The assessment of the United States quantitative easing policy: Evidence from global stock markets

• Date: 01 October 2017
• Author: Jung‐Bin Su,Ken Hung
• Published date: 01 October 2017
• DOI: http://doi.org/10.1002/ijfe.1590

RESEARCH ARTICLE

The assessment of the United States quantitative easing

policy: Evidence from global stock markets

Jung‐Bin Su

1

| Ken Hung

2

1

Department of Finance, China University

of Science and Technology, No. 245, Sec. 3,

Academia Rd., Nangang Dist, Taipei

11581, Taiwan

2

A.R. Sanchez, Jr. School of Business,

Texas A&M International University,

WHTC 217A, 5201 University Boulevard,

Laredo, TX, USA

Correspondence

Ken Hung, A.R. Sanchez, Jr. School of

Business, Texas A&M International

University, WHTC 217A, 5201 University

Boulevard, Laredo, TX, USA.

Email: ken.hung@tamiu.edu

JEL Codes: C13; C32; G15; E52

Abstract

This study assesses the performance of the quantitative easing policy imple-

mented by the United States (US) on the stock markets with a framework of

structure break. The empirical results show that the business cycle or the value

of gross domestic production has a negative impact on the stock markets for

most of the countries even if both gross domestic production and many stock

prices are procyclical. Moreover, the purchases of US Treasury securities and

mortgage‐backed securities respectively affect the stock markets synchronously

and laggardly. Notably, they both have a positive impact on the stock markets

during the study period. Finally, during the after structure break period, the

volatility can be easily affected by bad news, and the investors have the lower

profit or even the greater loss and bear the greater risk and variation of risk

owing to the global financial crisis caused by the US. On the basis of the above

findings, some policy implications are offered in this study.

KEYWORDS

business cycle, mortgage‐backed securities, quantitativeeasing, structure break, Treasurysecurities

1|INTRODUCTION

The global financial crisis was triggered by the accumulat-

ing losses on the United States (US) subprime mortgages

in the summer of 2007; it really started to show its effects

in mid‐2007 and into 2008. This crisis is regarded as the

worst financial crisis since the Great Depression of the

1930s by many economists. It played a significant role in

the failure of key businesses and a downturn in economic

activity that led to global recession from 2008 and 2012

and contributed to the European sovereign debt crisis.

During this period, global stock markets fell, large finan-

cial institutions collapsed or were bought out, and govern-

ments in even the wealthiest nations had to come up with

rescue packages to bail out their financial systems. Conse-

quently, the US, the United Kingdom (UK), and the

Eurozone implemented quantitative easing (QE)

1

policies

when their risk‐free short‐term nominal interest rates

were either at, or close to, zero. For example, in the US,

the Federal Reserve started to buy $600 billion of mort-

gage‐backed securities (MBS) in late November 2008

(QE1); in November 2010, the Federal Reserve announced

a second round of QE (QE2) and had bought $600 billion

of Treasury securities (TS) by the end of the second quar-

ter of 2011. A third round of QE (QE3) was announced on

September 13, 2012. The Federal Reserve decided to

launch a new $40 billion per month, open‐ended, bond‐

purchasing programme of agency MBS. In addition, the

Federal Open Market Committee announced that it

would in all probability maintain the federal funds near

zero at least through 2015. Hence, owing to the global

financial crisis and the subsequent QE policy being imple-

mented by the US and other countries, some structure

break (SB) may appear at the global stock market during

the abovementioned period.

Now, we turn to the recent literatures on QE.

Kobayashi, Spiegel, and Yamori (2006) employed a simple

capital asset pricing model specification to examine the

Received: 28 September 2015 Revised: 6 July 2017 Accepted: 16 September 2017

DOI: 10.1002/ijfe.1590

Int J Fin Econ. 2017;22:319–340. Copyright © 2017 John Wiley & Sons, Ltd.wileyonlinelibrary.com/journal/ijfe 319

impact of QE policy executed by the Bank of Japan on

Japanese bank equity values during the period from the

start of QE in March 2001 to the end of 2004. They found

that raw bank returns usually responded positively to the

expansion of the QE programme by increases in the ceil-

ing on long‐term Japanese government bond. Eric

Girardin and Moussa (2011) used the Markov switching

vector autoregressive methodology to examine the effec-

tiveness of Japanese monetary policy and quantify the

effect of QE on Japan's activity and prices during the

“lost”decade and found that QE was able not only to pre-

vent recession and deflation but also to provide consider-

able stimulation to both output and prices. The output

and prices are measured by the industrial production

and consumer price index, respectively. Lyonnet and Wer-

ner (2012) utilized the general‐to‐specific model selection

methodology to examine the effectiveness of the QE pol-

icy implemented by the Bank of England on the nominal

gross domestic production (GDP) growth since March

2009. They found that the QE announced in March 2009

had no apparent effect on the UK economy, whereas there

existed a stable relationship between a lending aggregate

and nominal GDP. Kagraoka and Moussa (2013)

