Nominally Denominated Sovereign Debt,
Risk Shifting, And Reputation
Herschel I. Grossman and John B. Van Huyck
[ Introduction | Conclusion | References | John's Web ]
Abstract: This paper analyzes a reputational equilibrium in a model in which nominally denominated sovereign debt serves to shift risk associated with the unpredictability of tax revenues from the sovereign to its lenders. The analysis answers the following set of related questions: Why would a sovereign refrain from inflating when faced with servicing a large quantity of nominal debt? If a sovereign does not plan to use inflation to repudiate its nominal debts, why would it want to issue nominal debt in the first place? What are the distinguishing features of those sovereigns who are willing and able to issue nominal debts?
© by the authors. All rights reserved.
Over the course of history, certain sovereigns--the USA and UK are prominent examples--have regularly issued debts that are nominally denominated in the units of the sovereign's own currency. Although this practice is far from universal, it is nevertheless a puzzling phenomenon, because such debt involves promises to pay a fixed number of units of an asset whose real value the sovereign itself controls through its power to issue fiat money. Presumably, a sovereign can issue nominally denominated debt only if lenders believe that the sovereign will not use its power to inflate in order to repudiate its debts. In practice, sovereign issuers of nominally denominated debt regularly validate this belief, even to the extent of deflating in periods, typically postwar periods, following the issuance of large quantities of debt.
These observations suggest the following set of related questions:
1. Why would a sovereign refrain from inflating, or even deflate, when faced with servicing a large quantity of nominally denominated debt?
2. If a sovereign does not plan to use inflation to repudiate its nominally denominated debts, why would it want to issue such debt in the first place?
3. What are the distinguishing features of those sovereigns who are willing and able to issue nominally denominated debt?
To answer these questions, this paper develops a theory that focuses on the unpredictability of tax revenues, net of nondiscretionary expenditures for such purposes as law, order, and national defense, and the desirability of shifting some of the associated risk from the sovereign to its lenders. If contracts contingent on net tax revenues are sufficiently costly to write, verify, and enforce, shifting risk by making debt servicing directly contingent on net tax revenues is not feasible. In this situation, if the price level and real net output are inversely related and real net output and net tax revenues are positively related, then nominally denominated debt can help to achieve the desired risk shifting. [Real net output, like net tax revenues, excludes expenditures necessary for the survival of the sovereign. The variability of real net output and net tax revenues can reflect, in addition to shocks to real gross output, fluctuations in the intensity of external or internal threats to national security.]
The sovereign's ability to issue nominally denominated debt requires that lenders expect the sovereign to refrain from attempting to produce unexpected inflation in order to repudiate its debts. The analysis assumes that, because monetary policy is verifiable, lenders can differentiate unexpected inflation that is associated with negative shocks to aggregate net output from unexpected inflation produced by a sovereign's efforts to repudiate its nominally denominated debts. In this theory, the choice by the sovereign to validate lender expectations about monetary policy--by, for example, maintaining or resuming convertibility to an external standard like gold--depends on the sovereign's concern for its trustworthy reputation. A trustworthy reputation is valuable because it provides continued ability to issue nominally denominated debt.
The analysis derives a reputational equilibrium in which the amount of nominally denominated debt is such that the short-run benefits to the sovereign from deliberately creating unexpected inflation are smaller than the long-run costs from the loss of a trustworthy reputation. Any unexpected inflation that occurs in the reputational equilibrium is the result of negative shocks to real net output and serves to shift risk from the sovereign to its lenders. The analysis shows that, depending on such factors as the variability of net tax revenues, the probability of the sovereign surviving in power, and the probability of the lenders forgetting a repudiation, the equilibrium amount of nominally denominated debt may or may not be adequate to permit efficient risk shifting. In extreme cases, reputational forces can fail to support anything other than a trivial equilibrium with zero nominal debt.
In the existing theoretical literature on nominally denominated sovereign debt, Lucas & Stokey (1983) develop a normative model in which efficiency requires that the sovereign commit itself to service its debts according to a contingent nominal servicing schedule and also commit monetary policy to achieve a specific price path. Such commitments would effectively convert nominally denominated debts into real contingent claims. Neither Lucas & Stokey nor the further discussion in Lucas (1986) explores the implications of the unenforceability of such commitments.
