Performance of Shanghai Composite Index and Sector Indices in The Beginning of Novel Coronavirus Pandemic

This paper aims to quantify the effect of the deadly novel coronavirus (COVID-19) pandemic outbreak on Chinese stock market performance. Shanghai Stock Exchange Composite Index and its component sector indices are examined in this study. The pandemic is represented by a lockdown dummy, new COVID-19 cases and a dummy for 3 February 2020. First, descriptive analysis is performed on these indices to compare their performances before and during the lockdown period. Next, regression analysis with Exponential Generalized Autoregressive Conditional Heteroscedasticity specification is estimated to quantify the pandemic effect on the Chinese stock market. This paper finds that health care, information technology and telecommunication services sectors were relatively more pandemic-resistant, while other sectors were more severely hurt by the pandemic outbreak. The extent to which each sector was affected by pandemic and sentiments in other financial and commodity markets were reported in details in this paper. The findings of this paper are resourceful for investors to avoid huge loss amid pandemic outburst and the China Securities Regulatory Commission in handling future pandemic occurrence to cool down excessive market sentiments.


Introduction
The novel coronavirus  pandemic was first detected in Wuhan, Central China in December 2019 (Wuhan Municipal Health Commission, 2019) (Note 1). This human transmitted disease spread at a rapid rate to other parts of China, and then to other parts of the world (Note 2) (Note 3). As of 28 April 2020, there were total of 2,883,603 confirmed cases, while the death toll had reached 198,842 cases (6.9%) (World Health Organization, 2020). To halt the outbreak, China swiftly imposed the unprecedented lockdown in Wuhan city, the epicenter of COVID-19, and nearby cities (so-called Wuhan lockdown), starting from 23 January 2020. By mid-February, some 200 other cities around the nation had implemented similar social distancing measures, control of movement, closed management and ban on travels (Note 4). Disregarding the sudden spike on 12 February 2020 (due to diagnostic changes), it took 2 weeks after the lockdown imposition for the daily new COVID-19 confirmed cases and deaths to exhibit declining trends, see Figure 1.
While China was able to curtail the outbreak of COVID-19 by containment, its production and economy were badly affected. Businesses and factories in China were essentially scaled down if not totally shutdown in the lockdown period. Subsequently, the country's industrial profit plunged by 38.3% in the first two months of the year 2020. Meanwhile, some 5 million employees lost their jobs in the same period, with urban unemployment surged to 6.2% in February 2020, up from 5.2% in December 2019. Moreover, barometers of economic growth such as retail sales of consumer goods and investment spending (on infrastructure, property, machinery and equipment) fell 29.5 and 24.5% respectively (Bermingham. and Wang, 2020;Chen, 2020; National Bureau of Statistics of China, 2020; Roper, 2020). As China's economy sank to all-time low in the first two months of the year, International Monetary Fund remarked on 20 March 2020 that "China's slowdown in the first quarter of 2020 will be significant and will leave a deep mark for the year" (Berger et al., 2020). In fact, China economy shrank 6.8% in the first quarter compared with the year before (Vaswani, 2020).  From the perspective of stock exchange, another barometer of economic growth (Note 5), Chinese stock markets had tumbled following the news of COVID-19 outbreak (Avalos and Zakraišek, 2020;Oxford Analytica, 2020). In this respect, the Shanghai and Shenzhen composite indices recorded the biggest single-day loss (2.75% and 3.25% respectively) in almost 9 months on 23 January 2020 (Chinanews.com, 2020a) (Note 6). On the next trading day (3 February, 2020), these two indices further suffered a gap down of 8% in the opening session. They eventually closed with a loss of 7.72% for Shanghai and 8.45% for Shenzhen composite indices, with the single day biggest loss in 12 months (Chinanews.com, 2020b). The current empirical study attempts to provide a comprehensive measure on the effect of the deadly COVID-19 outbreak on Chinese stock market performance, with reference to Shanghai Composite index and its component sectors indices.
