Exchange Rates and Fundamentals: A General Equilibrium Exploration

• Date: 01 February 2021
• Published date: 01 February 2021
• DOI: http://doi.org/10.1111/jmcb.12698

DOI: 10.1111/jmcb.12698

TAKASHI KANO

Exchange Rates and Fundamentals: A General

Equilibrium Exploration

Engel and West (2005) show that the observednear random-walk behavior

of nominal exchange rates is an equilibrium outcome of a partial equilib-

rium asset approach when economic fundamentals follow exogenous rst-

order integrated processes and the discount factor approaches one. In this

paper, I arguethat the unit market discount factor creates a theoretical trade-

off within a two-country general equilibrium model. The unit discount fac-

tor generates near random-walk nominal exchange rates, but it counterfac-

tually implies perfect consumption risk sharing and at money demand.

Bayesian posterior simulation exercises, based on post-Bretton Woodsdata

from Canada and the United States, reveal difculties in reconciling the

equilibrium random-walk proposition within the canonical model; in par-

ticular, the market discount factor is identied as being much smaller than

one. A relative moneydemand shock is identied as the main driver of nom-

inal exchange rates.

This is a revised version of a paper previously circulated under the title of “Exchange Rates and Fun-

damentals: Closing a Two-country Model.” I would like to thank the editor, Kenneth West; anonymous

referees, Kosuke Aoki, Martin Berka, Dave Cook, Mario Crucini, Charles Engel, Hiroshi Fujiki, Ippei

Fujiwara, Shin-ichi Fukuda, Ichiro Fukunaga, Shinsuke Ikeda, Susumu Imai, Kazuko Kano, Munechika

Katayama, Junko Koeda, Oleksiy Kryvtsov,Takushi Kurozumi, Enrique Martínez-García, Kiminori Mat-

suyama, ToshiMukoyama, Tomoyuki Nakajima, Jim Nason, Makoto Nirei, Michel Normandin, Yoshiyasu

Ono, Ricardo Reis, Makoto Saito, Moto Shintani, Etsuro Shioji, Christoph Thoenissen, and TakayukiTsu-

ruga; and seminar participants at Hitotsubashi University,the Institute for Monetary and Economic Studies

at the Bank of Japan, the Institute of Social and Economic Research at Osaka University, KyotoUniver-

sity, the National Graduate Institute of Policy Studies, the University of Tokyo, Victoria University of

Wellington, the University of Auckland, the Australian National University, the Reserve Bank of New

Zealand, the 2013 Summer Workshop on Economic Theory at Hokkaido University, the 2014 Canadian

Economic Association Meetings, the 2014 Japanese Economic Association Spring Meetings, the 2014

International Conference of the Society for Computational Economics (CEF 2014), the Conference on

Macroeconomic Issues at Sogang University,the second international workshop for nancial markets and

nonlinear dynamics (FMND 2015), and the 9th International Conference on Computational and Financial

Econometrics (CFE 2015) for their discussions and useful suggestions. Hiroshi Morita provides excel-

lent research assistance. I wish to thank the Seimeikai foundation, the Hitotsubashi Institute for Advanced

Study,and a grant-in-aid for scientic research from the Japan Society for the Promotion of Science (num-

bers 24330060, 17H02542, and 17H00985) for their nancial support. I am solely responsible for any

errors and misinterpretations in this paper.

T K is at Graduate School of Economics at Hitotsubashi University (E-mail:

tkano@econ.hit-u.ac.jp).

Received March 4, 2016; and accepted in revised form January 22, 2020.

Journal of Money, Credit and Banking, Vol. 53, No. 1 (February 2021)

© 2020 The Ohio State University

96 :MONEY,CREDIT AND BANKING

E31, E37, F41

Keywords: E31, E37, F41

Keywords: exchange rate, present-value model, economic fundamental,

random walk, two-country model, incomplete market, cointegrated TFPs,

perfect risk sharing.

F     exchange rates sys-

tematically beat a naïve random-walk model in terms of out-of-sample forecast per-

formance. Since the study of Meese and Rogoff (1983), this robust empirical property

of nominal exchange rate uctuations has stubbornly resisted theoretical attempts to

understand the behavior of nominal exchange rates as an equilibrium outcome of ob-

served fundamentals. Many open-economy dynamic stochastic general equilibrium

(DSGE) models also suffer from this problem. They fail to generate random-walk

nominal exchange rates along an equilibrium path because their exchange rate fore-

casts are closely related to observed macro-economic fundamentals.1

Engel and West (2005, hereafter EW) shed a new light on this vexing empirical

property of nominal exchange rates. Using a partial equilibrium asset approach, in

which nominal exchange rates reect the present discounted values of expected fu-

ture economic fundamentals, they show that if economic fundamentals are integrated

of order one (hereafter I(1)) and the discount factor approaches one, then a nominal

exchange rate follows a near random-walk process in equilibrium. This equilibrium

random-walk property is attributable to the fact that only the Beveridge–Nelsontrend

components of the I(1) economic fundamentals are reected in the present-value cal-

culation at the limit of the unit discount factor.

The I(1) property is likely to hold in actual data of economic fundamentals, par-

ticularly monetary fundamentals, which are identied in neoclassical open-economy

monetary models as the most important driver of nominal exchange rates.2Subse-

quent studies, thus, have focused on the empirical validity of the requirement that

the discount factor be close to one. Based on the monetary fundamentals, Sarno and

Sojli (2009) and Balke et al. (2013, hereafter BMW) identify a discount factor us-

ing partial equilibrium asset models similar to that of EW, examine data on different

currencies and sample periods, and estimate the discount factor to be distributed near

to one. In particular, the Bayesian unobservable component (UC) model of BMW

estimates relative money demand shocks to be a dominant underlying driver of a

long sample of the British pound/U.S. dollar rate. This empirical nding supports

the conjecture of EW that persistent unobservable economic fundamentals and a dis-

count factor near one play signicant roles in near random-walk nominal exchange

rates.

1. Engel (2014) provides a recent survey of studies on nominal exchange rates.

2. For example, the reduced-form regression exercises of Mark (1995), Rapach and Wohar (2002),

Mark and Sul (2001), and Cerra and Saxena (2010) show that the monetary fundamentals contain eco-

nomically signicant predictive components of nominal exchange rates. The model of this paper would

offer a general equilibrium foundation for these empirical results.