Real exchange rate dynamics: Relative importance of Taylor‐rule fundamentals, monetary policy shocks, and risk‐premium shocks

• Author: Cheolbeom Park,Chang‐Jin Kim
• Published date: 01 February 2019
• DOI: http://doi.org/10.1111/roie.12372
• Date: 01 February 2019

Rev Int Econ. 2019;27:201–219. wileyonlinelibrary.com/journal/roie 2018 John Wiley & Sons Ltd

|

201

Received: 17 April 2017

|

Revised: 22 May 2018

|

Accepted: 5 June 2018

DOI: 10.1111/roie.12372

ORIGINAL ARTICLE

Real exchange rate dynamics: Relative importance

of Taylor‐rule fundamentals, monetary policy

shocks, and risk‐premium shocks

Chang‐Jin Kim1

|

Cheolbeom Park2

1Department of Economics,University of

Washington, Seattle, Washington

2Department of Economics,Korea

University, Seoul, Korea

Correspondence

Cheolbeom Park, Department of Economics,

Korea University, Anamro 145, Seongbuk‐

gu, Seoul 02841, Korea.

Email: cbpark_kjs@korea.ac.kr

JEL classification: E52, F31, F41

Abstract

We first show that the solution to the real exchange rate

under the Taylor rule with interest rate smoothing can

have two alternative representations—one based on a

first‐order difference equation and the other based on a

second‐order difference equation. Then, by comparing

error terms from these two alternative representations and

analyzing their second moments, we evaluate the relative

importance of Taylor‐rule fundamentals, monetary policy

shocks, and risk‐premium shocks in the dynamics of the

real exchange rate. Empirical results suggest that the risk‐

premium shock is the largest contributor to real exchange

rate movements for all the countries examined, with the

Taylor‐rule fundamentals and monetary policy shocks

playing a limited role. These results are robust to various

alternative sets of parameter values considered for the

Taylor rule with interest rate smoothing.

1

|

INTRODUCTION

Inability to show a stable empirical link between the exchange rate and macroeconomic fundamen-

tals has been a well‐known puzzle since the work of Meese and Rogoff (1983). Obstfeld and Rogoff

(2001) refer to this as the “exchange‐rate disconnect puzzle.”1 However, recent studies have shown

some progress in relating the exchange rate to macroeconomic fundamentals. A growing body of liter-

ature shows the relevance and importance of Taylor‐rule fundamentals in the dynamics of the exchange

rate. For example, Benigno (2004) theoretically demonstrates that the persistence of the real exchange

rate is related to the degree of interest rate smoothing in the Taylor rule. Engel and West (2005, 2006),

Clarida and Waldman (2007), Mark (2009), Molodtsova and Papell (2009), and Kim, Fujiwara, Hansen,

202

|

KIM and PARK

and Ogaki (2015) report some empirical success explaining the dynamics of the exchange rate under

the Taylor rule. Given the success of the Taylor‐rule‐based models, this paper quantifies the relative

importance of Taylor‐rule fundamentals, monetary policy shocks, and risk‐premium shocks in real

exchange rate fluctuations within a Taylor‐rule‐based model of real exchange rate.2

The importance of monetary policy shocks in the real exchange dynamics has typically been inves-

tigated within structural vector autoregression (VAR) frameworks, with a focus on the exchange rate’s

delayed response to monetary policy.3 However, the results based on variance decompositions seem

to be sensitive to the identifying assumptions employed. For example, while Clarida and Gali (1994),

Juvenal (2011), and Faust and Rogers (2003) report a very small contribution of monetary policy

shocks to variations in the exchange rate, Eichenbaum and Evans (1995) and Rogers (1999) demon-

strate that monetary policy shocks account for a considerable portion of real exchange rate move-

ments. Based on structural estimation of a structural new Keynesian model, Bergin (2006) claims that

monetary policy shock, rather than risk‐premium shock, is the primary driver of exchange rates. He

argues that more than 50 percent of the variations in the exchange rate come from monetary policy

shocks. More recently, using a state‐space framework to model both the predictable and unpredictable

components of fundamentals, Balke, Ma, and Wohar (2013) show that directly observed monetary

fundamentals and money demand shifters contribute most to movements in exchanges rates, with de-

viations from uncovered interest parity or risk premium shocks playing a relatively minor role.

In this paper, we first present two alternative representations of the real exchange rate under the

Taylor rule with interest rate smoothing—one based on a first‐order difference equation solution and

the other based on a second‐order difference equation solution. We do not claim that our model has

multiple solutions but show that the unique solution can be expressed by two alternative combinations

of observable and unobservable variables. We then use the difference between these two alternative

representations to quantify the relative contributions of the Taylor‐rule fundamentals, monetary policy

shocks, and risk‐premium shocks to real exchange rate variations. Our approach is distinctive in that it

does not require any of the identifying assumptions often employed in structural VAR approaches or

the state‐space approach of Balke et al. (2013). We employ a simple Taylor‐rule‐based present‐value

model of real exchange in which we allow for interest rate smoothing and deviations from the UIP.

Methodologically, we estimate the present value of the Taylor‐rule fundamentals by employing VAR,

as in Engel and West (2006). Using the estimated present value of the fundamentals, we then calculate

the model‐based exchange rates and error terms from the two alternative representations of the real

exchange rate. By comparing individual components from and evaluating the second moments of each

of these error terms, we calculate approximate measures for the relative importance of the Taylor‐rule

fundamentals, monetary policy shocks, and risk‐premium shocks.

Our empirical results suggest that risk‐premium shocks are most important in explaining the dynamics

of real exchange rates, explaining 56 to 84 percent of the variations in the real exchange rate. Conversely,

we find that the contributions of monetary policy shocks and Taylor rule fundamentals are limited, with

monetary policy shocks explaining 6 to 17 percent and Taylor‐rule fundamentals accounting for up to 18

percent of real exchange rate variations. Such a limited contribution of monetary policy shocks is com-

parable to the findings of Clarida and Gali (1994), Juvenal (2011), and Faust and Rogers (2003), but it

is in sharp contrast to the findings of Eichenbaum and Evans (1995), Rogers (1999), and Bergin (2006).

This paper is organized as follows. In Section 2, we provide two alternative representations of the

solution to a Taylor‐rule‐based real exchange rate model under interest rate smoothing and discuss

the construction of error terms from the two alternative representations. In Section 3, we explain

how to evaluate the relative importance of the Taylor‐rule fundamentals, monetary policy shocks, and

risk‐premium shocks based on the empirical evaluation of the model implied by the two alternative

solutions. Section 4 provides our empirical results, and Section 5 discusses the conclusion of the paper.