The rest of this paper is structured as followed. The Brief Literature Overview section points out the urgency of conducting the current research. The Empirical data and descriptive analysis section describes the data sources and variables involved in this study. Besides, descriptive analysis is performed for data exploration. The Regression model section explains the ordinary least squares regression model with the EGARCH (1,1) specification employed in this study. The Regression results section presents the estimated results and interpretation of results. The last section on Conclusion and policy recommendation summarizes the key findings, suggests policy implications and concludes this paper.

Brief Literature Overview
In the recent decade, there has been a considerable surge in Chinese stock market research in conjunction to the intensifying China's influential roles on various aspects of the world including the stock exchanges. Majority of these studies focused on the efficiency (see among others, Chong et al., 2012;Han et al., 2019, Liu et al., 2020 andreferences therein) or relationship analysis between Chinese stock market with international stock markets (Zhang and Li, 2014;He et al., 2015;Hung, 2019), commodity markets or/and macroeconomic aspects (Ahmed and Huo, 2020;Zhuang et al., 2015, Jebran, 2019, Wei et al., 2019Wen et al. 2019;Yousaf and Hassan, 2019;Zhao et al., 2020 and references therein).
Remarkably, effect of financial crises on Chinese stock market was also commonly examined in the past (Jebran, 2019;Han et al., 2019;Hung, 2019;Lin and Tsai, 2019;Yousaf andHassan 2019, andWei et al., 2019), while studies on natural disasters were rare. Two contributions in this aspect are worth-mentioning. Specifically, Li et al. (2015) examine 60 natural disaster announcements over the period 2003-2013 and find that they produced significant negative stock returns in China. Earlier on, Chong et al. (2010) report significant positive abnormal returns for pharmaceutical stocks in China, while tourism stocks were significantly affected by the severe acute respiratory syndrome (SARS) caused by coronavirus (Note 7). These two studies adopted the event study approach. The current study attempts to estimate the nature, magnitude and significance of effect of COVID-19 pandemic on Chinese stock market with regression analysis that controls for the plausible financial and commodity markets spillover effects (Note 8) (Note 8).

Empirical data and descriptive analysis
The data set spanning from 2 September 2019 to 27 March 2020 are considered in this study. To estimate the effect of COVID-19 pandemic on the Chinese stock market, Shanghai Stock Exchange Composite Index (SSE) and its component sector indices are employed. They are from the sectors of energy (EN), materials (MT), industrials (IND), consumer discretionary (CD), consumer staples (CS), health care (HC), financials (FN), information technology (INT), telecommunication services (TS), and utilities (UL) (Note 9). COVID-19 pandemic is represented by a lockdown dummy (LD), the daily new confirmed cases in China (CNCC) and the first trading day after the Chinese New Year holidays (D1). LD assumes a value of 1 for observations ranging from 23 January 2020 to 27 March 2020, and zero otherwise. Other financial market variables specifically the New York Stock Exchange Composite Index (NYSE), the remninbi per US dollar exchange rate (ER), and commodity markets represented by spot gold price (GP) and crude oil WTI spot US dollar price (OP) are included for control purpose. Daily financial and commodity variables are obtained from Investing.com. Daily new confirmed cases in China is obtained from Worldometer. These time series are plotted in Figure  2.  ISSN 1946-052X 2021 ajfa.macrothink.org/ 5 respond negatively. In the meantime, the China exchange rate and the spot gold price were continuing their value appreciation trend, while the crude oil was extending its declining trend. After the Wuhan lockdown imposition, the Chinese and US stock markets did not only behave differently in the process but they eventually ended up with significant different performances. SSE and NYSE stood at 14,102.04 and 2,976.53 respectively on the Wuhan lockdown day. As of 27 March 2020, they remained 8,777.38 and 2,660.17. This is equivalent to a considerable loss of 60.66% for NYSE in the US and 11.89% for SSE in China within 41 trading days. Figure 3 shows that there was an increasing trend starting from November 2019 in majority of the sectors in Shanghai Stock Exchange, and so the composite index, probably due seasonal effects (Note 10). This trend was extended into the middle of January 2020 or later, although WHO was officially informed of the of the new virus on 31 December 2019. For the composite index, it reached its peak at 3,115.57 (H1) on 13 January 2020, and from there on it slipped to 2,746.61(L1) on 3 February 2020.

