key: cord-0956062-s1o65u0r
authors: Ongan, Serdar; Gocer, Ismet; Ongan, Ayse
title: Pandemic‐induced economic policy uncertainty and US stock exchanges
date: 2022-03-11
journal: J Public Aff
DOI: 10.1002/pa.2799
sha: 86ee6e789457d1982a92c9631133f9d7710e0892
doc_id: 956062
cord_uid: s1o65u0r

This study investigates the impacts of pandemic‐induced economic policy uncertainties (PIEPU) on the S&P500, Nasdaq‐100, and Dow Jones indexes (stock returns). To this aim, for the first time, newly created IDEMV (the Infectious Disease Equity Market Volatility index (henceforth, PIEPU index) is used. The Autoregressive Distributed Lag (ARDL) model and the Toda and Yamamoto (Journal of Econometrics, 1995, 66, pp. 225–250) causality test are applied for the 2009–2020 period. Empirical findings indicate that rises in the PIEPU index lead to falls of only the S&P500 and Dow Jones indexes. Corporations in the tech‐heavy Nasdaq100 index do not negatively respond to rises in the PIEPU index. Additionally, the negative impacts of the rises in the specifically COVID‐19 based‐constructed PIEPU (DCOVPIEPU) index on the S&P500 and Dow Jones indexes are higher than the negative impacts of the general PIEPU index. This can be interpreted to mean that the larger the magnitude and spread rate of a pandemic, the larger the negative impacts on stock returns. In the sample period of this study, COVID‐19 is the largest and most destructive pandemic compared to H1N1 and Ebola.

This study investigates the impacts of pandemic-induced economic policy uncertainties (PIEPU) on the S&P500, Nasdaq-100, and Dow Jones indexes (stock returns). To this aim, for the first time, newly created IDEMV (the Infectious Disease Equity Market Volatility index (henceforth, PIEPU index) is used. The Autoregressive Distributed Lag (ARDL) model and the Toda and Yamamoto (Journal of Econometrics, 1995, 66, pp. 225-250) causality test are applied for the 2009-2020 period. Empirical findings indicate that rises in the PIEPU index lead to falls of only the S&P500 and Dow Jones indexes. Corporations in the tech-heavy Nasdaq100 index do not negatively respond to rises in the PIEPU index. Additionally, the negative impacts of the rises in the specifically COVID-19 basedconstructed PIEPU (DCOVPIEPU) index on the S&P500 and Dow Jones indexes are higher than the negative impacts of the general PIEPU index. This can be interpreted to mean that the larger the magnitude and spread rate of a pandemic, the larger the negative impacts on stock returns. In the sample period of this study, COVID-19 is the largest and most destructive pandemic compared to H1N1 and Ebola.

COVID-19, Infectious Disease IDEMV Index, stock exchanges 1 | INTRODUCTION COVID-19 pandemic as a multidimensional phenomenon attracted many scholars to examine its impacts in different areas such as public health (Heymann & Shindo, 2020; Polychronis & Roupa, 2020) , education (Dennis, 2020 : Torda, 2020 , economy (Alhassan et al., 2020; Siddiquei & Khan, 2020; Singh & Neog, 2020) , environment (Alola & Bekun, 2020; Balsalobre-Lorente et al., 2020) , trade (Escaith & Khorana, 2021; Vidya & Prabheesh, 2020) . On the other hand, the COVID-19 pandemic (known as the "great lockdown" in the IMF report (2020) 1 to echo the Great Depression) has heightened unprecedented uncertainties in the economy and led to massive losses for businesses. According to Baker, Bloom, Davis, and Terry (2020) , the magnitude of uncertainty caused by this pandemic is larger than the 2008 financial crisis and close to that of the Great Depression. Hence, rising economic uncertainties may have negative impacts on stock exchanges. Therefore, this study aims to investigate the potential impacts of the pandemic-induced economic policy uncertainties (PIEPU) on US stock exchanges through the S&P500, Nasdaq-100, and Dow Jones indexes. To this aim, the newly created Infectious Disease Equity Market Volatility (IDEMV 2 ) index by Baker, Bloom, Davis, Kost, et al. (2020) is used. To the best of our knowledge, this is the first attempt using this index in an empirical study related to stock exchanges.

