A DEMAND-ORIENTED INDUSTRY-SPECIFIC VOLATILITY SPILLOVER NETWORK ANALYSIS OF CHINA’S STOCK MARKET AROUND THE OUTBREAK OF COVID-19

Fu Qiao¹⁺--- Yan Yan²

¹School of Economics and Management, University of Chinese Academy of Sciences, Beijing, China.
²Research Center on Fictitious Economy & Data Science, Chinese Academy of Sciences, Beijing, China.

Keywords:Volatility spillover network, BEKK-GARCH, Industry classification, China’s stock market, Demand, COVID-19.

JEL Classification: C58; G10.

ARTICLE HISTORY: Received:8 October 2020, Revised:26 October 2020, Accepted:6 November 2020, Published:23 November 2020

Contribution/Originality: This is one of very few studies that has investigated the volatility spillovers in the industry-specific networks of China’s stock market during the COVID-19 pandemic. The paper's primary contribution is finding the critical role that the service sectors play in the industry-specific network after the COVID-19 outbreak was contained.

1. INTRODUCTION

Frequently in the stock market, fluctuations in stock prices initially occur in companies belonging to one sector and gradually spread to other sectors. China’s stock market has become the second largest in the world. Up to January 2020, 3780 companies from a variety of sectors listed their shares on China’s stock market, which had a total value of more than 60.38 billion RMBs (approximately equal to 8.65 trillion US dollars). Thus, it is important for both investors and policymakers around the world to understand the complex linkage effects shown by fluctuations in the stock prices (or yields) of companies in different sectors in China’s stock market.

One of the typical measurements of the linkage characteristics between different variables is the spillover effect, which can be measured by using the generalized autoregressive conditional heteroskedastic (GARCH) family of models (such as the BEKK-GARCH and DCC-GARCH models) or the variance decomposition model under the vector autoregression (VAR) framework (Diebold & Yilmaz, 2012; Jiang, Jiang, Nie, & Mo, 2019; Singh, Kumar, & Pandey, 2010).

Studies on industry-specific volatility spillover networks have highlighted the measurement of the linkage level. These studies were initially motivated by finding the arbitrage opportunities between upstream and downstream industry sectors in the supply chain. In addition, studies, such as those by Yarovaya, Brzeszczyński, and Lau (2016) and Yin, Liu, and Jin (2020), further found that volatility spillovers also exist between industry sectors without a direct input–output relationship. However, the existing literature does not answer the question of how industry-specific volatility spillover networks reflect economic demand and its changes. We believe that there are two reasons for this. First, it is more difficult to provide a proper explanation for the findings in analyses on industry-specific volatility spillover networks than those done across countries or regions. Part of the detected spillovers in the networks might match economic theory, such as the spillovers between the energy and finance sectors, or those between the transportation and consumption sectors (Gonzalez-Navarro & Quintana-Domeque, 2016; Singh, Nishant, & Kumar, 2018). However, the rest of the spillovers might not be properly explained. Second, some scholars have criticized the arbitrariness when selecting industry classification standards. Mateus, Chinthalapati, & Mateus (2017) pointed out that the industry classification standard should be cautiously selected depending on the research targets. When necessary, self-built industry indices should also be used for pursuing more meaningful numerical results and theoretical implications.

Some early studies on this topic showed that exogenous shocks to the macroeconomy of a country do not lead to fluctuations in the prices of all securities in the country at the same time (Campbell, Lettau, Malkiel, & Xu, 2001; Ewing, Forbes, & Payne, 2003; Wang, 2010). Inspired by these studies, we further considered how the demand structure influences industry-specific spillover networks. The demands of a country mainly comprise consumption, investment and export. One of the critical factors of the profit and asset price of companies is whether or not their goods or services successfully meet a part of the demand (Acemoglu & Guerrieri, 2008). The influence of exogenous shocks on economic demand should, therefore, be reflected in the structural change in the industry-specific volatility spillover networks.

To highlight the economic demand structure, we chose the industry classification standard constructed by SWS Research Co., Ltd., which is the largest securities research institute in Mainland China . Using the SWS standard, we identified the GARCH-BEKK-based demand-oriented industry-specific volatility spillover networks of China’s stock market. Each node in the networks represents a level 2 industry securities index in the SWS industry list. We chose the minute-per-minute return data between January 2 and March 20, 2020 for 102 SWS industry indices as the sample. In this period, the exogenous shock of the outbreak of COVID-19 dramatically changed economic demand in China.