Sector indices performance
Sector-wise, both consumer discretionary and staples sector indices reached their peaks (H1) at 2,548.69 and 11,128.32 respectively on 13 January 2020. Besides, information technology as well as telecommunication services and health care sector indices further climbed to H1 at 4, 611.15 (22 January); 3,442.73 (22 January) and 6,967.30 (21 January) correspondingly, before the trend reversed. In this conjunction, it is evident from Figure 2 that, the reversal trend ended with a big dip for all sectors on 3 February 2020 (Note11). In fact, on this first trading day after the extended Chinese New Year celebrations, the composite index plunged by 7.72% amid the fear of the COVID-19 pandemic rapid outbreak when millions of workforces were commuting to work. Meanwhile, all sectors were extensively down without exception, see Table 1 which highlights of percentage returns of SSE stock indices in the first three months of the year 2020.

Figure 2. Time series plots of financial and commodities variables
After the deep market correction on 3 February 2020, sending the indices to their new lows (at L1) in 12 months, all these indices exhibited technical rebound with different strengths ( Figure  2). In general, this round of rebound ended on 5 March (H2), and then they started to decline again until 23 March (L2), with the exception of information technology and telecommunication sectors. As we can observe in Column 4 of Table 1, the energy and financials sectors set for the comparatively weakest rebound (Note12) before their indices continued with another fall to find their lower lows (L2 < L1) at 1,019.20 and 4,614.08 respectively on 23 March 2020 (Note 13). ISSN 1946-052X 2021 ajfa.macrothink.org/ As for information technology, telecommunication sectors and consumer staples, they exhibited the strongest strength in rebound to succeed their higher highs (H2 > H1) (see Figure  2) of 5,578, 3,803 and 11,229.08 on 24 February, 25 February and 5 March respectively (Note 14). Remarkably, they managed to pose higher lows (L2 > L1), in the following round of price falling action, indicating a better future prospect than other sectors. The index of health care (H2 at 6 February) sector also managed to find their higher highs (compared H1) but like the majority of other sectors, it ended up with lower lows (L2 < L1). Indices which rebounded for only lower highs (H2 < H1) and ended up eventually with lower lows (L2 < L1) were materials, industrials, energy, consumer discretionary and financials.
Table 1 also reveals that the indices of consumer discretionary (-19.45%), materials (-16.44%) and information technology (-15.32%) performed the worst in terms of periodic returns from 5 March to 23 March 2020. On the other hand, telecommunication services (-5.69%) and utilities (-7.86%) sectors experienced the smallest loss over the same period.   . Moreover, all their returns were turning from positive pre-lockdown to negative during lockdown, with the exception of utilities sector index which was already in the negative pre-lockdown. This means the price performance of these sectors was very much affected by the lockdown as a result of COVID-19 outbreak. In sharp contrast, health care (HC), information technology (INT), telecommunication services (TS) sectors were able to record higher maximum, average and minimum prices during the lockdown period. In fact, health care is the only sector that was able to maintain its positive return amid the lockdown, while information technology (-4.06%) and telecommunication services (-0.98%) sectors suffered relatively milder losses in indices returns in the lockdown period.
