitz, who argued that in a changing and uncertain world there is a huge benefit to countries in having access to international capital markets.4 In early times, capital market access might have enabled countries to get food during times of an exceptionally bad harvest. In modern times, countries may need to borrow to fight recessions or to engage in highly productive infrastructure projects.

  Eaton and Gersovitz argued that the benefits of continued capital market access could induce governments to maintain debt repayments absent any legal system whatsoever to force their cooperation. They based their analysis on the conjecture that governments need to worry about their “reputation” as international borrowers. If reneging on debt damages their reputation, governments will not do so lightly. The Eaton and Gersovitz approach is appealing to economic theorists, especially because it is relatively institution-free. (That is, the theory is “pure” in that it does not depend on the particulars of government, such as legal and political structures.) In principle, the theory can explain sovereign borrowing in the Middle Ages as well as today. Note that the reputation argument does not say simply that countries repay their debts now so they can borrow even more in the future. If that were the case, international borrowing would be a Ponzi scheme with exploding debt levels.5

  This “reputation approach” has some subtle problems. If the whole edifice of international lending were built simply on reputation, lending markets might be even more fragile than they actually are. Surely fourteenth-century Italian financiers must have realized that England’s Edward III might die from battle or disease. What would have become of their loans if Edward’s successor had had very different goals and aspirations? If Edward had successfully conquered France, what need would he have had for the lenders in the future?6 If institutions really do not matter, why, over most of history, has the external debt of emerging markets been denominated largely in foreign currency and written so that it is adjudicated in foreign courts?

  Bulow and Rogoff raised another important challenge to the notion that institutions and international legal mechanisms are unimportant in international lending.7 Countries may, indeed, be willing to repay debts to maintain their right to borrow in the future. But at some point, England’s debt burden would have had to reach a point at which the expected value of repayments on existing debt exceeded any future borrowing. At some point, a country must reach its debt limit. Why wouldn’t Edward III (or his successor) have simply declared the Italian debts null and void? Then England could have used any payments it might have made to its financiers to build up gold reserves that could be used if it experienced a shortfall in the future.

  The reputation approach therefore requires some discipline. Bulow and Rogoff argue that in modern times sophisticated investing strategies (e.g., those used in foreign stock markets) might offer as good, or almost as good, a hedge against default as any potential stream of foreign lending. In another work, Bulow and Rogoff contend that instead of relying simply on reputation, repayment of much foreign borrowing, especially by emerging markets, might be enforced by the legal rights of creditors in the lenders’ own countries.8 If a country tried to move to self-insurance, many of the investments it might need to make would involve overseas purchases. Creditors might not be able to seize assets directly in the borrowing country, but, armed with sufficient legal rights, they might well be able to seize the borrower’s assets abroad, particularly in their own countries, but potentially also in other countries with highly developed legal systems. Of course, the right to seize assets abroad will also make it difficult for a defaulting country to borrow from other international lenders. If a country defaults on foreign bank A and then attempts to borrow from foreign bank B, bank B has to worry whether bank A will attempt to enforce its prior claim when it comes time for the country to repay. In this sense, the reputation and legal approaches are not so different, though the resemblance can become significant when it comes to policy questions about how to design and operate the international financial system. For example, establishing an international bankruptcy court to replace domestic courts may be virtually irrelevant if legal rights are of little consequence in any event.

  Emphasizing legal rights also leads one to focus on other costs besides being cut off from future borrowing. A government contemplating default on international loans must also contemplate the potential disruption to its trade that will result from the need to reroute trade and financing to circumvent creditors. Fourteenth-century England depended on selling wool to Italian weavers, and Italy was the center of the trade in spices, which England desired to import. Default implied making future trade with and through Italy difficult, and surely this would have been costly. Nowadays, trade and finance are even more closely linked. For example, most trade, both within and across countries, is extremely dependent on very short-term bank credits to finance goods during shipment but prior to receipt. If a country defaults on large long-term loans, creditor banks can exert significant pressure against any entity that attempts to finance trade credits. Countries can deal with this problem to some extent by using government foreign exchange reserves to help finance their trade. But governments are typically ill equipped to monitor trade loans at the microeconomic level, and they cannot easily substitute their own abilities for bank expertise. Last but not least, creditors can enforce in creditor countries’ courts claims that potentially allow them to seize any defaulter country’s goods (or assets) that cross their borders. Bulow and Rogoff argue that, in practice, creditors and debtors typically negotiate a partial default so that one seldom actually observes such seizures.