Recent studies have reported the influence of COVID-19 on both the macroeconomies and financial markets of different countries or regions. Some of them focused on the impact of the disease on the financial market in a single country, or the overall impact on global financial markets (Gupta & Chatterjee, 2020; Lewis, 2020; Procacci, Phelan, & Aste, 2020). Furthermore, according to Huang et al. (2020), industry-specific networks were identified based on macroeconomic data rather than data from financial markets. These studies provided us with a good incentive to design further research to illustrate how industry-specific volatility spillover networks can reflect change in economic demand.

Our study extends the literature and contributes the following:

(1) From the perspective of demand, we developed a new idea for reconstructing the structure of the industry-specific spillover network. By reorganizing the industry securities indices into demand-oriented sector groups, a better linkage between the theories of macroeconomics and the industry-specific network analysis of the financial market can therefore be obtained.
(2) We provided an early report of the structural change in the industry-specific volatility spillover networks of China’s financial market around the outbreak of COVID-19. We further analyzed how the changes in this network reflected the changes in economic demand as a result of the disease.
(3) A list of new economic implications was found from the numerical results. First, during the entire study period, there were stable spillovers from the capital goods sector group to the consumption goods sector group. The spillovers from the capital goods and equipment manufacturing sector groups, which represent the demand for investment, to other sector groups rose significantly at the beginning of the COVID-19 outbreak. However, these spillovers declined approximately one month later. Second, the level of spillovers from the unclassified services sector group was continuously rising during the whole study period. This rising trend reflected that the intensive use of non-pharmaceutical interventions (NPIs) (Lai et al., 2020) in China caused supply restrictions to services and, therefore, had an overall impact on all types of demand.

The next section introduces the data selection and preprocessing strategies used, section 3 discusses the methodology, section 4 presents the empirical study of the demand-oriented industry-specific volatility spillover network analysis based on the SWS industry classification standard, section 5 presents a further discussion, and section 6 concludes the study.

2. DATA

The study period was from January 2 to March 20, 2020. Considering the size of the spread and the progress in containing COVID-19 both inside and outside China, we divided the study period into three subperiods. Period 1 is between January 2 and January 23, 2020; period 2 is between February 3 and February 28, 2020, and period 3 is between March 2 and March 20, 2020. Periods 1, 2 and 3 have 16, 20 and 15 trading days, respectively.

We chose the SWS standard as the industry classification standard. According to this standard, securities in China’s stock market are divided into 28 sectors, which are further divided into 104 industry groups. As shown in Table 1, the SWS standard integrated the sectors into four demand-oriented sector groups.

As a supplement of Table 1, we listed the names and the codes of all industry group indices according to the SWS standard in the Appendix A. In addition, because the level 3 categories (industry) were not mentioned in this paper, we will refer to the industry group securities index as the “industry securities index” in the following sections.

 The comprehensive descriptive statistics of the log return series of all industry securities indices in different periods can be found in the Appendix B.

3. METHODOLOGY

3.1. BEKK-GARCH-Based Volatility Spillover Network

To test volatility spillovers between multiple variables, a set of bivariate BEKK-GARCH models is required. After all testing has been completed, we can identify the BEKK-GARCH-based volatility spillover networks.

3.2. Node Importance Ranking Indicators

3.2.1. Connectivity and Relative Influence

When dividing the nodes into groups, additional indicators were required to assess the importance of the groups according to Billio, Getmansky, Lo, and Pelizzon (2012). The outward spillover effect from the nodes in one group to those in other groups can be defined as “total out to other” (TOTO). Similarly, the inward spillover to one group from other groups can be defined as “total in from other” (TIFO). The TOTO and TIFO are shown in Equation 11 and Equation 12:

3.2.2. Weighted K-Shell Decomposition

In addition to the number of neighbors, the location of a node in the network is also critical for the assessment of its importance. Kitsak et al. (2010), therefore, proposed k-shell decomposition to evaluate the locational importance of the nodes. K-shell decomposition is the method that reshapes the networks into a layered structure according to their connectivity patterns.

Vanilla k-shell decomposition fails to consider the intensity of connections, thus it cannot rank the nodes for weighted networks. Garas, Schweitzer, and Havlin (2012) extended vanilla k-shell decomposition to weighted k-shell decomposition. The alternative measure for node degree is shown in Equation 14.