Having analyze the indices price behavior and return performance before and during the lockdown to prevent COVID-19 outbreak, another key finding can be derived from Table 2. Specifically, the standard deviation of all indices were larger during lockdown in relation to pre-lock period, with the exception of industrial sector. This is probably because investors' reaction was more sensitive to COVID-19 news such that the overall Chinese stock market was relative more volatile during lockdown. Table 2. Descriptive statistics of SSE indices before and during lockdown

Regression model
This study adopts regression model to quantify the effect on COVID-19 pandemic on the Chinese stock market. The following regression model is estimated: for = 0, 1, … ,7 are the parameters to be estimated. The model residuals series is denoted by and is the heteroscedastic variance of . In the meantime, for = 0, 1, … ,3 are the parameters to be estimated in the variance equation.
Notably, as this study estimates the effect of the COVID-19 outbreak, three proxies are included. LD is the dummy variable for lockdown as a result of COVID-19 outbreak. LD assumes a value of one for dates (inclusively) ranging from 23 January 2020 to 27 March 2020, and zero otherwise. CNCC stands for daily new confirmed COVID-19 cases in China. A dummy variable (D1) is included to estimate the effect of the incident of drastic plunged in the Chinese stock market on the first trading day after the prolonged Chinese New Year holidays, amid the fears of rapid spread COVID-19 (Note 16). D1 takes a value of one for the day 3 February 2020 and zero for all other dates.
It is expected that the Chinese stock market is adversely affected by the lockdown, and so the coefficient of LD ( ) should be negative, for lower index price during the lockdown period. On the other hand, CNCC signals the condition of the coronavirus outbreak with higher (lower) value means more (less) confirmed COVID-19 cases in China. A rational investor would take less cases as favorable news of improving condition and more cases as deteriorating condition. Thus, < 0 is expected. On the other hand, should be negative to confirm with the observations of market plunged. Importantly, the magnitude of , and will reveal the extent of the COVID-19 effect on the Chinese stock market.
Equation (1) is estimated with Exponential General Autoregressive Conditional Heteroscedasticity EGARCH (1, 1) specification as shown in Equation (2) to model the residuals. In this respect, EGARCH (1, 1) is commonly used to model and predict the financial time series that exhibit time-varying volatility (Note 17) (Note 18). Remarkably, and are of particular interest.
is the asymmetry parameter. If ≠ 1, then positive and negative shocks/news, , will have asymmetric effect on the price volatility on the next day. If = 0, there is no asymmetrical behavior detected. is the persistence parameter. The larger (smaller) the , the longer (shorter) it takes for the shocks/news to decay. In this respect, following Olbrys and Majewska (2017), this study measures the persistence by calculating the half-life, which is the time taken for to be reduced to one-half of their original size. Specifically, Half-life = ln(0.5)/ln(| |). (3) In addition, for the conditional volatility process as represented in Equation (2) to be stationary, the condition for | | < 1 must be met. Otherwise the process is explosive and the index is unpredictable.

Regression results
The estimated regression models over the sample period spanning from 2 September 2019 to 27 March 2020 are reported in Table 3 (Note 19), while the half-life and diagnostic results are given in Table 4. First and foremost, it is evident from Table 3 that the lockdown imposition to circumvent COVID-19 outbreak had exerted a significant impact on the Shanghai stock  ISSN 1946-052X 2021 ajfa.macrothink.org/ exchange, as the estimated coefficient is significant at 5% statistical level for the over market composite index and all its component sectors. Second, different sectors received different COVID-19 effect in terms of lockdown, depending on their nature of business. Specifically, negative effect is observed in the Shanghai Stock Exchange overall market as well as the energy, materials, industrials, consumer discretionary, consumer staples, financials and utilities sectors. In contrast, positive effect is revealed in the health care, information technology and telecommunication services sectors in specific. These sectors had benefited from their important contributions to the fight against COVID-19. Third, estimated coefficient for the lockdown to was -67.99 points for Shanghai Stock Exchange Composite Index as dependent variable. It reveals, overall, that the composite index had slumped by an average of 67.99 points significantly (as noted in the first point) in response to the lockdown, compared to the its average value before lockdown, holding other variables constant. Similarly, on average, the indices of the energy, materials, industrials, consumer discretionary, consumer staples, financials and utilities sectors were down by 129.80, 52.88, 60.35, 64.02, 693.23, 272.62, 148.83 points respectively during the lockdown. Conversely, the indices for the health care, information technology and telecommunication services sectors were up by 237.46, 654.92 and 159.81 corresponding.