  At some level, neither the reputation-based model of Eaton and Gersovitz nor the institutional approach of Bulow and Rogoff seems quite adequate to explain the scale and size of international lending or the diversity of measures creditors bring to bear in real-life default situations. Trade depends not only on legal conventions but also on political resistance to tariff wars and on a broader exchange of people and information to sustain business growth and development.

  Indeed, whereas a country’s reputation for repayment may have only limited traction if construed in the narrow sense defined by Eaton and Gersovitz, its reputation interpreted more broadly—for instance, for being a reliable partner in international relations—may be more significant.9 Default on debt can upset delicate balances in national security arrangements and alliances, and most countries typically have important needs and issues.

  In addition to loans, foreign direct investment (FDI) (for example, when a foreign company builds a plant in an emerging market) can also be important to development. A foreign company that wants to engage in FDI with a defaulting country will worry about having its plant and equipment seized (a prominent phenomenon during the 1960s and 1970s; examples include Chile’s seizure of its copper mines from American companies in 1977 and the nationalization of foreign oil companies’ holdings in the early 1970s by the Organization of Petroleum-Exporting Countries). A debt default will surely cast a pall over FDI, costing the debtor country not only the capital flows but also the knowledge transfer that trade economists find typically accompanies FDI.10

  In sum, economists can find arguments to explain why countries are able to borrow abroad despite the limited rights of creditors. But the arguments are surprisingly complex, suggesting that sustainable debt levels may be fragile as well. Concerns over future access to capital markets, maintaining trade, and possibly broader international
relations all support debt flows, with the relative emphasis and weights depending on factors specific to each situation. That is, even if lenders cannot directly go in and seize assets as in a conventional domestic default, they still retain leverage sufficient to entice a country to repay loans of at least modest size. We can dismiss, however, the popular notion that countries pay back their debts so that they can borrow even more in the future. Ponzi schemes cannot be the foundation for international lending; they must eventually collapse.

  How does the limited leverage of foreign creditors relate to the fragility of confidence we emphasized in the preamble? Without going into great detail, it is easy to imagine that many of the models and frameworks we have been alluding to produce highly fragile equilibria in the sense that there are often multiple outcomes that can be quite sensitive to small shifts in expectations. This fragility comes through in many frameworks but is most straightforwardly apparent in cases in which highly indebted governments need to continuously roll over short-term funding, to which we will turn next.

  Illiquidity versus Insolvency

  We have emphasized the important distinction between willingness to pay and ability to pay. Another important concept is the distinction between a country that faces a short-term funding problem and one that is not willing and/or able to service its debts indefinitely. In most of the literature, this distinction is typically described as the difference between “illiquidity and insolvency.” Of course, the reader now understands that this literal analogy between country and corporate debt is highly misleading. A bankrupt corporation may simply not be able to service its debts in full as a going concern. A country defaulter, on the other hand, has typically made a strategic decision that (full) repayment is not worth the necessary sacrifice.

  Often governments borrow internationally, either at relatively short horizons of one to three years or at longer horizons, at interest rates linked to short-term international debt. Why borrowing tends to be relatively short term is a topic of its own. For example, Diamond and Rajan contend that lenders want the option of being able to discipline borrowers that “misbehave,” that is, fail to invest resources, so as to enhance the probability of future repayment.11 Jeanne argues that because short-term borrowing enhances the risk of a financial crisis (when often debt cannot be rolled over), countries are forced to follow more disciplined policies, improving economic performance for debtor and creditor alike.12 For these and other related reasons, short-term borrowing often carries a significantly lower interest rate than longer-term borrowing. Similar arguments have been made about borrowing in foreign currency units.

  In either event, when a country borrows short term, not only is it faced with financing interest payments (either through its own resources or through new borrowing) but it must also periodically roll over the principal. A liquidity crisis occurs when a country that is both willing and able to service its debts over the long run finds itself temporarily unable to roll over its debts. This situation is in contrast to what is sometimes casually labeled an “insolvency” problem, one in which the country is perceived to be unwilling or unable to repay over the long run. If a country is truly facing merely a liquidity crisis, a third party (for example, a multilateral lending organization such as the International Monetary Fund) can, in principle, make a short-term bridge loan, with no risk, that will keep the borrower on its feet and prevent it from defaulting. Indeed, if creditors were fully convinced that a country had every intention of repaying its debts over the longer term, the debtor would hardly be likely to run into a short-term liquidity problem ever again.