3.2.3. Betweenness Centrality

3.3. Earth Mover’s Distance (EMD)

None of the indicators introduced in Section 3.2 highlights measuring the distributional change in the spillover intensity. Therefore, we introduce the EMD to consider the intensity distribution change in spillovers across different groups. The EMD is a cross-bin distance that is defined as the minimal cost that must be paid to transform one histogram into another (Rubner, Tomasi, & Guibas, 2000). An intensity distribution of spillovers can be represented by countable clusters. Each cluster is represented by its mean and by the fraction of the distribution that belongs to that cluster. We refer to such a representation as the signature of the distribution. Then, the distributional change in spillover intensity between periods 1 and 2 can be formalized and solved as a transportation problem.

4. DEMAND-ORIENTED INDUSTRY-SPECIFIC VOLATILITY SPILLOVER NETWORK ANALYSIS

4.1 Network before the COVID-19 Outbreak

Table 3 reports the intergroup spillovers of the network in period 1. We can find the significant asymmetry in the spillover between the Kg and Cg groups. Both the gross and net spillovers from Kg to Cg are the highest among all gross and net intergroup spillovers, which are 49.02 and 16.22. In addition, the net spillover from Kg to Ke is 14.8. Relatively, there is only slight asymmetry in the rest of spillovers. The net spillover from Cg to Ke is much weaker than that from Kg to Cg. Thus, integrating the Kg and Ke groups as a whole sector group meeting the investment demand has a net spillover to the Cg group.

In period 1, the Kg group is a main spillover contributor from all perspectives. The outward spillovers from Kg to other groups account for 23.2% of total spillovers in the network. In contrast, the Cg group is the main receiver of spillovers, which receives 22.3% of the total spillovers. The Ke group also receives 20.3% of the total spillovers, which is only slightly lower than that of the Cg group. In addition, the other three groups have net spillovers to the Ke group.

4.2. Network at the Beginning of the COVID-19 Outbreak

Table 5 reports the intergroup spillovers of the network in period 2. We can see that the total spillover intensity and the number of spillover paths in period 2 are significantly higher than those in period 1. This means that the outbreak of COVID-19 intensified the overall spillovers in China’s stock market. The gross and net outward spillovers from Kg to Cg are 61.37 and 36.32, respectively, which account for a higher proportion of total spillovers than those in period 1. Specifically, the proportion of the gross spillover from Kg to Cg increased from 9.1% to 9.5% of the total spillovers of the network. The proportion of the net spillover increased even more rapidly from 3.0% to 5.6%. In addition, the Ke and Kg groups, as an integral whole, still have a net spillover effect on the Cg group.

In period 2, the outward spillovers from Kg to other groups account for 24.0% of the total spillovers of the network. The Cg group received 24.3% of the total spillovers. In addition, the other three groups have net outward spillovers to the Cg group, including the Ke group. Therefore, Kg and Cg can be viewed as the major contributor and receiver, respectively, of the spillovers in period 2.

4.3. Network after Preliminary Containment of Covid-19

Table 7 reports the intergroup spillovers of the network in period 3. Both the total spillover intensity and the number of spillover paths in period 3 are less than those in period 2. The most significant difference between the intergroup spillovers in period 3 and those in periods 1 and 2 is the occurrence of the net outward spillover from the Us group to the Kg group. The Us group, therefore, has the net outward spillovers to other three groups. In addition, the Ke and Kg groups, as an integral whole, still have the net spillover to the Cg group.

In period 3, the outward spillovers from the Us group account for 20.4% of the total spillovers of the network. The inward spillovers received by the Cg group accounted for 21.9% of the total spillovers. Therefore, the Us and Cg groups were the major contributor and receiver, respectively, of the spillovers in period 3.

In conclusion, there are both stable patterns and significant changes in the demand-oriented industry-specific volatility spillover networks of China’s stock market during the study period. First, viewing the Kg and Ke sector groups as an integral whole, they maintained significant net spillovers to the Cg group. Such spillovers were relatively stronger in period 2 than in the other two periods. Second, the Kg group always had a net outward spillover to the Ke group. Furthermore, the importance of the Us group increased and finally became the major contributor of the spillovers in period 3. Figure 2 depicts the simplified spillover paths of networks in different periods by sector group.

According to Figure 1, the spillover paths from Kg to Cg and Ke were stable in all periods. The outbreak of COVID-19 led to an increasing rise in the importance of the spillover paths from the Us group to other groups. In particular, the path from Us to Cg became one of the major paths of the spillover networks of China's stock market after the outbreak.

