Is finance a veil? Lead‐and‐lag relationship between financial and business cycles: The case of China

DOIhttp://doi.org/10.1111/eufm.12193
Published date01 September 2019
Date01 September 2019
DOI: 10.1111/eufm.12193
ORIGINAL ARTICLE
Is finance a veil? Lead-and-lag relationship
between financial and business cycles: The case
of China
Chung-Hua Shen
1,2
|
Jun-Guo Shi
3,
*
|
Meng-Wen Wu
4
1
Institute of Banking and Money,
Nanjing Audit University, Nanjing,
People's Republic of China
2
Department of Finance and Banking,
Shih Chien University, Taiwan, Republic
of China
Email: chshen01@ntu.edu.tw
3
College of Finance and Statistics,
Hunan University, Changsha 410079,
People's Republic of China
Email: shijunguo@hnu.edu.cn
4
Department of Business Administration,
National Taipei University, Taiwan,
Republic of China
Email: mwwu@mail.ntpu.edu.tw
Abstract
This study examines the lead-and-lag relationship between
financial cycles (FCs) and business cycles (BCs) by using
Chinese provincial data. We construct FCs of the financial
sector on the basis of three financial variables: credit-to-
GDP (gross domestic product) ratios, house prices, and
equity prices. We use the panel dynamic logit model to
investigate the lead-and-lag effect between two sectors.
Results show that each province has its own unique FCs and
BCs. Hence, financial policies should be different in
dissimilar provinces. Next, we find that FCs lead BCs and
not vice versa. Furthermore, the leading effect is stronger in
rich provinces than in poor areas.
KEYWORDS
business cycle, credit-to-GDP ratio, direct financing ratio, financial
cycle, panel dynamic logit model
JEL CLASSIFICATION
E32, E44, G21, P34
We are grateful for the constructive comments from the Editor John Doukas, the Associate Editor, and the three anonymous
reviewers for constructive comments and suggestions that have helped improve our paper significantly. We appreciate
helpful comments from the participants at the European Financial Management 2017 Symposium on Finance and Real
Economy at Xiamen University, China on April 79, 2017. All errors remain our own.
*Present address: College of Finance and Statistics, Hunan University, Changsha 410079, People's Republic of China.
Eur Financ Manag. 2018;135. wileyonlinelibrary.com/journal/eufm © 2018 John Wiley & Sons, Ltd.
|
1
978 © 2018 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/eufm Eur Financ Manag. 2019;25:978–1012.
1
|
INTRODUCTION
The 2008 financial crisis triggered a major rethink in macroeconomics (Drehmann, Borio, &
Tsatsaronis, 2012). Moreover, the crisis led to an intensive debate about the influence of the financial
crisis on broad economies. Dominant pre-crisis paradigms view finance largely as a sideshow to
macroeconomic fluctuations. Also, finance is usually seen effectivelyas a veil. As a first approximation,
this factor can be ignored when seeking to understand business fluctuations (Woodford, 2003).
However, the crisis demonstrated that this presumption is dangerously false. The past several years have
witnessed recessions in all advanced economies and many emerging markets. A common feature of
these recessions is that they are accompanied by various financial disruptions, including severe
contractions in credit and sharp declines in asset prices. Borio, Disyatat, and Juselius (2013) argued that
information about the financial boom and bust should be considered in improving measures of potential
output and output gaps. In addition, recessions associated with financial disruptions are generally
deeper and long lasting (Claessens, Kose, & Terrones, 2012; Jordà, Schularick, & Taylor, 2013).
These developments have led to an intensive debate in the profession about the links between
finance and macroeconomics. The debate on the direction of causality between financial development
and economic growth has long been an ongoing study (King & Levine, 1993a, 1993b; Shen & Lee,
2006). Past studies commonly support the positive view that finance boosts economic growth (Beck,
2014; King & Levine, 1993a, 1993b). However, some researchers find the opposite to be true (Ghartey,
2015; Vazakidis & Adamopoulos, 2010), or, even worse, that finance development reduces economic
growth (Shen & Lee, 2006; Shen, Lee, Chen, & Xie, 2011).
1
Against this backdrop, studies on interactions between the two sectors have been propelled to the
forefront of research (Caballero, 2010; Woodford, 2010). The current study aims to re-examine this
issue by using a different set of variables, namely, the financial cycle (FC) and the business cycle (BC),
using data from 31 Chinese provinces. In particular, given that our two cycles are discrete numbers, we
study whether boom leads the expansion and bust leads the recession, and vice versa, which is different
from the literature on financial development and economic growth that focuses on the link between the
two continuous variables. By knowing the leadlag relation between FCs and BCs, governments can
better introduce financial, monetary, and fiscal regimes. Moreover, enterprises can better arrange
investment and business activities, which will help to mostly reduce the negative impact of FCs on BCs
or BCs on FCs. Such arrangements can also help avoid a severe recession caused by a financial crisis.
