Tracking exchange rate management in Latin America

• Published date: 01 April 2015
• DOI: http://doi.org/10.1016/j.rfe.2015.02.004
• Author: César Carrera
• Date: 01 April 2015

Tracking exchange rate management in Latin America☆

César Carrera

BancoCentral de Reserva del Perú(BCRP), Jr. Miroquesada441, Lima, Peru

UniversidadPeruana de CienciasAplicadas (UPC), Peru

abstractarticle info

Availableonline 13 March 2015

JEL classification:

C16

F31

F41

Keywords:

Exchangerate

Forex

Benford'sLaw

Internationalreserves

FX intervention

One wayto track exchange-rate deviationsfrom its long-runvalue is to examine numericalpatterns in exchange

rates to see if those patterns appear to have been subjected to some degree of policy management.We apply

Benford's Law to exchangerates in Latin American countries, computing and comparingthe distribution of ex-

change-rateobserved values with those of Benford'sLaw. For most cases we find that the exchange rate for the

US dollar does not satisify Benford'sLaw, however this law holds when the euro is considered. This result may

be explained by the factthat these countries are characterized forhaving different degrees of dollarizationand

interventionin the US dollar forexmarket while there is almostno policy interventionin the euro forex market.

Ourapproach is an alternativeview of how these characteristicsplaya role inducingdeviationswith respect to an

implied equilibrium exchangerate.

© 2015 ElsevierInc. All rights reserved.

1. Introduction

Understanding the implications of how countriesmanage their cur-

rency is critical to a full appr eciation of the challenges for central

bankers in Latin America. The exchange rate is a key variable for any

open economy becauseit reflects trading and financial conditions. The

value of exportsand imports, as well as capital market flows,is subject

to thevolatility of the exchangerate. The variabilityof the exchangerate

is also intrinsically linkedto a country's short-terminterest rate and in-

ternationalreservepolicies and is importantfor the determinationof in-

terestrate parity conditions.In a financial crisisin which currencies lose

value relative to the US dollar, exchange rate management issues can

come to the forefront. Moreover,in a few cases, Latin Americanecono-

mies have experienced a phenomenon called dollarization (common

use of the dollar rather than domestic currency for transactions), and

some smaller economies have fully replaced their domestic currency

and now use U.S.dollars instead.

In this research, we are interested in monitoringthe relative degree

of exchange rate management using numer ical pattern analysis. We

note that the forex (FX) market is the world's largest and most liquid

market,with trillions of dollarsbeing traded on any given daybetween

millions of parties all around th e world in many currencies. Latin

American central banks, as in the rest of the world, however,conduct

most of theirreserve activityin US dollars relative tothe local currency.

And even though currency diversification is slowly increasingin reserve

managementactivities, ourresearch focuses first on US dollar exc hange

rates, withsome later observations relative to the euro.

Our approach is to study the numerical patterns present in actual

exchange rates and to look for deviation s that are suggestive of ex-

change rate management activiti es. This approach is in contrast to

manystudies, as we will look beyondjust the accumulationor depletion

of a centralbank's internationalreserves. Specifically, we want to estab-

lish a basecase of freely-floating exchang e rates based on Benford'sLaw.

Benford'sLaw (the law of the first digit or the law of the leadingdigits)

refers to the frequency distribution of digits in many sources of data.

Under this law, the larger the digit the lessfrequent this digit is likely

to be the first digit. The implied distribution of first digits can be de-

scribed on a logarithmic weakly mon otonic distribution. Benford's

Law can also be extended for the expected distribution for the second

digitand beyond. For this law tohold, data have to be distributedacross

multiple orders of magnitude ( for example, the list of numbers that

represents the populations of U.S . cities). This law does not hold

if there is any cutoff that exclude s a portion of the underlying data

above a maximum or belowa minimum value (for example, cities de-

fined as a settlementwith a population between 30,000and 99,000).

To anticipate some of thehighlights of our results, we find definite

deviationsfrom the freely-floatingbase case using Benford's Law for se-

lected Latin American currencies relative to the US dollar, suggesting

varying degrees of currency management. The numericalpatterns for

Reviewof Financial Economics 25 (2015)35–41

☆I would like to thank Phillipe Bachetta, La wrence Christiano, Thomas Wu, Adrian

Armas, Diego Winkelried , Sofia and Catherine Carrera, Nelson Ramirez- Rondan, and

Hilda Guay for their valua ble comments and suggestions. Yrasema D ioses and Maria

Robles providedexcellent research assistance. The pointsof view expressed throughout

this document are the author's own and are not necessarily shared by the institutions

with whichhe is currently affiliated.

E-mailaddresses: cesar.carrer a@bcrp.gob.pe,pcefccar@upc.edu.pe.

http://dx.doi.org/10.1016/j.rfe.2015.02.004

1058-3300/©2015 Elsevier Inc. All rightsreserved.

Contents listsavailable at ScienceDirect

Review of Financial Economics

journal homepage: www.elsevier.com/locate/rfe