The findings of this section have inspiring economic implications. First, some studies, such as Justiniano, Primiceri, and Tambalotti (2010), proved that fluctuations in investment demand caused by exogenous factors are the main cause of fluctuations in China’s economic demand. Our numerical findings further show that the structural change in volatility spillover networks of China’s stock market can reflect the critical role that investment demand plays in the fluctuation of China's economic demand since the outbreak of COVID-19. On the one hand, the Kg and Ke groups, as an integral whole, are the stable spillover contributors to the Cg group in the networks. On the other hand, the outward spillover effect from the Kg and Ke groups to the Cg group rapidly rose at the beginning of the COVID-19 outbreak. After the outbreak was preliminarily contained, these spillovers significantly fell. One of the main types of damage caused by COVID-19 was the nationwide closure and idling of plants in all trades, which has undisputedly had an enormous impact on investment demand in China. The increased uncertainty of investment demand led to the fluctuation in stock prices of securities in the industry sectors, which supplies goods or services to meet the investment demand. Therefore, regarding the change in spillovers from the Ke and Kg sector groups to the Cg group provides empirical evidence, from the perspective of the financial market, for the economic theories proposed in literature by Greenwood, Hercowitz, and Krusell (2000).

Second, the increasingly rising importance of the Us sector group in spillover networks reveals the occurrence of supply restrictions on the service industry caused by the implementation of NPIs. To contain COVID-19, the Chinese government implemented immediate NPIs nationwide. The majority of businesses in the service sector were forced to shut down, and a large percentage of transportation services in China had to idle, despite the enormous freight and passenger traffic demands. The uncertainty of COVID-19 transformed into the uncertainty of the operational environment of the companies in the service sector and, consequently, their asset prices. According to Xu and Zhang (2020), service supply restrictions will lead to an imbalance between supply and demand and will negatively affect economic growth. As a result, companies in the service sector contribute more volatility spillovers to those in other sectors. When overseas market demand is strong, a country is still able to achieve high economic growth under the condition of service supply restrictions. However, once the overseas market demand becomes insufficient, service supply restrictions will seriously damage the economy. As introduced in section 2.1, COVID-19 began to spread outside China in period 3. As a global pandemic, COVID-19 will surely lead to insufficient overseas demand for Chinese products. As a result, the importance of the Us group in the networks in period 3 is even higher than in period 2. In addition, according to Herrendorf and Fang (2019), to compare the period in which developed countries were at a similar stage of development as China is currently, there is severe supply restriction on most service industries currently in China. Service supply restriction is an overall problem rather than a structural problem in the Chinese economy. The outbreak of COVID-19 was only an exogenous shock that intensified the problem. Therefore, we believe that our findings are still representative, although not all service industry sectors are classified as members in the Us group according to the SWS standard.

5. FURTHER DISCUSSIONS

We further discussed the demand-oriented industry-specific volatility spillover networks of China's stock market from three aspects. First, we calculated the earth mover’s distance (EMD) of the distributions of the spillover intensity of both inter- and intra-sector groups in different periods. Second, defining the major spillover paths as those with the top 20% highest intensity among a set of paths, we discussed the major spillover paths between different sector groups and their changes in different periods. Third, from various perspectives, we selected the systemically important nodes of the networks in different periods.

5.1. EMDs Between Spillover Intensity Distributions in Different Periods

Figure 2 depicts the intensity distributions of the spillovers between sector groups. Intensity distributions of the spillovers, both intergroup and intragroup, are right-skewed. Few spillover paths have high intensity. Most of the subfigures show that the intensity distributions of spillovers in period 1 are similar to their counterparts in period 3. The intensity distributions in period 2, in contrast, are significantly different from those in periods 1 and 3. This reveals that in period 2, the industry-specific volatility spillover network of China’s stock market has significant structural changes. As an exception, the intensity distribution of the intragroup spillover paths of the Us group, and of intergroup spillover paths between the Us and Cg groups in period 2, are more similar to their counterparts in period 3, rather than to those in period 1. This exception is also consistent with the findings in section 4 and proves that COVID-19 had a more long-lasting impact on the service sector than on other sectors.

Table 8 shows the EMDs between spillover intensity distributions in different periods. Between periods 1 and 2, most of the high EMDs were connected with the distribution changes in spillovers between the Us group and other groups. Specifically, the EMD of the change in the intensity distribution of spillovers from Ke to Us is 6.34%, and those from the Us group to the Ke and Cg groups are 5.11% and 4.38%, respectively. Between periods 2 and 3, most of the high EMDs were connected with the distribution changes in spillovers between the Kg group and other groups. Specifically, according to the EMD, the intensity distribution of the intragroup spillovers of the Kg group changed by 4.43%. The EMD of the change in the intensity distribution of spillovers from Kg to Us and those from Cg to Kg are 4.31% and 3.71%, respectively.

