The spillover effects of U.S. monetary policy on emerging market economies

• Published date: 01 July 2019
• Author: Geun‐Young Kim,Hail Park,Peter Tillmann
• Date: 01 July 2019
• DOI: http://doi.org/10.1002/ijfe.1720

RESEARCH ARTICLE

The spillover effects of U.S. monetary policy on emerging

market economies

Peter Tillmann

1

| Geun‐Young Kim

2

| Hail Park

3

1

Faculty of Economics and Business

Studies, Justus Liebig University Giessen,

Giessen, Germany

2

Research Department, Bank of Korea,

Seoul, Republic of Korea

3

Department of International Business

and Trade, Kyung Hee University, 26,

Kyungheedae‐ro, Dongdaemun‐gu, Seoul

02447, Republic of Korea

Correspondence

Hail Park, Department of International

Business and Trade, Kyung Hee

University, 26, Kyungheedae‐ro,

Dongdaemun‐gu, Seoul 02447, Republic of

Korea.

Email: hailpark@khu.ac.kr

JEL Classification: E58; F32; F42

Abstract

The recent implementation of unconventional monetary policies in advanced

economies and the preparations for an eventual return to normalization have

renewed the interest in spillover effects of monetary policy on emerging

market economies. This paper estimates a series of VAR‐X models for a set of

10 emerging economies, that is, VARs in which U.S. policy enters exogenously.

The contribution of this paper is (a) to use an identified shock component

of the U.S. (shadow) federal funds rate as a consistent policy instrument for

conventional and unconventional policies, (b) to account for changes in the

transmission of U.S. monetary policy over time, and (c) to quantify asymmetries

in the transmission of tightening and easing shocks. The results point to sub-

stantially nonlinear and asymmetric spillover effects, which pose challenges

to policymakers.

KEYWORDS

Asymmetries, conventional–unconventional monetary policy, emerging markets, spillovers, VAR‐X

model

1|INTRODUCTION

Monetary policy responses to the global financial crisis can

shed light on the spillover effects of monetary policy to

emerging economies. When the Federal Reserve (“Fed”)

adopted a series of unconventional monetary policies

known as quantitative easing (“QE”) and forward guidance

in 2008, the side effects were transmitted through global

financial markets. In particular, emerging economies felt

the surge in liquidity after the adoption of QE. In 2014,

the Fed has eventually unwound its asset purchases

(“tapering”) after announcing the tapering in 2013 and is

expected to “lift‐off”the federal funds rate in the near

future, that is, the Fed is about to normalize its monetary

policy soon.

Spillover effects pose major challenges to policymakers

in emerging countries (Q. A. Chen, Filardo, He, & Zhu,

2015; J. Chen, Mancini‐Griffoli, & Sahay, 2014; Lavigne,

Sarker, & Vasishtha, 2014; MacDonald, 2015; Tillmann,

2015). The flood of liquidity under QE is believed to lead

to an easing of monetary conditions, a surge in asset prices,

and domestic currency appreciations. Furthermore, abun-

dant liquidity is raising financial stability concerns. Many

central banks and regulators have responded to this by a

monetary tightening and/or the use of macroprudential

policy tools. The reversal of capital flows after 2014 has

led to a sharp depreciation and a drop in asset prices in

emerging market economies. In order to choose the right

policy response, understanding the nature of these spill-

overs is essential.

This paper presents a large collection of findings on the

nature of spillover effects. We empirically quantify the

recent spillover effects and make four contributions. First,

we focus on the effects of both conventional and uncon-

ventional monetary policy. This is accomplished by using

the identified shock component of the (shadow) federal

funds rate as a measure of U.S. monetary policy, which

allows us to estimate a consistent model under both policy

Received: 15 May 2017 Revised: 23 March 2018 Accepted: 4 January 2019

DOI: 10.1002/ijfe.1720

Int J Fin Econ. 2019;24:1313–1332. © 2019 John Wiley & Sons, Ltd.wileyonlinelibrary.com/journal/ijfe 1313

regimes. Second, we study the notion of a “New Normal”,

that is, a change in the pattern of spillovers after the finan-

cial crisis by comparing a precrisis and a postcrisis regime.

Third, we allow for tightening and easing measures to have

different effects. This enables us to quantify the asymmet-

ric nature of the spillover effects. Fourth, we relate the

strength of the spillover effects to the macroeconomic fun-

damentals of emerging market economies. As a result, we

can identify which particular form of weakness gives rise

to a large exposure to U.S. policy shocks.

