Page 1 of 24
Archives of Business Research – Vol. 12, No. 2
Publication Date: February 25, 2024
DOI:10.14738/abr.122.16548.
Chiang, T. C., & Chen, P.-Y. (2024). Evidence of Geopolitical Risk and Economic Policy Uncertainty on Equity Prices in Emerging
Markets. Archives of Business Research, 12(2). 105-128.
Services for Science and Education – United Kingdom
Evidence of Geopolitical Risk and Economic Policy Uncertainty
on Equity Prices in Emerging Markets
Thomas C. Chiang
Drexel University, Department of Finance
Pei-Ying Chen
College of Finance, Feng Chia University
No.100, Wenhwa Road, Taichung City, Taiwan, R.O.C.407224
ABSTRACT
This paper presents evidence of the impact of changes in downside risk(ΔVaRt),
economic policy uncertainty (∆EPUt
), the geopolitical risk (ΔGPRt
) in six Asian
equity markets while controlling for COVIDE-19. Applying the month data for the
period of 1990.01– 2021.12 to Asian markets, evidence shows that national stock
returns are negatively correlated with ΔVaRt
, ∆EPUt ∆GPRt and ∆GPRt
2
, suggesting
that these risk/uncertainty factors harmfully affect stock markets. Evidence shows
that a rise in U.S. geopolitical risk adversely spillovers to Asia stock markets.
Spreads of COVID-19 damage local stock returns via interacting with changes in EPU
and GPR. Purpose: Present evidence of COVIDE-19 and the impact of changes in
downside risk(ΔVaRt), economic policy uncertainty (∆EPUt
), the geopolitical risk
(ΔGPRt
) with different forms of nonlinear specifications in six Asian equity markets.
Design/methodology/approach: The rolling regression method is used to derive
time-varying coefficients of investors’ response to changes of policy uncertainty,
which is sensitive to react to other uncertainty variables. Findings: Evidence shows
that national stock returns are negatively correlated with ΔVaRt
, ∆EPUtand∆GPRt
2
,
suggesting that these risk/uncertainty factors adversely affect local stock markets.
In addition, the evidence shows that a rise in U.S. geopolitical risk (∆GPRt
2,US) can
adversely spillover to domestic stock markets. A spillover effect is found from
COVID_19 development. Research limitations/implications: The findings are based
on data from six Asian markets. More data may be collected, which would expand
the analysis to a broader coverage of the global market. Practical implications:
Evidence suggests that in addition to the risk from financial markets, the
geopolitical risk (GPR) and COVID-19 are important factors in determining whether
a risk premium should be received. From a policymaker’s point of view, geopolitical
stability is the foundation for maintaining financial market stability.
Originality/value: This paper presents evidence on the time-varying behavior of
investors’ response to ∆EPUt
. In addition to ∆EPUt and the ΔGPRt
, the nonlinear
components of ∆GPRt
2
, U.S.∆GPRt
2,US and interaction to COVID-19 have a harmful
effect on national stock prices in Asian markets.
Keywords: Downside risk, Economic policy uncertainty, Geopolitical risk, Safe haven,
Asset pricing, Asian markets, JEL Classification: G10, G11, G14, G15
Page 2 of 24
106
Archives of Business Research (ABR) Vol. 12, Issue 2, February-2024
Services for Science and Education – United Kingdom
INTRODUCTION
The mean-variance approach has been one of most crucial relations in analyzing capital
investments. The attractiveness of this model lies in its simplicity in focusing on two key
parameters: return and variance (risk). An advance of this approach is due to Merton’s
(1973,1980) Intertemporal Capital Asset Pricing model (ICAPM), which implies a positive risk- return tradeoff relation in stock markets. Studies from French, Schwert, and Stambaugh (1987),
Ghysels, Santa-Claraand Valkanov (2005), Kinnunen (2014), and Cederburg and O’Doherty
(2019) provide supporting evidence on the positive tradeoff hypothesis. However, an extensive
literature review shows that the conditional volatility and conditional mean move in an
opposite direction as reported in the publications of Glosten, Jagannathan and Runkle (1993),
Lettau and Ludvigson (2001) and Brandt and Kang (2004). Therefore, the evidence on the issue
of risk-return relation displays mixed signs.
As noted by Guo and Whitelaw (2006), the conventional approach in conducting empirical
analysis is likely to miss a component for the changes in investment opportunities that are
expounded in Merton’s (1973) model. It follows that some of the model specifications are likely
to commit an omitted variable problem. The researchers thus have reoriented their direction
by focusing on the impact of economic policy uncertainty (EPU) on stock prices. Bloom (2009,
2014) and Baker, Bloom and Davis (2016) and Chiang (2019a) demonstrate that heightened
EPU will impede economic activity and delay business investment decision, which would
produce a harmful impact on stock prices. The studies by Caggiano, Castelnuovo and Groshenny
(2014), Li (2017), Christou, Cunado, Gupta and Hassapis (2017) and Chiang (2020) provide
supporting evidence for the impact of EPU on the stock returns.
