The Macroeconomic Implications of a Key Currency

Abstract

We identify incentives generated by the Bretton Woods II system that may have contributed to the sub-prime liquidity crisis now working its way through the international monetary system. We then evaluate the persistent conjecture that the liquidity crisis is or will become a balance of payments crisis for the United States. Given that it happens, the additional costs associated with a sudden stop of net capital flows to the United States could be quite substantial. But we observe that emerging market governments have continued to acquire US assets even as yields have fallen, and the incentives for continuing to do so remain strong. Moreover, the Bretton Woods II system, which has clearly been the most resilient of the forces driving current markets, continues to generate low real interest rates in industrial countries and growth in emerging markets that will help limit the damage from the liquidity crisis. Copyright © 2009 Blackwell Publishing Ltd.

Full text

Page 1

NBER WORKING PAPER SERIES
THE MACROECONOMIC IMPLICATIONS OF A KEY CURRENCY
Matthew Canzoneri
Robert E. Cumby
Behzad Diba
David Lopez-Salido
Working Paper 14242
http://www.nber.org/papers/w14242
NATIONAL BUREAU OF ECONOMIC RESEARCH
1050 Massachusetts Avenue
Cambridge, MA 02138
August 2008
We are grateful for comments from the participants of a conference in honor of Mike Wickens at
the University of York. The opinions expressed here are solely those of the authors and do not necessarily
reflect the views of the Board of Governors of the Federal Reserve System, nor of anyone else associated
with the Federal Reserve System, nor of the National Bureau of Economic Research.
© 2008 by Matthew Canzoneri, Robert E. Cumby, Behzad Diba, and David Lopez-Salido. All rights
reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission
provided that full credit, including © notice, is given to the source.

Page 2

The Macroeconomic Implications of a Key Currency
Matthew Canzoneri, Robert E. Cumby, Behzad Diba, and David Lopez-Salido
NBER Working Paper No. 14242
August 2008
JEL No. F3,F4,F41
ABSTRACT
What are the macroeconomic consequences of the dominant role of the dollar in the international monetary
system? Here, we present a calibrated two country model in which exports are invoiced in the key
currency, and government bonds denominated in the key currency are held internationally to facilitate
trade. Domestic government bonds and money are held in each country to facilitate domestic transactions.
Our model generates deviations from uncovered interest parity that are as volatile as some empirical
estimates, but much too small by others. Our model also speaks to some other empirical anomalies,
such as the Backus - Smith puzzle. Shocks affecting asset supplies – such as bond financed tax cuts,
and open market operations – have large effects in our model because they generate non-Ricardian
changes in household wealth. Generally, shocks emanating from the key currency country do more
to destabilize the world economy than equal sized shocks coming from the other country. Similarly,
monetary and fiscal policy innovations in the key currency country are more potent than those in the
other country. On the other hand, the key currency country is more vulnerable to financial market
turbulence, such as a sell off of key currency bonds, which can lower consumption dramatically.
Matthew Canzoneri
Department of Economics
Georgetown University
Washington, DC 20057
canzonem@georgetown.edu
Robert E. Cumby
Georgetown University
School of Foreign Service
Washington, DC 20057-1045
and NBER
cumbyr@georgetown.edu
Behzad Diba
Department of Economics
Georgetown University
Washington, DC 20057
dibab@georgetown.edu
David Lopez-Salido
Federal Reserve Board
Mail Stop 71
20th and C Streets, N.W.
Washington, DC 20551
David.J.Lopez-Salido@frb.gov

Page 3

1 This phrase is sometimes attributed to Charles de Gaulle, but is evidently due to Giscard
d’Estaing, in 1965. See Gourinchas and Rey (2007), footnote 4, for an account.
2 Aliber (1964) cites numerous political discussions of the costs and benefits of being the
key currency country, and he discusses academic arguments to the effect that the U.S. might be
constrained in its ability to lower its payments deficit and in its choice of policies to achieve full
employment.
3 Goldberg and Tille (2008; Table 1) report that large shares of various countries’ imports
and exports are invoiced in dollars. Lane and Milesi-Ferretti have a series of papers
documenting the proclivity of foreigners for dollar assets; see for example Lane and Milesi-
Ferretti (2007).
1. Introduction
The implications of a key currency were hotly debated in the 1950's and 60's, both in
academia and in political circles. Valery Giscard d’Estaing, while still a French finance minister,
asserted that the U.S. enjoyed an “exorbitant privilege” because of the role of the dollar in the
international monetary system.1 The U.S. could finance a current account deficit by issuing dollar
assets at low rates of interest. Why? Because a large fraction of international trade was invoiced
in dollars, and dollar assets were desired by importers and exporters to facilitate trade; dollars were
also desired by central banks as official reserves.2
The dollar continues to play a dominant role in the international monetary system.3 And the
dollar figures prominently in current policy discussions: the U.S. has financed most of its recent
current account deficits by issuing dollar denominated debt, and there is much speculation in the
press about what might happen if foreign dollar holders decided to diversify into other currencies.
In this paper, we build a model in which international trade is invoiced in the key currency,
and bonds denominated in the key currency are held to facilitate international trade (and as official
reserves). Since key currency bonds provide liquidity services, they are held at a low rate of
interest; this is source of the “exorbitant privilege” accruing to the key currency country. Back of