In contrast, Bohn's (1988) analysis of nominally denominated debt as a device to shift risk emphasizes the potential time inconsistency of monetary policy commitments. To rationalize a finite inflation rate, Bohn assumes that sovereigns are averse to inflation, whether expected or not, and derives a myopic equilibrium in which the expected inflation rate is sufficiently high and the amount of debt sufficiently small that the marginal cost of unexpected additional inflation equals the marginal benefit from reduced real debt servicing. Bohn's analysis does not consider the possible role of reputational forces in supporting a larger amount of debt.
In Grossman & Van Huyck (1987), we develop the idea of sovereign debt as a contingent claim that allows excusable default in verifiably bad states of the world, and we derive a reputational equilibrium in which the sovereign always chooses to meet its contingent real servicing obligations rather than to repudiate its debts. The present paper considers a complementary problem in which claims that are directly contingent on net tax revenues are assumed to be infeasible. This infeasibility motivates the issuance of nominally denominated debt and the associated use of inflation that is unexpected, but not deliberate, to enforce the contingent claim.
[ Top | Introduction | Conclusion | References | John's Web ]
This paper has analyzed a reputational equilibrium in a model in which, because net real tax revenues and unexpected inflation are inversely related, nominally denominated sovereign debt serves to shift the risk associated with net real tax revenue from the sovereign to its lenders. The model assumes that lenders can differentiate unexpected inflation that is associated with negative shocks to aggregate output from unexpected inflation that would result if a sovereign set its exchange rate opportunistically, without regard for its announced exchange rate policy, in an attempt to repudiate its debts. A trustworthy reputation is valuable because it provides continued ability to issue nominally denominated debt. In the reputational equilibrium, the amount of nominally denominated debt is such that the short-run gains from repudiation of the debt through unexpected devaluation and inflation are smaller than the long-run costs from the loss of a trustworthy reputation. Accordingly, the sovereign always resists the temptation to repudiate.
Depending on the value to the sovereign of a trustworthy reputation, the equilibrium amount of debt can be sufficient to permit efficient and complete risk shifting or it can be constrained in a way that limits risk shifting. The factors that enhance the value of a trustworthy reputation and, thus, enable the sovereign to issue a large amount of nominally denominated debt include high variability of real net national product and real net tax revenues, high risk aversion, a low rate of pure time preference for the sovereign, a high probability per period that the sovereign will survive in power, and a high probability per period that lenders would not forget a past repudiation.
In a recent revisionist history, DeLong (1986) argues that Britain's net sterling denominated asset position after the First World War was positive and so large that its return to gold at the prewar parity was actually an opportunistic policy decision. This decision, however, was not historically unique, and DeLong does not suggest that similar conclusions would apply to Britain's return to gold at the prewar parity after the Napoleonic Wars or the United States' return to gold after the Civil War and the First World War. The analysis in the present paper implies that the primary and common consideration in each of these deflations was the desire to maintain a trustworthy reputation, without which efficient financing of the next war would not be possible. A reading of the public debate over each of these policy decisions seems consistent with this hypothesis. Indeed, the present analysis suggests that the main realized benefit of Britain's return to gold after the First World War at the prewar parity was to enable Britain to mobilize the resources to survive the Second World War.
[ Top | Introduction | Conclusion | References | John's Web ]
H. Bohn, "Why Do We Have Nominal Government Debt?" Journal of Monetary Economics, 21, January 1988.
J. B. DeLong, "Was Britain's Return to Gold Worth It?", unpublished, November 1986.
H.I. Grossman, "A Generic Model of Monetary Policy, Inflation, and Reputation." NBER Working Paper No. 2239, revised July 1987.
H.I. Grossman and J.B. Van Huyck, "Seigniorage, Inflation, and Reputation," Journal of Monetary Economics, 18, July 1986.
H.I. Grossman and J.B. Van Huyck, "Sovereign Debt as a Contingent Claim: Excusable Default, Repudiation, and Reputation," NBER Working Paper No. 1673, revised December 1987.
R.E. Lucas, Jr., "Principles of Fiscal and Monetary Policy," Journal of Monetary Economics, 17, January 1986.
R.E. Lucas, Jr. and N.L. Stokey, "Optimal Fiscal and Monetary Policy in an Economy Without Capital," Journal of Monetary Economics, 12, July 1983.
[ Top | Introduction | Conclusion | References | John's Web ]