EXCHANGE RATE MANIPULATION AND CONSTRUCTIVE AMBIGUITY

• Published date: 01 November 2015
• DOI: http://doi.org/10.1111/iere.12139
• Author: Ricardo T. Fernholz
• Date: 01 November 2015

INTERNATIONAL ECONOMIC REVIEW

Vol. 56, No. 4, November 2015

EXCHANGE RATE MANIPULATION AND CONSTRUCTIVE AMBIGUITY∗

BYRICARDO T. FERNHOLZ1

Claremont McKenna College, U.S.A.

I consider dynamic models in which investors are heterogeneously informed about both foreign exchange interven-

tions and exchange rate fundamentals and show that transparency sometimes exacerbates misalignment between the

exchange rate and fundamentals. Although transparency reveals information about fundamentals, it also increases the

precision of the exchange rate as a signal of those fundamentals that remain unknown (the signal-precision effect of

transparency). If a central bank announcement reveals little information about fundamentals, then this effect dom-

inates and transparency magnifies misalignment. One implication is that ambiguity can increase the effectiveness of

intervention to support a declining currency during times of crisis.

1. INTRODUCTION

Over the past decade, a growing body of evidence has demonstrated that all but a few

countries exert some control over the value of their exchange rates. According to Calvo and

Reinhart (2002), this “fear of floating” is common not only among countries that openly admit

it, but also among those that claim not to let currency prices affect policy. Just as central banks

broadly agree about the desire to control their exchange rates, they broadly disagree about

the policies that should accompany these interventions, especially with regard to transparency.

In this article, I develop dynamic models of foreign exchange intervention that address these

questions.

I focus on the issue of central bank transparency, specifically on the implications of

credible and truthful public announcements about the size and timing of foreign exchange

interventions as opposed to deliberate attempts to be secretive and create uncertainty about

those interventions. Although there are other important aspects of central bank intervention

policy, the question of transparency is among both the most important and the most disputed.

Indeed, there is extensive evidence that central banks from around the world hold opposing

views about the implications of predictability versus unpredictability and that they implement

different policies for different reasons (Canales-Kriljenko, 2003; Chiu, 2003; Archer, 2005).

Two examples from the financial crisis highlight this lack of policy consensus. Both Mexico

and Russia faced intense capital outflows and speculative pressure as the price of risky assets

throughout the world declined in the months after the collapse of Lehman Brothers in Septem-

ber 2008.2The Bank of Mexico has a longtime commitment to transparent foreign exchange

intervention, but at the height of this crisis in early February 2009, the Bank became convinced

that transparency was hurting its efforts to stabilize the peso and abruptly switched to a secretive

and purposely ambiguous policy. In that month alone, the Bank spent nearly two billion dollars

∗Manuscript received December 2011; revised June 2014.

1I am grateful to Pierre-Olivier Gourinchas for invaluable guidance on this project and also to David Ahn and Barry

Eichengreen for their many important and helpful suggestions. I also thank Christian Bauer, Philip Ernstberger, Yuriy

Gorodnichenko, Matteo Maggiori, Kristoffer Nimark, Maurice Obstfeld, Helen Popper, Demian Pouzo, and seminar

participants at UC Berkeley and Universit¨

at Trier for helpful comments. All remaining errors are mine. Please address

correspondence to: Ricardo Fernholz, Robert Day School of Economics and Finance, Claremont McKenna College,

500 E. Ninth St., Claremont, CA 91711. E-mail: rfernholz@cmc.edu.

2Between August 2008 and March 2009, both the Mexican peso and the Russian ruble lost more than one third of

their values against the U.S. dollar before eventually stabilizing at slightly higher levels.

1323

C

(2015) by the Economics Department of the University of Pennsylvania and the Osaka University Institute of Social

and Economic Research Association

1324 FERNHOLZ

of its reserves in unannounced interventions.3In this same period, the Bank of Russia fought a

protracted battle with the markets over the falling ruble. Its well-publicized attempts to initially

guide the currency to an orderly and predictable depreciation eventually gave way to a looser,

more ambiguous policy in which the target band for the ruble was substantially widened and

made more flexible.4Ultimately, the Bank of Russia’s extensive interventions contributed to a

loss of more than 200 billion dollars in foreign exchange reserves (nearly 40% of the Bank’s total

reserves) in a period of only six months. In both of these cases, policymakers appear to have been

uncertain about the best way to complement their interventions and to help effectively stabilize

and defend their currencies. In this era of enormous foreign exchange reserves and large-scale

interventions, a better understanding of the implications of these different policies is important.

The main prediction of my analysis is that central bank transparency can, in fact, magnify

any existing misalignment between the exchange rate and fundamentals. In all my models, the

equilibrium exchange rate is linear and of the form

e=fundamentals −risk premium +misalignment,(1)

so exchange rate misalignment refers to the part of the exchange rate other than fundamentals

and the risk premium.5This article’s main results state that transparency often magnifies the lat-

ter misalignment term from Equation (1). This occurs because a transparent intervention policy

improves the precision of the exchange rate as a signal of fundamentals (the signal-precision

effect of transparency) and thus compels rational Bayesian investors to weigh that public signal

more heavily in their expectations. Although transparency reveals some information about fun-

damentals (the truth-telling effect of transparency) and thus also diminishes the signal value of

the exchange rate, this extra information can be outweighed by the increased precision provided

by a public announcement. It is precisely in these cases, when central bank announcements do

not credibly reveal sufficient information about fundamentals, that exchange rate misalignment

worsens.6Figure 1 plots the relationship between exchange rate misalignment and information

revelation. As shown, transparency magnifies misalignment for low levels of information reve-

lation, but there exists a threshold at which transparency starts to reduce this misalignment. In

effect, partial transparency is worse than no transparency, whereas full transparency is best.

This conclusion has several implications. First, because central banks often intervene in

foreign exchange markets to reduce perceived misalignment between the market exchange

rate and its long-run equilibrium value (Moreno, 2005), my results imply that transparency may

sometimes undermine the effectiveness of such interventions. Central banks also intervene

to reduce exchange rate volatility (Canales-Kriljenko, 2003; Moreno, 2005), and my results

imply that transparency may sometimes undermine the effectiveness of such interventions as

well. Indeed, I show that exchange rate misalignment as described by Equation (1) above is an

important contributor to exchange rate volatility, so it follows that if transparency can increase

misalignment, then transparency can also increase volatility.

Arguably the most important implication of my results, however, is that a policy of ambiguity

will often increase the effectiveness of central bank intervention during periods of crisis and

3Although these interventions were intentionally kept secret, the Bank of Mexico did reveal their size afterward.

For a discussion of the Bank’s normally transparent policy, see Sidaoui (2005).

4In the second half of 2008, the Bank of Russia widened the target band for the ruble to 16.9% (top to bottom) via

a series of small adjustments. It then widened the band further to 28.9% in a little over one week in January 2009. Two

examples of some of the press coverage surrounding this episode are the articles “The Flight from the Rouble” and

“Down in the Dumps” from The Economist, November 20, 2008, and February 5, 2009, respectively.

5This concept of misalignment is closely related to the concept of market depth from the finance and market

microstructure literature as discussed by Vives (2008). In particular, more misalignment typically implies less market

depth.

6In all of the models I present in this article, the concept of exchange rate misalignment is closely related to the

concept of exchange rate informativeness from market microstructure theory. In most cases, more exchange rate

misalignment is equivalent to a less informative exchange rate.