Fourth, the estimated coefficient for daily new confirmed COVID-19 cases in China (CNCC) is essentially negligible in size and it was insignificant at all for all sectors, except for financials sector ( = −0.01, < 0.05) . It reveals, in general, that the market was insignificantly affected by the new developments of COVID-19 cases.
Fifth, the estimated sign of coefficient is negative in all sectors. Moreover, the effect is significant for all sectors, excluding the energy and financials sectors. This indicates that the fears of COVID-19 outbreak did not only persisted but it had mounted over this long holidays with no share trading activities. Subsequently, the whole Chinese stock market reacted negatively when the market resumed trading on 3 February 2020.
Sixth, exchange rate had significant and negative effect on Shanghai Stock Exchange Composite Index (-301.66 points), materials (-612.02), industrials (-53.43), consumer discretionary (-345.26), consumer staples (-391.20), financials (-945.88), information technology (-1196.89) and telecommunication services (-1046.23). It suggests that lower (higher) indices are associated to appreciation (depreciation) of renminbi with respect to the US dollar (Note 20). In other words, depreciation of renminbi is favorable for the growth of these sectors in China. Conversely, the results indicate that appreciation of renminbi is favorable for the energy and utilities sectors.
Seventh, the US stock market as the world's leading stock market on the Chinese stock market is minute if not totally negligible as the magnitude is less than one point in all sectors. Eighth, the Chinese stock market is significant and positively related to the spot gold market. The coefficient is significant and positive for all indices with no exception. Ninth, the Chinese stock market is significant and positively affected by the crude oil market, with the exception of consumer staples (negative effect) and health care (insignificant effect).
Tenth, the estimated is insignificantly different from zero in all sectors, implying the absence of asymmetrical reactions on favorable or unfavorable news in the Chinese stock market. Last but not least, the estimated are all less than one in magnitude, implying the conditional volatility process is stationary and the index is predictable. Moreover, the half-life values calculated from estimated suggest fast market digestion of shocks as it took the Chinese stock market less than 3 trading days for shocks to be reduced to one-half its original size.

Conclusion and policy implications
The current empirical study attempts to provide a comprehensive measure on the effect of the deadly COVID-19 outbreak on Chinese stock market performance, with reference to Shanghai Composite index and its component sectors indices. The data set spanning from 2 September 2019 to 27 March 2020 are considered in this study. Shanghai Stock Exchange Composite Index and its component sector indices (energy, materials, industrials, consumer discretionary, consumer staples, health care, financials, information technology, telecommunication services, and utilities are examined. COVID-19 is represented by a lockdown dummy (LD), the daily new confirmed cases in China (CNCC) and a proxy for the first trading day after the Chinese New Year holidays. Financial market variables specifically the New York Stock Exchange Composite Index (NYSE) and exchange rate (ER), and commodity markets represented by spot gold price (GP) and crude oil WTI spot US dollar price (OP) are included for control purpose.