Our main empirical tool for the analysis of policy

transmission is a vector autoregression (VAR) model of

standard financial variables of emerging markets, which

we augment by introducing an exogenous change in U.

S. monetary policy. The fact that U.S. policy enters exog-

enously transforms the model into a VAR‐X model. In

order to focus on unexpected changes in U.S. monetary

conditions, we estimate an auxiliary VAR model for the

U.S. economy and feed the identified policy shock into

the VAR‐X model for emerging market economies. From

the perspective of a prototypical emerging economy, U.S.

monetary policy can clearly be considered exogenous.

Rather than estimating a panel model for a large set of

emerging market economies, this paper estimates a series

of country‐specific time‐series VAR‐X models. We think

this approach is better able to shed light on the large

degree of heterogeneity across countries.

We find that spillovers are divergent. Both the sign and

the size of spillovers changed over regime. We do not find,

however, that spillovers are systematically higher post‐

2008. There are not only structural breaks over time but

also asymmetries as regards the sign of the policy shock

itself. In many countries, the spillovers after a tightening

shock and an easing shock in the U.S. differ in terms of

their absolute magnitude and their dynamics. From them,

fundamentally weak economies appear to suffer from

stronger spillovers than fundamentally strong economies.

The results are relevant for policymakers in emerging

economies: Our findings suggest that not all spillovers are

equal. To the extent spillovers are asymmetric, the appro-

priate policy response cannot be linear. For example, if

macroprudential policies are considered to offset the con-

sequences of unwarranted spillover effects, these instru-

ments have to be adjusted to be an efficient tool to

contain the effects of U.S. policy shocks. In light of our

results, the need to fine‐tune macroprudential policies

challenges the efficient design of appropriate policies.

The interest rate lift‐off and the subsequent return to an

interest rate policy are likely to lead to renewed tensions

in emerging market economies—which will put policies

to a test. Our results show that strong macroeconomic

fundamentals are the best way to make emerging econo-

mies less vulnerable to spillovers.

The main contribution of this paper is the use of an

estimation framework that (a) allows for a consistent

comparison of the effects of conventional and unconven-

tional policies and (b) offers a maximum degree of flexi-

bility. This flexibility pertains to the distinction between

tightening and easing shocks and the role of structural

breaks. On element of this, among many others, is the

use of a consistent notion of “shocks.”

This paper proceeds as follows. Section 2 introduces

the VAR‐X model, which is our main modelling frame-

work. The main results are presented in Section 3.

Section 4 offers a concluding summary and draws policy

implications.

2|AVAR‐X MODEL OF

MONETARY SPILLOVERS

To estimate the spillover effects of U.S. monetary policy

on emerging economies, we estimate a series of VAR‐X

models, that is, VAR models augmented by an exoge-

nous variable. Shocks to the exogenous variable are

the focus of our study. VAR‐X models have recently

been used in the literature: Bassett, Chosak, Driscoll,

and Zakrajsek (2010) feed an exogenous credit supply

shock in an otherwise standard VAR model of the

U.S. economy to study the effects of an exogenous

change in lending. T. Wu and Cavallo (2012) include

an exogenous oil price shock in a VAR system. In a

paper similar in focus to ours, Miyajima, Mohanty, and

Yetman (2014) estimate a panel VAR model of Asian

economies where monetary policy in the U.S. enters as

an exogenous variable. Compared with their paper, we

offer a unifying perspective on conventional and uncon-

ventional monetary policy and also allow for shifts in

the strength of policy transmission over time. We also

prefer a country‐by‐country perspective over a panel in

order to link the strength of the transmission to eco-

nomic fundamentals.

In our model, the vector of the first differences of n

endogenous variables, Y

t

, is determined as follows

Yt¼A0þALðÞYtþb×ΔRUS

tþεt;

where A(L) is a polynomial in the lag operator. A mea-

sure of U.S. monetary policy is included as an exogenous

variable, ΔRUS

t. The impact effect of U.S. policy on the

endogenous variables is reflected by the vector b.

1

Because the stance of U.S. policy is clearly exogenous

from the perspective of an emerging market economy,

and we are interested in the endogenous responses to exog-

enous policy only, there is no need for further identifying

assumptions. Hence, we do not intend to identify other

1314 TILLMANN ET AL.