employed the macrodynamic yield model proposed by

Nelson and Siegel (1987) to examine the effects of the

QE strategy on the yield curve in Japan. They showed that

the monetary policy shock has a significant effect on the

level factor of yield curve only during the high interest

rate period. Schenkelberg and Watzka (2013) used the

structural vector autoregressive approach to assess the

macroeconomic effects of QE measures adopted by the

Bank of Japan during a prolonged zero lower bound epi-

sode over the period from March 1995 to September

2010. They showed that the QE shock led to a significant

and transitory decrease in long‐term interest rates and

positively and significantly but transitorily affected indus-

trial production. Nakazono and Ueda (2013) utilized a

term structure model to evaluate the effects of the QE pol-

icy on market expectations in Japan from 2001 to 2006 via

using individual survey data on inflation expectations, as

well as interest rate expectations, and found that there

existed a kinked relationship between interest rate expec-

tations and inflation rate expectations at around the 0%

threshold level of inflation expectations. Karras (2013)

used the approach of Cover (1992) and Karras (1996a,

1996b) to examine whether the effects of QE are subject

to the asymmetries by estimating monetary base shocks

and their effects on real GDP and industrial production

in US and found that with or without QE, monetary base

contractions have larger effects than monetary base

expansions, and the effectiveness of monetary base shocks

declines with their size, indicating that there is evidence

of sign and size asymmetries, respectively. Fratzscher,

Lo Duca, and Straub (2013) used the linear regression set-

ting to examine the impact of the US Federal Reserve'sQE

policies on portfolio flows and asset prices both in the US

and globally and found that the Federal Reserve's mea-

sures in the early phase of the crisis (QE1) were highly

effective in lowering sovereign yields and raising equity

markets, especially in the US relative to other countries.

The Federal Reserve's measures since 2010 (QE2) boosted

equities worldwide, whereas they had muted impact on

yields across countries. Lo Duca, Nicoletti, and Vidal Mar-

tínez (2016) used the panel linear regression setting to

examine the impact of unconventional monetary policy

in the US on global bond markets and found that pur-

chases and holdings of MBS and TS by the Federal

Reserve have a strong impact on gross corporate bond

issuance across advanced and emerging economies.

Most of literature on QE in recent years has focused

on its effects by examining macroeconomic and financial

variables via the linear regression of a mean equation.

The above variables include the inflation (e.g., Kagraoka

& Moussa, 2013; Nakazono & Ueda, 2013), industrial pro-

duction (e.g., Eric Girardin & Moussa, 2011; Karras, 2013;

Schenkelberg & Watzka, 2013), GDP (e.g., Karras, 2013;

Lyonnet & Werner, 2012), consumer price index (e.g., Eric

Girardin & Moussa, 2011; Schenkelberg & Watzka, 2013),

bank equity value (e.g., Kobayashi et al., 2006), exchange

rate (e.g., Schenkelberg & Watzka, 2013), interest rate (e.

g., Kagraoka & Moussa, 2013; Nakazono & Ueda, 2013;

Schenkelberg & Watzka, 2013), and others (e.g.,

Fratzscher et al., 2013; Lo Duca et al., 2016). Unlike previ-

ous studies, this study uses the AR(1)‐QGARCH‐SGED

model

2

with an SB to assess the performance of QE policy

implemented by the US via examining the impact of busi-

ness cycle

3

and the unconventional monetary policy on

global stock markets for the two subperiods (i.e., the

periods of before SB and after SB). The unconventional

monetary policy is the purchases of TS and MBS by the

Federal Reserve in US, whereas the global stock markets

are represented by the nine stock indices in the following

eight countries: US, Belgium, Germany, UK, Japan,

Singapore, China, and India. Moreover, under the frame-

work of the parametric techniques, the long position in‐

sample value‐at‐risk (VaR) forecasts based on the above

model are calculated, and the value of VaR and the loss

caused by the violation of VaR model are respectively

used to explore the risk and the loss suffered by the inves-

tors in the above eight countries during the before and

after SB periods. Our results show that the business cycle

or the value of GDP has a negative impact on the stock

market for most of the countries even if both the GDP

and stock price are procyclical for most of cases. More-

over, the purchases of TS and MBS in the US have a pos-

itive impact on the stock markets during the before and

320 SU AND HUNG