Few key findings could be drawn from the numerical analysis over the Chinese stock markets data. First, when the market resume trading after a prolonged Chinese New Year celebrations, the composite index unreservedly plunged by 7.72% amid the extended fear of the COVID-19 rapid outbreak. With no exception, all sectors were extensively down and the daily performances of these sectors on the 3 February 2020, from the most wounded to the least were: telecommunication services (-9.99%), information technology (-9.72%), industrials (-9.03%), consumer discretionary (-8.91%), materials (-8.64%), energy (-8.44%), utilities (-7.68%), financials (-7.28%), consumer staples (-6.85%) and health care (-3.22%). In fact, on 23 January 2020, when Wuhan city was locked down, there was already one round of market considerable drop. The daily performances of these sectors in the same manner were: telecommunication services (-4.79%), materials (-3.57%), consumer staples (-3.53%), consumer discretionary (-3.28%), information technology (-3.07%), industrials (-2.96%), energy (-2.78%), financials (-2.73%), utilities (-7.68%) and health care (-2.08%). Second, after the deep market correction on 3 February 2020, sending the indices to their new lows in 12 months, all these indices exhibited technical rebound with different strengths. In general, the energy and financials sectors set for the comparatively weakest rebound, while information technology, telecommunication sectors and consumer staples exhibited the strongest strength in rebound (Table 1). These findings suggest investors who hold shares should close their trades to avoid huge loss, as soon as quality information on negative shocks are received. Besides, short-term investors could buy low whenever there is a substantially dip in the market and close the trade within 5 to 10 trading days upon technical rebound for sizable profits. On the other perspective, the China Securities Regulatory Commission should immediately prohibit short-selling for considerate period in handling future events of pandemic outbreak and other crisis. This could avoid market over-reaction and to reduce market volatility.
Third, sub-sample analysis for the composite and sector indices before and during lockdown, with equal window length reveals that all the indices returns of materials, industrials, consumer discretionary and financials sectors dropped from positive to negative after the lockdown as a result of COVID-19 outbreak. Contradictory, health care is the only sector that was able to maintain its positive return amid the lockdown, while information technology and telecommunication services sectors suffered mild losses in indices returns in the lockdown period. Fourth, the standard deviation of all indices were larger during lockdown in relation to pre-lockdown period, with the exception of industrial sector, implying the Chinese stock market as a whole turned relatively more volatile during lockdown (Table 2). In this respect, risk adverse investors are advised to stay away from the market amid pandemic outbreak or crisis in the future.
First, different sectors received significant but different COVID-19 effect in terms of lockdown, depending on their nature of business. Specifically, overall, Shanghai Stock Exchange Composite Index slumped by an average of 67.99 points significantly in response to the lockdown. In the meantime, the indices of the energy, materials, industrials, consumer discretionary, consumer staples, financials and utilities sectors were down by 129.80, 52.88, 60.35, 64.02, 693.23, 272.62, 148.83 points respectively during the lockdown. Conversely, the indices for the health care, information technology and telecommunication services sectors were up by 237.46, 654.92 and 159.81 corresponding. This reveals that health care, information technology telecommunication services sectors are potential avenues for profit-seeking in the time of pandemic outbreak.
Second, the fears of COVID-19 outbreak had mounted over the prolonged Chinese New Year holidays with no share trading activities. Subsequently, the all indices in the Chinese stock market was negative and significant when the market resumed trading on 3 February 2020, confirming the significance of the findings from descriptive analysis. On the other hand, the estimated results suggest that the Chinese stock market was insignificantly affected by the new developments of COVID-19 cases, probably the effect was overshadowed by the other two COVID-19 proxies.
Third, it is found that depreciation of renminbi is favorable for the growth of materials, industrials, consumer discretionary, consumer staples, financials, information technology and telecommunication services sectors in China. Conversely, appreciation of renminbi is favorable for the energy and utilities sector, probably due to the oil importing nature of China. As such, the Chinese government should maintain a fair renminbi in order to promote its economic growth by expanding its exports on consumer and industrial products as well as information technology and telecommunication services.
Fourth, the Chinese stock market is significant and positively related to the spot gold market and crude oil market in general. Thus, these two commodity markets are not suitable avenue for investment diversification. However, investors could detect the sentiments in these commodity markets for plausible trading signals for the Chinese stock market.
Fifth, this study finds no asymmetrical reactions on favorable and unfavorable shocks in the Chinese stock market. Nevertheless, these shocks were digested quickly as it took the Chinese stock market less than 3 trading days for shocks to be reduced to one-half its original size. This suggests that the COVID-19 effect on the Chinese stock market in general and the various sectors in the country specifically is not persistent and hence these sectors will recover very soon upon removal of lockdown and when the pandemic is over.