Stock Market's Response to Real Output Shocks in China: A VARwAL Estimation

• Published date: 01 September 2023
• Author: Numan Ülkü,Kexing Wu
• Date: 01 September 2023
• DOI: http://doi.org/10.1111/cwe.12500

©2023 Institute of World Economics and Politics, Chinese Academy of Social Sciences

China & World Economy / 1–25, Vol. 31, No. 5, 2023 1

Stock Market’s Response to Real Output Shocks

in China: A VARwAL Estimation

Numan Ülkü, Kexing Wu*

Abstract

This paper studies the connection between the stock market and real output in China

and compares it with benchmark countries, employing a novel vector autoregression

ZLWKDV\PPHWULFOHDGV 9$5Z$/PRGHO,W PDNHVWZRFRQWULEXWLRQV )LUVWLW¿QGVWKDW

WKHWLPH SUR¿OHRIWKH &KLQHVHVWRFNPDUNHW¶VUHVSRQVH WRUHDO RXWSXWVKRFNVVXJJHVWV

QRHYLGHQFHRI DGLVWRUWHGUHODWLRQVKLSGXH WRPDQLSXODWLRQRI&KLQHVH UHDORXWSXWGDWD

RUGRPLQDWLRQRI WKH&KLQHVH VWRFNPDUNHWE\ LQGLYLGXDOLQYHVWRUV5DWKHU WKH&KLQHVH

stock market is relatively more responsive to real output, in line with the larger share

RIPDQXIDFWXULQJ LQWKH&KLQHVH HFRQRP\(OHFWULFLW\RXWSXW DQGLQGXVWULDOSUR¿WVWZR

GLৼHUHQWOHVVPDQLSXODEOHWLPHVHULHV \LHOGVLPLODUUHVXOWV6HFRQG LWSUHVHQWVWKH¿UVW

XVHRI 9$5Z$/LPSXOVHUHVSRQVHV WRGHWHFW VWRFNPDUNHW EXEEOHV9$5Z$/FDSWXUHV

WKHEXEEOHLQ&KLQDVXFFHVVIXOO\2YHUWKHIXOOVDPSOHSHULRG&KLQD¶VVWRFNPDUNHW

DSSHDUVWRKDYHEHHQOHVVSURQHWREXEEOHVWKDQWKH86VWRFNPDUNHW

Keywords: China, real output shocks, stock market, VARwAL model

JEL codes: C58, E44, G14, P34

I. Introduction

Starting with Fama (1990), the dynamic interaction between stock market returns and

UHDORXWSXWJURZWKKDVEHHQRQH RIWKHSULQFLSDOLQGLFDWRUVXVHGE\¿QDQFLDOHFRQRPLVWV

to assess the connectedness of the stock market to the real economy, which is a key

dimension of market efficiency. The interaction between the stock market and real

output in China, the world’s second largest economy and stock market, is particularly

interesting for two reasons. First, given suspicions raised regarding the reliability of

Chinese real output statistics,1 does the interaction between the stock market and real

*Numan Ülkü (corresponding author), Visiting Professor, Institute of Economic Studies, Charles University,

Czechia. E-mail: numan.ulku@fsv.cuni.cz, numan.ulku@outlook.com; Kexing Wu, Postgraduate student,

School of Commerce, University of South Australia, Australia. E-mail: Todd1221@hotmail.com. This work was

supported by the Cooperatio Program, Research Area Economics, at Charles University, Czechia.

1Allegations of China’s meddling in output data have been raised by analysts at global investment banks and

have been discussed in the literature by Rawski (2001), Holz (2003, 2004, 2014), Chow (2006), Fernald et al.

(2013), and Lyu et al. (2018), among others.

© 2023 The Authors. China & World Economy published by John Wiley & Sons Australia,

Ltd on behalf of Institute of World Economics and Politics, Chinese Academy of Social Sciences.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License,

which permits use and distribution in any medium, provided the original work is properly cited,

the use is non-commercial and no modifications or adaptations are made.

Numan Ülkü, Kexing Wu / 1–25, Vol. 31, No. 5, 2023

©2023 Institute of World Economics and Politics, Chinese Academy of Social Sciences

2

RXWSXWLQ &KLQDGL൵HUIURP WKDWLQ FRPSDUDEOHFRXQWULHVLQ ZD\VWKDW DUHV\PSWRPDWLF

of data meddling? Second, given the dominant role of individual investors and the

influence of government interventions,2 is the Chinese stock market more prone to

misalignments from macroeconomic fundamentals? Or, does it serve the same role as a

leading economic indicator as in other countries?

In this article, we employ a novel vector autoregression with asymmetric leads

(VARwAL) model to describe the Chinese stock market’s response to real output news

on China and compare with several benchmark economies – the US, Japan, South

Korea, and India. The VARwAL model, which adds exogenous leads of one variable in

a VAR model, is a special case of noncausal and mix VARs.3 The VARwAL’s advantage

LVWKDW LWSURYLGHVD FRPSOHWHWLPH SUR¿OHRIWKHVWRFN PDUNHW¶VUHVSRQVHWR UHDORXWSXW

shocks, combining the forward-looking (anticipatory) and lagged responses of the stock

market in a single impulse response function, conveniently interpreted in an event-

study format. The VARwAL model is particularly useful in assessing the stock market’s

timeliness in responding to output shocks and comparing across countries. It enables us

to detect any dissimilarity between China and benchmark countries at any point in the

WLPHSUR¿OHRIWKHVWRFNPDUNHW¶VUHVSRQVH

Studying the stock market’s connection with the real output on China using the

VARwAL model enables us to make two contributions to the financial economics

literature. First, the granularity of the complete time profile of the stock market’s

response to real output shocks provided by the VARwAL output allows us to

conclusively assess whether alleged manipulation of China’s output data, dominance

of individual investors, or government interference in the stock market alters the joint

dynamics of the stock market and real output. Second, we offer a new tool for stock

PDUNHWEXEEOHGHWHFWLRQ XVLQJWKH9$5Z$/LPSXOVHUHVSRQVHVDQGZH SUHVHQWLWV¿UVW

successful use in detecting the 2015 bubble in China’s stock market.

1. Background literature

The main studies of the interaction between the stock market and real economic activity

in developed markets (Fama, 1990; Lee, 1992; Choi et al., 1999; Binswanger, 2000,

2004; Gallinger, 2004; Shanken and Weinstein, 2006; Laopodis, 2011) confirmed the

leading role of the stock market as an indicator, whereas Binswanger (2000, 2004)

reported a loss of connectedness of the stock market to the real economy starting

2See the review provided by Allen et al. (2019).

3A noncausal VAR contains lead terms of the endogenous variables. Standard VARs that only contain lags are

called causal VARs in this context. A mix 9$5 LVDVSHFL¿FDWLRQ WKDWFRQWDLQVERWK ODJDQGOHDG WHUPV6HH

Lanne and Saikonnen (2013), Lanne and Nyberg (2015), Lanne and Luoto (2016), and Davis and Song (2020).

© 2023 The Authors. China & World Economy published by John Wiley & Sons Australia,

Ltd on behalf of Institute of World Economics and Politics, Chinese Academy of Social Sciences.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License,

which permits use and distribution in any medium, provided the original work is properly cited,

the use is non-commercial and no modifications or adaptations are made.