Despite the researchers’ positive contributions, their attention is placed on the risk associated
with financial sources. However, it is well recognized that geopolitical stability is the foundation
for maintaining financial market stability. Investors tend to pay attention to information that is
revealed in the daily news associated with nuclear tensions, war threats, terrorist threats, etc.
All of these activities and their related uncertainties no doubt have a profoundly disturbing
investors’ psychology and, hence, have significant impact on investors’ behavior, which would
affect their judgment in making portfolio decisions. Along this line, Pástor and Veronesi (2013)
suggest that political uncertainty commands a risk premium. In their recent study of U.S.
political policy risk on world asset prices, Brogaard, Dai, Ngo, and Zhang et al. (2020) state that:
“political uncertainty can spillover from one country and influence the financial outcomes in
other countries”. Caldara and Iacoviello (2022) provide various components of measuring
geopolitical risk and conflict of related tensions.
The outbreak of COVID-19 has threatened human health and impeded economic activities since
the spring of 2020. Its ensuing damages can be seen in the significant financial losses suffered
by many investors in a short period of time (Zhang et al., 2020). In response, governments have
provided assistance, by using fiscal programs and easing monetary supplies. These government
interventions, however, can induce further policy uncertainty that interacts with the often
rising and erratic spread of coronavirus infections. Thus, an investigation of stock returns and
policy uncertainty needs to factor into the interacting effects of EPU and GPR with changes in
COVID-19 infections.
Page 3 of 24
107
Chiang, T. C., & Chen, P.-Y. (2024). Evidence of Geopolitical Risk and Economic Policy Uncertainty on Equity Prices in Emerging Markets. Archives of
Business Research, 12(2). 105-128.
URL: http://doi.org/10.14738/abr.122.16548
This paper expands on the empirical studies of stock return behavior in several ways. First,
instead of using conditional volatility to model financial risk as a way of explaining stock market
variations, this study employs VaR as a measure of risk because of its information content,
which covers the variance and higher stock return moments in describing asset pricing. Second,
this study employs a news-based EPU index to describe stock market performance. Third, this
paper uses a news-based GPR index to capture the impact arising from military-tensions,
nuclear threats, or changes of governments that cause a harmful effect on stock prices. Fourth,
to factor in the global influence, this study also includes U.S. GPR in the test equation. This study
also incorporates the interacting effects of EPU and GPR with changes in COVID-19 infections.
Lagged dividend yield and time series patterns are used as control variables. Thus, the model
in this study is more inclusive because the arguments are not restricted to a single measure of
risk variable, an approach that is typically used in conventional studies.
The remainder of this paper is organized as follows. Section 2 provides a brief literature review
that leads to the specification of an econometric model for testing stock returns. Section 3
presents the testable hypothesis. Section 4 describes the data and variable definitions in the
empirical analysis. Section 5 presents empirical evidence for the test equation and extends the
analysis by applying rolling repression estimations. Section 6 tests the impact of uncertainty by
incorporating the C) VID-19 effect on stock returns. Section 7 concludes the empirical findings.
LITERATURE REVIEW
Empirical analysis of risk-return relation usually employs a regression model by relating stock
return to a measure of risk. French et al. (1987) pioneers this research and demonstrate the
(excess) stock return is positively correlated with the conditional variance or standard
derivation. Tests of the null hypothesis performed by Ghysels et al. (2005), (2010), Bali and
Engle (2010) and Chiang, Li and Zheng (2015) that relate the conditional mean of stock returns
to the conditional variance and find evidence of a positive and statistically significant relation.
However, Nelson (1991), Glosten et al. (1993), Brandt and Kang (2004) and Ang, Hodrick, Xing
and Zhang (2006) test the same hypothesis and fail to find supporting evidence.
The cause of these mixed signs stems from a number of reasons, but most noticeable are the
measure of risk variable and a possible missing variable. Focusing on the characteristic that
conditional variance models fail to deal with the extreme tail risk has prompted researchers to
shift their attention and consider downside risk as measured by using Value-at-Risk (VaR) in
their approach. Bali, Demirtas and Levy (2009) report that downside risk can significantly
predict stock returns in the U.S. market; Chen, Chiang, and Härdle (2018) document that stock
returns are positively correlated with the lagged downside risk in G7 markets, and Chiang
(2019b) uncovers evidence that lagged downside risk presents a significant predictive power
in Chinese stocks.
A shortcoming of using historical information to construct a risk variable is based on a premise
of parametric stability, which usually fails to capture updated information. A news-based index
such as an EPU index, however, appears to more effectively reveal on-going disruptive events
that could affect decision-making. Following the studies by Bloom (2009, 2014), Baker et al.
(2016), we employ a news-based EPU index that highlights three words: {Economic, Policy,