Exchange Rates and Monetary Policy Uncertainty

• Published date: 01 June 2017
• Author: PHILIPPE MUELLER,ANDREA VEDOLIN,ALIREZA TAHBAZ‐SALEHI
• DOI: http://doi.org/10.1111/jofi.12499
• Date: 01 June 2017

THE JOURNAL OF FINANCE •VOL. LXXII, NO. 3 •JUNE 2017

Exchange Rates and Monetary Policy

Uncertainty

PHILIPPE MUELLER, ALIREZA TAHBAZ-SALEHI, and ANDREA VEDOLIN∗

ABSTRACT

We document that a trading strategy that is short the U.S. dollar and long other cur-

rencies exhibits significantly larger excess returns on days with scheduled Federal

Open Market Committee (FOMC) announcements. We show that these excess returns

(i) are higher for currencies with higher interest rate differentials vis-`

a-vis the United

States, (ii) increase with uncertainty about monetary policy,and (iii) increase further

when the Federal Reserve adopts a policy of monetary easing. We interpret these ex-

cess returns as compensation for monetary policy uncertainty within a parsimonious

model of constrained financiers who intermediate global demand for currencies.

ANNOUNCEMENTS BY THE Federal Open Market Committee (FOMC) are among

the most highly anticipated events by investors around the world. These an-

nouncements, which occur regularly at prespecified dates, serve as the Federal

Reserve’s main channel for communicating its monetary policy decisions to the

market. Given the close link between currency markets and monetary policy,

it is only natural to expect that FOMC announcements can have large impacts

on exchange rates. The active nature of the currency markets (with a daily

turnover of over five trillion U.S. dollars) coupled with high market concen-

tration and participants’ ability to operate with high leverage ratios means

that even small price movements in this market can potentially translate into

economically significant effects.

∗Philippe Mueller and Andrea Vedolinare with the London School of Economics. Alireza Tahbaz-

Salehi is with Columbia University. The authors thank the Editor, Ken Singleton, an Associate

Editor, and two anonymous referees for very helpful comments and suggestions. We are also

grateful for comments received from Daron Acemoglu, Gianluca Benigno, Andrea Buraschi, Mike

Chernov, Pasquale Della Corte, Marco Di Maggio, Jack Favilukis, Fabio Fornari, Thomas Gilbert,

Christian Julliard, Roman Kozhan, Matteo Maggiori, Emi Nakamura, Ricardo Reis, H´

el`

ene Rey,

Maik Schmeling, Ali Shourideh, Vania Stavrakeva, Christian Wagner, Jonathan Wright, Stephen

Zeldes, and Irina Zviadadze, as well as from seminar participants at Columbia Business School,

Federal Reserve Bank of San Francisco, London Business School, London School of Economics,

NOVALisbon, Swiss National Bank, University of Bangor, University of British Columbia (Sauder),

University of St. Gallen, University of York,University of Zurich, the Society of Economic Dynamics

Annual Meeting (Toulouse), the Tel Aviv Finance Conference, the ECB Workshop on “Financial

Determinants of Exchange Rates,” the Imperial Hedge Fund Conference, and the UCL workshop

on the “Impact of Uncertainty Shocks on the Global Economy.”We gratefully acknowledge financial

support from the Systemic Risk Centre at LSE. Vedolin gratefully acknowledges financial support

from the Economic and Social Research Council (Grant ES/K002309/1). We have read The Journal

of Finance’s disclosure policy and have no conflicts of interest to disclose.

DOI: 10.1111/jofi.12499

1213

1214 The Journal of Finance R

In this paper, we document that announcements by the FOMC have an eco-

nomically and statistically significant impact on the excess returns of a host

of different currencies vis-`

a-vis the U.S. dollar. By relying on high-frequency

data, we document that a trading strategy that is short the U.S. dollar and

long other currencies exhibits significantly larger average excess returns on

days with scheduled FOMC announcements compared to all other days. We

also document that the excess returns earned on announcement days (i) con-

sist of a pre- as well as a postannouncement component, (ii) are higher for

currencies with higher interest rate differentials vis-`

a-vis the United States,

(iii) increase with market participants’ uncertainty about monetary policy, and

(iv) are higher when the Federal Reserve adopts a policy of monetary easing.