Page 4

2
4 At the end of 2007, foreign holdings of U.S. Treasury securities amounted to about 90
percent of quarterly U.S. consumption. We have computed the average liquidity premium on
Treasury securities to be about 0.0065 per quarter. These imply a value of the exorbitant
privilege of 0.585 percent of consumption.
5 The literature prior to the mid 1970's generally assumed that bonds were imperfect
substitutes; see Branson and Henderson (1985) for a survey. Then, following the popularity of
the Dornbusch (1976) model, and empirical work in the late 1970's and early 1980's, modelers
gravitated towards the perfect substitutes assumption.
6 Wickens (2008) provides a brief review of this literature. Lewis (1995) provides a more
extensive review of the earlier literature.
the envelope calculations suggest that this privilege is worth slightly more than a half of a percent
of U.S. consumption each year.4 And, one of our tasks will be to investigate the potential
consequences of a sudden reversal of the recent buildup of foreign official holdings of dollars.
The recent academic literature has shied away from these issues, and one reason for this is
undoubtedly that currently popular models almost invariably assume that bonds denominated in
different currencies are perfect substitutes: portfolio managers do not distinguish between them if
their expected returns (when evaluated in a common currency) are equal. Expected returns equalize
in these models, resulting in the standard uncovered interest parity (or UIP) condition.5 These
models are not suitable for analyzing the asymmetries implied by a key currency: there is no special
role for dollar denominated assets.
While macroeconomic modeling has generally assumed that bonds are perfect substitutes,
a lively empirical literature has questioned the validity of the UIP condition. There appear to be
large, frequent and persistent deviations from UIP in the data.6 And, there have been some efforts
to introduce deviations from UIP into theoretical models, either because the data required it or the
question being analyzed required it.
Deviations from UIP have been introduced in a variety of ways. Blanchard et al (2005)

Page 5

3
7 See Pesenti (2008).
8 See Kollmann (2002, 2005), and McCallum and Nelson (2000).
9 Lewis (1995) surveys early contributions to this literature. See Alverez, Atkeson and
Kehoe (2007) for a more recent application.
10 Holmstrom and Tirole (1998) provide a rationale for government provision (and
management) of liquidity. They show that with aggregate uncertainty, the private sector cannot
adequately supply its liquidity needs. The government can raise welfare by issuing bonds, which
will command a liquidity premium over private debt.
follow a much earlier literature that simply assumes bonds denominated in different currencies are
imperfect substitutes and posits asset demand functions that depend on relative rates of return and
financial wealth. They expand upon a model by Kouri (1976) to discuss the causes and long run
consequences of recent U.S. current account deficits. Some versions of the IMF’s GEM assume
portfolio adjustment costs,7 and this creates portfolio balance effects in the model’s dynamics. Other
modelers have simply added an exogenous UIP shock, calibrated to match the volatility of exchange
rates and (ex-post) UIP deviations observed in the data.8 And some take a more structural approach,
modeling the deviations from UIP as endogenous, time varying risk premia.9
In this paper, we model deviations from UIP as endogenous, time varying liquidity premia.
More specifically, we build a fully articulated NOEM model in which government bonds
denominated in different currencies provide different kinds of transactions services, making them
imperfect substitutes. There are two countries in our model – the key currency country and a
country representing the rest of the world. Both countries price their exports in terms of the key
currency. Consequently, households in each county use the key currency bond to facilitate their
purchases of imported goods; and within each country, households use their own government bond
and money to facilitate their purchases of domestic goods.10 Key currency bonds also serve as