The IDEMV was constructed as a news-based index. In the construction of this index, first, articles across approximately 3000 US newspapers are scanned and terms that mention at least one term in each of ID, E, M, and V are counted in the following four sets: ID: (epidemic, pandemic, virus, flu, disease, Coronavirus, Mers, Sars, Ebola, H5N1, H1N1) E: (economic, economy, financial) M: ("stock market", equity, equities, "Standard and Poors") V: (volatility, volatile, uncertain, uncertainty, risk, risky) Second, the raw IDEMV counts are scaled by the count of all articles on the same day. Lastly, the resulting series are multiplicatively rescaled to scale equity market volatility (EMV) series in the EMV tracker, 3 which is formulized as follows:

where # is the count of newspaper articles in the EMV sets. The EMV t is the value of the overall EMV tracker, which was constructed based on different weighted article categories 4 that mention one or more terms in each category. The sample formula presented above was constructed to show the importance of "monetary policy," which is one of the items in these categories affecting EMV. To show the importance of fiscal policy in the EMV Tracker (2020), we would substitute "monetary policy" with "fiscal policy" in this formula.

As a result of all these steps and instructions so far, we can conclude that the IDEMV, constructed on the volatilities through the EMV tracker, including ID terms, will give us a kind of pandemic-induced economic policy uncertainty index. Therefore, the IDEMV index will henceforth be replaced with the pandemic-induced economic uncertainty (PIEPU) index in this study. Especially in regards to the COVID-19 pandemic, using this index may provide a wide range of application fields to researchers who want to test the impacts of PIEPU in their empirical models. This daily index is available as of 1985. It is believed that the findings of this study will also support some empirical studies (Arouri et al., 2016; Christou et al., 2017; Ongan & Gocer, 2017; Paule-Vianez et al., 2020; Peng et al., 2018) that test the impacts of changes in the economic policy uncertainty (EPU 5 ) index on US stock returns. The EPU index, created by Baker et al. (2016) , was also constructed based on newspaper articles (with similar technical instructions to those for the PIEPU index). However, the EPU index (2020) does not specifically consider the uncertainties caused by pandemics, like the PIEPU index, which was used in this study.

We obtained the data of the S&P500, Nasdaq-100, and Dow Jones indexes from the Federal Reserve Bank of St. Louis (FED, 2020) . The data of the IDEMV (henceforth, PIEPU) index, created by Baker, Bloom, Davis, Kost, et al. (2020) , were obtained from https://www.policyuncertainty.

com/infectious_EMV.html. We used daily series and the sample period of the study is January 05, 2009 to May 18, 2020, with 2862 observations.

To investigate the potential impacts of pandemic-induced economic uncertainties on US stock returns, we apply both the Autoregressive Distributed Lag (ARDL) model and Toda and Yamamoto (1995) causality test. Appropriate analysis methods and casualty test were selected according to the results of the unit root tests.

In this model, we use the following linear ARDL model by Pesaran et al. (2001) in error correction form:

where SR and PIEPU are stock returns (indexes) and PIEPU index, respectively.

p and q are optimum lags determined by using Akaike information criteria (AIC). ε t is the innovation at time t: We expect the sign of α 4 to be negative, since rises in the PIEPU index will lead to falls in SR in the long run. This model is separately applied to the S&P500, Nasdaq, and Dow Jones indexes. Additionally, to consider-include the specific impacts of the COVID-19 pandemic on stock indexes, Model 1 is transformed to the following Model 2 with an additional new variable (index):

where DCOVPIEPU is specifically COVID-19 based-constructed PIEPU variable (index). We created this index based on the equation:

is given the value of 0 and 1 before and after January 21, 2020, respectively. This is the date of the first confirmed case of this pandemic in the United States. We expect the sign of α 6 to be negative and its size to be higher than α 4 in Equation (2), because, in our sample period, the COVID-19 is the largest pandemic when compared to the H1N1 and Ebola.

To support the results of the ARDL model, we apply the Toda and Yamamoto (1995) PIEPU and DCOVPIEPT indexes to the S&P500, Nasdaq-100 and, Dow Jones indexes in the following models:

where p is the optimal lag and d max is the degree of maximum integration of variables. The null hypothesis of this test is "there is no causality." If β 2i ¼ 0, then we will decide that there is no causality from the PIEPU index to SR; if α 2i ¼ 0, then we will decide that there is no causality from DCOVPIEPU to SR.

Before running the models, we must first make sure the series are stationary. To this aim, we apply the ADF (Dickey & Fuller, 1981) , PP (Phillips & Perron, 1988) , and KPSS (Kwiatkowski et al., 1992) unit root tests. The results of these three tests are reported in Table 1 .

Test results in Table 1 indicate that the series is stationary at different levels. Hence, to test long-run relations, we apply the bounds testing cointegration approach by Pesaran et al. (2001) . The results of bound testing are reported in Table 2 . The null hypothesis of this test is that there is "No cointegration."

Test results in Table 2 indicate that the series have significant long-run (cointegration) relations, since the F-statistics of the models are higher than the upper bounds at 1% and 5% levels. Hence, to estimate the coefficients, we apply the ARDL model. Test results of this approach for Model 1 (without specific COVID-19 effect) and Model 2 (with specific COVID-19 effect) are reported in Tables 3 and 4, respectively.

Test results of normalized long-run estimates of Model 1 (without specific COVID-19 effect) in Table 3 

(1) S&P 500

(2) Dow Jones (3) Nasdaq-100

Model 1 9.14** 11.05*** 10.26*** 

(1)

(2)

(1) 

(1)

(2)

(1)

( (2) (3) Note: p-values are in the parenthesis. *, **, and *** denote significances at 10%, 5%, and 1% levels. and Dow Jones indexes are higher (À0.04: À0.05 in Table 4 ) than the negative impacts of the general EIEPU index (À0.018: À0.026 in Table 3 ). This can be interpreted to mean that the larger the magnitude and spread rate of a pandemic, the larger the negative impacts on stock returns. In the sample period of this study, the COVID-19 is the largest and most destructive pandemic compared to the H1N1 and Ebola.

Additionally, test results of the Toda and Yamamoto (1995) causality test are reported in Table 5 . It should be noted that this causality test approach was selected because the series of the study are stationary at different levels.

Test results in Table 5 indicate that there are causalities from the DCOVPIEPU index only to the S&P500 and Dow Jones indexes.

These findings completely affirm-support the results of the ARDL 

We confirm that this work is original and has not been published elsewhere nor is it currently under consideration for publication elsewhere.

T A B L E 5 Toda and Yamamoto (1995) 

The data that support the findings of this study are available from the corresponding author upon reasonable request. 

The impact of Covid-19 spread on stock markets: The case of the GCC countries

Does life expectancy, death rate and public health expenditure matter in sustaining economic growth under COVID-19: Empirical evidence from Nigeria?

Pandemic outbreaks (COVID 19) and sectoral carbon emissions in the United States: A spillover effect evidence from Diebold and Yilmaz index

Economic policy uncertainty and stock markets: Long-run evidence from the US

Measuring economic policy uncertainty Quart

Policy news and stock market volatility (Working Paper No. 2019-53)

The Unprecedented Stock Market Impact of COVID-19

Covid-induced economic uncertainty

Consequences of COVID-19 on the social isolation of the Chinese economy: Accounting for the role of reduction in carbon emissions

The outbreak of COVID-19 pandemic and its impact on stock market volatility: Evidence from a worst-affected economy

Relative stock market performance during the coronavirus pandemic: Virus vs. policy effects in 80 countries

Economic policy uncertainty and stock market returns in Pacific rim countries: Evidence based on a Bayesian panel VAR model

Consider higher education opportunities after COVID-19. The Successful Register

Likelihood ratio statistics for autoregressive time series with a unit root

The impact of the COVID-19 pandemic on merchandise trade in commonwealth countries

COVID-19: What is the next for public health

The Great Lockdown: Worst Economic Downturn Since the Great Depression

The impact of COVID-19 pandemic on stock markets: An empirical analysis of world major stock indices

Testing the null hypothesis of stationarity against the alternative of a unit root: How sure are we that economic time series have a unit root

The impact of the COVID-19 lockdown on stock market performance: Evidence from Vietnam

Testing the causalities between economic policy uncertainty and the US stock indices: Applications of linear and nonlinear approaches

Effect of economic and monetary policy uncertainty on stock markets. Evidence on return, volatility and liquidity

Asymmetric dependence between economic policy uncertainty and stock market returns in G7 and BRIC: A quantile regression approach

Bounds testing approaches to the analysis of level relationships

Testing for a unit root in time series regression

Health Workers' knowledge and perception regarding the risk of spread of COVID-19 during the pandemic: A systematic review

COVID-19's disaster are perlous than global financial crisis: A rumor or fact? Financial Research Letters

Economic implications of coronavirus

Contagion effect of COVID-19 outbreak: Another recipe for disaster on Indian economy

Statistical inference in vector autoregressions with possibly integrated processes

How COVID-19 has pushed us into a medical education revolution

Implications of COVID-19 pandemic on the global trade networks. Emerging Markets Finance and Trade

Ongan earned his Ph.D. degree in 1995. He became Associate Professor in Macroeconomics in 2005 and Full Professor in 2010. He has been teaching Intermediate Macroeconomics and Comparative Economics at the University of South Florida, Tampa. Before this university, he taught at