We interpret these findings through the lens of a parsimonious model of ex-

change rate determination in the spirit of Gabaix and Maggiori (2015), in which

constrained financiers with short investment horizons intermediate global de-

mand for currencies. These financiers can actively engage in currency trad-

ing, but have a downward-sloping demand for risk taking, which limits their

risk-bearing capacity. Such a limit can arise for a variety of reasons, such as

value-at-risk constraints or agency problems. Crucially, in addition to the “fun-

damental risk” in global demand for and supply of currencies, financiers in our

model also face “monetary policy uncertainty” due to potential future changes

in interest rates.

Using this framework, we show that an increase in uncertainty regarding

future interest rates in the United States results in higher excess returns for

other currencies: financiers are willing to engage in currency trading and to

bear this extra risk only if they are compensated accordingly with higher re-

turns. As such, all else equal, an increase in monetary policy uncertainty due

to an upcoming FOMC announcement results in the depreciation of foreign

currencies against the U.S. dollar, followed by an expected appreciation in

the future. We also establish that the increase in excess returns in response

to monetary policy uncertainty is higher for currencies with larger interest

rate differentials vis-`

a-vis the United States. This is due to the fact that, even

though an increase in the interest rate differential induces an exchange rate

adjustment, financiers’ risk-bearing constraints all but ensure that this adjust-

ment does not offset the increase in the interest rate differential one-for-one,

thus resulting in higher excess returns.

The fact that higher currency excess returns are meant to compensate fi-

nanciers for the uncertainty in monetary policy means that such returns

materialize irrespective of the interest rates set by the Fed upon the an-

nouncement. We thus interpret the impact of monetary policy uncertainty on

currency excess returns as a “preannouncement” effect. However, the actual

realization of the monetary policy shock also impacts foreign currencies’ ex-

cess returns by affecting financiers’ balance sheets, leading to what we call a

“postannouncement” effect. Indeed, we show that our model predicts that an

ex post adoption of an expansionary monetary policy (corresponding to an in-

terest rate reduction by the Fed) further increases foreign currencies’ excess

returns.

Exchange Rates and Monetary Policy Uncertainty 1215

We empirically study currency risk premia on announcement days by relying

on 20 years of high-frequency data from 1994 to 2013 for the 10 most traded

currencies. We find that, in line with our theoretical model, a simple trad-

ing strategy that is short the U.S. dollar and long the other currencies yields

significantly higher returns on announcement days compared to nonannounce-

ment days. We also document that returns earned on the eight announcement

days account for a significant fraction of currencies’ yearly excess returns. No-

tably, the large increase in average excess returns on announcement days is

not accompanied by an equally large increase in realized risk, resulting in

significantly higher Sharpe ratios on announcement days.

Since investors do not typically trade only in individual currencies but rather

go long and short portfolios of currencies simultaneously, we also test our

model’s predictions for currency portfolios sorted based on the interest rate

differential vis-`

a-vis the United States. Our empirical results indicate that ex-

cess returns earned on announcement days are larger for currency portfolios

with higher interest rates, an observation consistent with our model. In par-

ticular, we find that a portfolio consisting of currencies with low interest rates

earns an average daily return of 5.19 basis points (bps) during days when the

Federal Reserve makes an announcement, compared to an average of −0.51 bps

on all other days. This difference becomes larger (and highly statistically signif-

icant) for the portfolio consisting of high interest rate currencies, with a daily

return of 14.47 bps on announcement days compared to 1.73 bps on all other

days.

Our explanation for the large returns earned on announcement days is that

they reflect a premium for heightened uncertainty about monetary policy.Using

different proxies for monetary policy uncertainty, we find that an increase in

market participants’ uncertainty is indeed associated with higher returns on

FOMC announcement days.

We next study the intraday pattern of returns in further detail by decom-

posing currency returns into their pre- and postannouncement components.

To this end, we split the day into two nonoverlapping time windows that fall

before and after the exact time of the announcement. We find that returns

earned over both windows are larger on announcement days compared to the

corresponding windows on nonannouncement days. We also test our model’s

prediction regarding the relationship between the returns earned during the

postannouncement window and the stance of monetary policy by stratifying

our sample into easing and tightening periods depending on the policy adopted

by the Fed. Using a monetary policy indicator constructed from high-frequency

data on various interest rate futures, we find that, in line with our model’s

prediction, postannouncement returns are higher when the Federal Reserve

adopts an expansionary policy.

The observation that the stance of monetary policy and interest rate differen-

tials are tightly linked to currency excess returns means that trading strategies

that take these factors into account should exhibit higher returns compared to

simpler strategies that do not. We leverage these observations to construct

trading strategies that improve upon the simple strategy that always shorts