Stock Market's Response to Real Output Shocks: Connection Restored but Delayed

• Author: Numan Ülkü,Duminda Kuruppuarachchi
• Date: 01 December 2015
• DOI: http://doi.org/10.1111/irfi.12056
• Published date: 01 December 2015

Stock Market’s Response to Real

Output Shocks: Connection

Restored but Delayed*

NUMAN ÜLKÜ AND DUMINDA KURUPPUARACHCHI

Department of Accountancy and Finance, University of Otago, Dunedin, New Zealand

ABSTRACT

We propose a vector autoregression with asymmetric leads model to combine

the forward-looking, contemporaneous, and delayed responses of the stock

market to output news. Using this approach, we document that the stock

market’s connection to real output, shown by Binswanger to have been

broken since the early 1980s, has been restored after 1998, however, via a

delayed response. Subperiods mainly differ in terms of delayed response,

portraying an interesting evolution of market participants’ response to mac-

roeconomic information based on the realized persistence of output shocks.

JEL classification: G14, E32

I. INTRODUCTION

Using regressions of current stock market returns on future industrial produc-

tion (IP) growth rates and current IP growth on past returns, Fama (1990)

documents stock market’s forward-looking behavior, consistent with the

present value model.1However, Binswanger (2000) shows that this result

no longer holds since the early 1980s.2Louis and Eldomiaty (2010)

confirm Binswanger’s conclusion using a comprehensive battery of vector

autoregression (VAR) and vector error correction (VEC) specifications. Beetsma

and Giuliodori (2012) report that gross domestic product (GDP)’s negative

response to stock market volatility shocks has become smaller after the 1980s.

This note proposes the vector autoregression with asymmetric leads

(VARwAL) model to combine stock market’s forward-looking, contemporane-

ous, and delayed responses to real output shocks, enabling us to depict a

complete time profile of the stock market’s response. Using this approach, we

confirm, but also refine, Binswanger’s result. Then, we show that the full

*We thank Dorian Owen for helpful comments.

1 Several studies confirm a strong positive connection between the stock market and real output

under alternative methodologies (e.g., Lee 1992; Gallinger 1994; Choi et al. 1999; Nasseh and

Strauss 2000; Hassapis and Kalyvitis 2002).

2 Binswanger (2004a, b) reaches a similar conclusion for Japan and the aggregated G-7 European

economy. Gallinger (1994) also notes that ‘causality’ from stock returns to real output is

weaker after 1980.

bs_bs_banner

International Review of Finance, 15:4, 2015: pp. 613–622

DOI: 10.1111/irfi.12056

© 2015 International Review of Finance Ltd. 2015