Studying the interaction between the two cycles using China's provincial
2
data is important because the
country has become the second largest economy since 2010.
3
In addition, China's economy currently
affects many countries significantly (Mirza, Narayanan, & Leeuwen, 2014).
Our study first dates the busts and booms of FCs. Economists have been investigating methods of
dating the expansion and recession of BCs for more than 100 years; however, the dating approach of
FCs is still in its infancy. Claessens, Kose, and Terrones (2011) presented the concept of FCs, which are
separately measured by the three circular movements of credit scale, house prices, and equity prices.
They used Bry and Boschan's (1971) approach, which was later extended by Harding and Pagan
1
Shen et al. (2011) found an inverted U-shaped curve relationship between financial development and economic growth.
Before the turning point, financial development positively affects economics, after which financial development decreases
the economic growth.
2
For convenience, we define provinceas representing provincial areas, including provinces, autonomous districts, and
municipalities.
3
Refer to the website: http://www.chinanews.com/fortune/2011/02-15/2844193/html
2
|
SHEN ET AL.SHEN ET AL.
|
3
979
(2002), to date the peaks and troughs of the cycles of these variables. Egert and Sutherland (2012) also
dated the cycles of real share prices, real house prices, and real credit. However, these studies do not
develop a method to date the peak and trough of the aggregate FC index, hereafter referred to as the FC
index, of the whole financial sector. Drehmann et al. (2012) advanced the field by innovatively
constructing an FC index from the credit-to-GDP (gross domestic product) ratio, credit scale, house
prices, and equity prices. To the best of our knowledge, the BC has been widely investigated, whereas
the FC has been ignored because studies commonly argue that the financial sector reflects the
movement of the real sector. However, our empirical results provide a different inference from this
common belief. Our evidence suggests that the financial sector is not a veil but has real effects on the
real world that FCs lead the real sector.
We construct an FC index for each Chinese province by incorporating the three individual financial
variables of credit-to-GDP ratio, housing price growth, and equity prices (Drehmann et al., 2012).
However, we suggest using different proxies for credit and equity prices when considering unique
features of Chinese provinces. First, studies that select credit measure in the credit-to-GDP ratio
overwhelmingly use aggregate claims on the private sector by deposit money banks,as collected
from the International Financial Statistics (line 32d). However, this commonly used credit measure
does not appropriately evaluate the Chinese credit market for two reasons. First, this credit measure
focuses on bank lending to the private sector, whereas more than half of bank lending is to state-owned
enterprises in China (Cull & Xu, 2000). Second, this credit measure stresses the lending provided by
depository banks; by contrast, increasing amounts of lending in China are provided by the non-bank
sector, which is termed the shadow bank(Lu, Guo, Kao, & Fung, 2015; Shen, Lee, Wu, & Guo,
2016). Considering the two weaknesses, the commonly used credit measure is substantially
underestimated compared with the true credit in the system. Hence, the Chinese government developed
a new credit measure, which is referred to as aggregate financing to the real economy,hereafter
referred to as aggregate financing, to measure the aggregate credit in the Chinese market. This new
credit measure includes lending to the state-owned sector and lending provided by the non-bank sector.
The present study uses this aggregate financing as a proxy for credit measure. The results are
dramatically different; the FC positively leads the BC when using aggregate financing, whereas the FC
negatively leads the BC when using the conventional credit measure.
Collecting equity price for each province is also a challenge. The stock exchange market is usually
designed for a nation and not a province; thus, we lack a stock index for each province.
4
Hence, we
adopt the direct financing ratio of non-financial institutions, which is equal to the sum of stock
financing and bond financing over aggregate financing. In this mechanism, stocks and bonds, including
convertible bonds, are the main financial instruments. The direct financing ratio of non-financial
institutions is highly related to equity price because the stock financing ratio is higher and fluctuates to
a larger degree than the bond financing ratio.
5
We adopt the novel panel dynamic logit (PDL) model to estimate the lead-and-lag relationship
between the two cycles because our two cycles are binary numbers for the data of each province.
Bartolucci and Nigro (2010a) emphasized that the conventional panel logit yields biased estimates
when lagged binary dependent variables are present on the right side of a regression. They developed a
panel dynamic model in logit to improve this bias. We first examine our issue using a sample of
31 provinces. Then, we conduct two types of robust testing to examine the robustness of our results.
4
Two stock exchange markets (i.e., Shanghai and Shenzhen stock exchanges) are available nationwide in China.
5
Refer to the Chinese Regional Financial Operation Report by the People's Bank of China. Website: http://www.pbc.gov.
cn/zhengcehuobisi/125207/125227/125960/126049/index.html
SHEN ET AL.
|
3
T
b
t
g
p
s
u
T
t
f
a
p
p
p
r
v
f
t
w
f
t
e
o
t
S
r
2
2
G
d
2
T
F
(
6
e
c
n
4
S
2
|
SHEN ET AL.
980

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT