The Impact of Covid-19 Spread on Stock Markets: The Case of the GCC Countries

This paper attempts to investigate the effects of 2020 COVID-19 world-wide spread on stock markets of GCC countries. Coronavirus spread has been measured by cumulative cases, new cases, cumulative deaths and new deaths. Coronavirus spread has been measured by numbers per million of population, while stock market return is measured by Δ in stock market index. <br><br>Papers conducted in this topic tend to analyze Coronavirus spread in the highly infected countries and focus on the developed stock markets. Countries with low level of infection that have emerging financial markets seem to be less attractive to scholars concerning with Coronavirus spread on stock markets. This is why we try to investigate the GCC stock markets reaction to COVID-19 spread. <br><br>Findings show that there are significant differences among stock market indices during the research period. Besides, stock market returns seem to be sensitive to Coronavirus new deaths. Moreover, this has been confirmed for March without any evidence about these effects during April and May 2020. <br>

Usually, global financial crisis plays out in countries across the globe and consequently manifests in four overlapping phases. Although each phase has a policy focus, each phase of the crisis affects the others, and, until the crisis has passed, no phase seems to have a clear end point. Nanto (2009) summarized the four phases of the global financial crisis as follows: contain the contagion and strengthen financial sectors; coping with macroeconomic effects; regulatory and financial market reform and dealing with political, social, and security effects. Orlweski (2008) identifies five distinctive stages of the current global financial crisis as follows: the outbreak of the subprime mortgage crisis; the proliferation of credit risk, with the broadening of losses of financial institutions; the eruption of liquidity crisis; the commodity price bubble and the ultimate freeze of credit markets.
When it comes to COVID-19, it's important to consider not only the characteristics of financial crisis, but also the economic shocks, where poverty kills poor people, but the outbreak of COVID-19 has another story. McKibbin & Fernando (2020) shows that if diseases are generated in poor countries due to overcrowding, poor public health and interaction with wild animals, they can kill people of any socioeconomic group in any society. Unfortunately, politicians continue to ignore the needs of investment in public health and development and the scientific evidence on the role of public health in improving the quality of life and as a driver of economic growth.
Coronavirus spread has been increased during the research period in all of the GCC countries, where Saudi Arabia has recorded the highest indicators and Bahrain has registered the lowest ones (https://www.worldometers.info/coronavirus).  (1) shows that stock market returns tend to be decreasing during March 2020 and seem to be flattened during April and May. This paper tries to explain the behavior of stock market returns due to Coronavirus spread indicators.
This paper addresses a main question about the stock market reaction to Coronavirus spread. This has been applied on the Gulf Cooperation Council (GCC) countries, on daily basis over the period from March 1, 2020 till May 31, 2020. So, this paper tries to address the following questions: 1-Are there significant differences among stock market indicators during months of Coronavirus spread compared with the earlier months.

2-Does Coronavirus spared affect stock market return?
After this introduction, section 2 illustrates the related literature. Section 3 explains how to develop hypotheses and measure variables. Section 4 presents descriptive and diagnostic statistics. Section 5 is for testing hypotheses and section 6 is for robustness checks. Section 7 summarizes the paper and provides remarks about conclusions.

Literature Review
This section tries to present some of previous work, which has been conducted in the field of stock market reaction to the positive or negative informational contents, especially to the announcement about Coronavirus spread. Besides, it covers some recent papers about Coronavirus economic effects.
Stock markets seem to be sensitive to bad news and this sensitivity may differ according to countries and industries, where Alber (2013a) supports the effects of "industry effect" on stock market reaction to global financial crisis in Egyptian, Kuwaiti, American and British stock markets during the period from 2007 to 2011. Besides, Saleh (2017) shows that the factors of fears and hesitation to startup an investment is one of the most important factor that brings entrepreneurs to fail in their investments, comparing with the institutional or individual investors in the stock market, where the brokers absorb the psychological fears and chocks.
On the other hand, good news may affect stock market returns, where Alber (2013b) addresses the effects of quality announcement on performance Egyptian listed companies. This has been conducted using a sample of 11 events, covering announcements of international and national quality accreditation during the period from 2006 to 2012. Using event study methodology, results indicate that, hypotheses regarding the significance of differences between ARs and CARs could be accepted. Another study by Alber (2020a) attempts to investigate the effects of Coronavirus spread on stock markets of the worst 6 countries over the period from March 1, 2020 till April 10, 2020. Results indicate that stock market return seems to be sensitive to Coronavirus cumulative cases. Besides, robustness check confirms these effects for China, France, Germany and Spain. However, these effects haven't been confirmed for Italy and United States. Using the same indicators of Coronavirus spread in Belgium, France, Germany, Italy, Netherlands Spain and UK, Alber (2020b) has NOT supported the anticipated effects during the period from Febreuary15, 2020 till May 24, 2020 on daily basis. After splitting the research Comparing with literature, it's important to pinpoint that it considers not only both of infection and death indicators, but also, both of cumulative and new ones. Moreover, Coronavirus spread has been measured relatively, where all measures are adjusted per million of country population. Besides, this study focuses on GCC countries, while many studies concern with the US, China and Europe.

Measuring Variables and Developing Hypotheses
Coronavirus spread has been measured by "Cumulative Coronavirus Cases" (CCC), "Cumulative Coronavirus This paper aims at testing the following two hypotheses: 1-There's no significant differences among stock market indicators during months of Coronavirus spread compared with the earlier months.
2-There's no significant effect of "Coronavirus spread" on "stock market return".
Regarding the first hypothesis, Wilcoxon Signed Rank test has been conducted to investigate the significance of differences among stock market indicators during March, April and May 2020 compared with the January and February 2020.

SMR = α + β NCD + ε (9)
All of the Coronavirus indicators are positively correlated and this is why we test their effects separately to avoid the problem of multicollinearity.

Descriptive and Diagnostic Statistics
The first COVID-19 cases in Europe were reported in UAE, on February 15, 2020, and the first death was in Bahrain on March 16, 2020. Tables (2) illustrates the information of the research sample that includes 7 stock markets, over the period from March 1, 2020 till May 31, 2020 as follows: Tables (2) illustrates descriptive statistics of the research variables during the research period and table (3) indicates the correlation coefficients as follows:  193 193 193 193 193 Source: Outputs of data processing using EViews 10.  Source: Outputs of data processing using EViews 10.
Regarding normality, Jarque-Bera values indicate that all variables are normally distributed at p-value of 0.01 for most of the research variables. Regarding multicollinearity, the correlation coefficients among independent variables range from 0.62168 to 0.93389, which indicates that multicollinearity problem does exist and this is why we will use these variables separately.

Testing Hypotheses
The first hypothesis is about investigating the significance of differences between average changes in sector indices of stock markets, according to month of analysis. The null hypothesis H 0 could be shown as: The alternative hypothesis H a states that: Wilcoxon Signed Rank test has been conducted to investigate the significance of differences among stock market indicators during March, April and May 2020 compared with the January and February 2020.  When conducting Wilcoxon Signed Rank test, comparing the monthly return, results support the significance of difference between Feb., 2020 compared with Mar., 2020, with Z of -2.366 and sig. of 0.018, which means that the drop of monthly return from -5.502% to -18.995% is significant. So, for the first hypothesis, we can assume that there's a significant differences between monthly returns of GCC stock markets during Coronavirus spread compared with the earlier month.
The second hypothesis is about investigating the effect of each of "Coronavirus Cumulative Cases" (CCC), "New Coronavirus Cases" (NCC), "Cumulative Coronavirus Deaths" (CCD) and "New Coronavirus Deaths" (NCD) on "Stock Market Return" (SMR). Model (1) attempts to assess the effects of CCC model (2) concerns with investigating the effects of NCC. Besides, Model (3) investigates the effect of CCD and Model (4) is for NCD.    When splitting the whole research period into 3 sub-periods, results support the effect of New Coronavirus Deaths (NCD) effect on Stock Market Return (SMR) with explanation power of 30.7% in March. R 2 has been increased from 2.1% to 30.7% and this may be due to the homogeneity of the research period. Figure (3) shows that the decline of market indices has been continued till the end of March 2020, while curves seem to be more flattened in April and May. Stock markets of GCC countries seem to be different than those of China, France, Germany, and Spain, where stock market return seems to be sensitive to cumulative cases, not to new deaths.
So, for the first hypothesis, the null hypothesis is rejected and the alternative one could be accepted. However, for the second hypothesis, the null hypothesis could be accepted for CCC, NCC and CCD, while it's rejected for NCD. Figure (1) shows that the curves tend to be flattened starting from April 2020.
Another Robustness check has been conducted by reprocessing data after excluding only one stock market and provides the following results:   Source: outputs of data processing using EViews 10. http://ibr.ccsenet.org International Business Research Vol. 13, No. 11; Results provide that stock markets of the GCC countries seem to be sensitive to Coronavirus spread measured by New Coronavirus Deaths (NCD). These effects haven't been confirmed supporting the rejecting of the null hypothesis of the second hypothesis, providing higher explanation powers; where R 2 is 38.45% when excluding KSA, is 36.75% for ABD, is 31.44% for DUB, is 30.74% for KUW, is 31.75% for OMA, is 44.5% for BAH and is 10.48% for QAT.

Summary and Concluded Remarks
This paper attempts to investigate the effects of 2020 Covid-19 world-wide spread on stock markets of GCC countries on daily basis over the period from March 1, 2020 till May 31, 2020. Coronavirus spread has been measured by cumulative cases, new cases, cumulative deaths and new deaths. Coronavirus spread has been measured by numbers per million of population, while stock market return is measured by Δ in stock market index.
Papers conducted in this topic tend to analyze Coronavirus spread in the highly infected countries and focus on the developed stock markets. Countries with low level of infection that have emerging financial markets seem to be less attractive to scholars concerning with Coronavirus spread on stock markets. This is why we try to investigate the GCC stock markets reaction to Covid-19 spread .