  Sachs illustrates an important caveat.13 Suppose that the money a country borrows is provided by a large group of lenders, each of which is small individually. It may be in the collective interest of the lenders to roll over short-term debt. Yet it can also produce equilibrium if all lenders refuse to roll over the debt, in which case the borrowing country will be forced into default. If no single lender can provide enough money for the country to meet its payments, there may be both a “default” and a “no-default” equilibrium. The example given by Sachs is, of course, a very good illustration of the theme of financial fragility and the vulnerability of debtors to the “this-time-is-different” syndrome. A borrower can merrily roll along as long as lenders have confidence, but if for some (possibly extraneous) reason confidence is lost, then lending collapses, and no individual lender has the power or inclination to stave it off.

  The concept of illiquidity versus insolvency is one we already illustrated in the preamble with bank runs and one that we will see again in other guises. Technically speaking, countries can sometimes be exposed to “multiple equilibria,” implying that the difference between a case in which a country defaults and one in which it does not default can sometimes be very small. For a given structure of debt and assuming all actors are pursuing their self-interest, there can be very different outcomes depending on expectations and confidence.

  Theorists have developed many concrete examples of situations in which default can occur as a result of a “sunspot” that drives a country from a no-default to a default equilibrium.14 The possible existence of multiple equilibria and the idea that investors may temporarily become skittish about a country can also play an important role in rationalizing intervention into sovereign lending crises by the governments of creditor countries and international institutions. The danger, of course, is that it is not always easy to distinguish between a default that was inevitable—in the sense that a country is so highly leveraged and so badly managed that it takes very little to force it into default—and one that was not—in the sense that a country is fundamentally sound but is having difficulties sustaining confidence because of a very temporary and easily solvable liquidity problem. In the heat of a crisis, it is all too tempting for would-be rescuers (today notably multilateral lenders such as the IMF) to persuade themselves that they are facing a confidence problem that can be solved with short-term bridge loans, when in fact they are confronting a much more deeply rooted crisis of solvency and willingness to pay.

  Partial Default and Rescheduling

  Until now, we have somewhat glossed over the point of exactly what constitutes default. In practice, most defaults end up being partial, not complete, albeit sometimes after long negotiations and much acrimony. Creditors may not have the leverage (from whatever source) to enforce full repayment, but they typically do have enough leverage to get at least something back, often a significant share of what they are owed. Even the most famous cases of total default have typically ended in partial repayment, albeit often quite small and many decades later. Russia’s Bolshevik government refused to repay Tsarist debts in 1918, but when Russia finally re-entered the debt markets sixty-nine years later, it had to negotiate a token payment on its defaulted debt.

  In most cases, though, partial repayment is significant and not a token, with the amount repaid presumably determined by the types of complex cost-benefit considerations we have already been discussing. Precisely because partial repayment is often the result of long and contentious negotiations, interested bystanders often get sucked in. For example, Bulow and Rogoff show how well-intentioned third parties such as international lending institutions (e.g., the IMF) or the governments of creditor countries may be gamed into making side payments to facilitate a deal, much as a realtor may cut her commission to sell a house.15 Country borrowers and their creditors potentially have bargaining power vis-à-vis outside parties if failed negotiations interfere with trade and cause broader problems in the global financial system, such as contagion to other borrowers.16 As we have noted, the creation of the IMF since World War II has coincided with shorter but more frequent episodes of sovereign default. This phenomenon is quite consistent with the view that default episodes occur even more frequently than they otherwise might, because both lenders and borrowers realize that in a pinch they can always count on subsidies from the IMF and the governments of creditor countries. (Later literature has c
ome to term this gaming of third parties with deep pockets the “moral hazard” of international lending.)

  A bargaining perspective on sovereign default also helps explain why, in addition to outright defaults (partial or complete), we include “reschedulings” in our definition of sovereign defaults. In a typical rescheduling, the debtor forces its creditors to accept longer repayment schedules and often interest rate concessions (relative to market interest rates). The ratings agencies (including Moody’s and Standard and Poor’s) rightly regard these episodes as negotiated partial defaults in which the agreed rescheduling minimizes the deadweight costs of legal fees and other expenditures related to a more acrimonious default in which a country and its creditors simply walk away from the table, at least for a time. Our data set does make a distinction between reschedulings and outright defaults, although from a theoretical perspective the two are quite similar.

  One final but critical point is this: the fact that countries sometimes default on their debt does not provide prima facie evidence that investors were irrational. For making loans to risky sovereigns, investors receive risk premiums sometimes exceeding 5 or 10 percent per annum. These risk premiums imply that creditors receive compensation for occasional defaults, most of which are only partial anyway. Indeed, compared to corporate debt, country defaults often lead to much larger recoveries, especially when official bailouts are included.