The analysis in this section is a meaningful supplement for the analysis based on the sector influence indicator in section 4. From period 1 to period 2, the Us group is the sector group of which the spillover effect strength distribution had the most significant change. It revealed that, at the beginning of the outbreak of COVID-19, the nationwide implementation of NPIs is reflected immediately in the distributional characteristics of the spillover networks of China’s stock market. The significant distributional change in the spillover strength concerning the Kg group from period 2 to period 3 also shows that the risk for investment demand destruction has been controlled to some extent. This is mainly due to the successful containment of COVID-19 and the resumption of work that is strongly supported by both the central and local governments of China. This means that the influence of the pandemic on the investment demand fell rapidly after the pandemic was contained in China, while the influence on service sectors was long-lasting.

5.2. Systematically Important Nodes in the Spillover Network

We selected the systematically important nodes in the volatility spillover networks in different periods from various perspectives.

According to Figure 3 (a), nodes in the commerce and trading sector (1202 and 1203), the construction sector (1721, 1723 and 1724), the utility sector (1161) and the transportation services sector (1175 and 1178) were the main contributors of the spillovers in all periods. Compared to period 1, a larger number of nodes in the Ke and Us groups were the main spillover contributors of the network in period 2. Compared to period 2, a larger number of nodes in the Cg and Us groups were the main spillover contributors of the network in period 3. According to Figure 3 (b), a list of nodes in the Cg group (1014, 1017, 1111, 1112, 1212, 1123, 1141, 1143 and 1156) and in the Us group (1171, 1176, 1123, 1223 and 1752) were the main spillover receivers of the network during the whole study period. After period 2, a larger number of nodes in the Kg (1712, 1037 and 1053) and Ke (1731, 1732, 1733, 1101 and 1084) groups were identified as the main spillover contributors of the network.

According to Figure 4, compared to the network in period 1, in period 2, a larger number of nodes in the Kg and Ke groups had a relatively higher betweenness centrality and k-shell level. However, in the network in period 3, a larger number of nodes in the Us group became the center of the networks. In conclusion, the analysis in section 5 further validates the main result in section 4. The spillovers from the Kg and Ke sector groups, as an integral whole, rose in period 2 and fell in period 3; the change in the spillovers from the Cg group is in contrast. Regarding the spillovers from sectors meeting the consumption demand of other sectors in China's stock market, our findings are consistent with those in the literature of Yang, Chen, and Zhang (2020) (in Chinese). However, we further illustrated how spillover networks of China’s stock market reflected the relative rise and fall of the uncertainty of investment demand and consumption demand in China during the spread of COVID-19. Yang et al. (2020) also proposed that the service industry, which has suffered due to the pandemic, is a potential threat to economic recovery in China. In addition, Huang et al. (2020) found that more difficulties would be faced by industry sectors relying on transportation services during economic recovery. Based on the spillover networks of the stock market, our findings provide evidence for these studies.

6. CONCLUSIONS

According to our empirical analysis, first, the Ke and Kg sector groups, as a whole, had stable net spillovers to the Cg sector group in all of three different periods during the breakout of COVID-19. Second, the net spillovers from the Ke and Kg groups to the Cg group rose in period 2 but fell in period 3. Third, as of period 2, the importance of the Us sector group became increasingly higher. The Us group finally played the main contributor to the spillover network of China’s stock market in period 3. We conducted further discussions from various perspectives, and all discussions validated our main result. We emphasize the need to discuss the demand change in a country. Our findings also have meaningful insights regarding economic recovery in the context of containing the spread of COVID-19. The investment demand in China suffered more than the consumption demand from the exogenous shock of COVID-19 at the beginning of the outbreak. However, when the pandemic was contained, the risk in investment demand in China was also controlled to some extent. The increasingly critical role that the Us sector group began to play revealed that the supply restriction in services is still a long-lasting threat to the next stage of Chinese economic recovery, especially under the condition that foreign demand is destroyed by COVID-19. We believe that NPIs are necessary for all countries and regions suffering from COVID-19. Thus, being aware of the overall influence of the service sector is critical for investors and policymakers globally.

REFERENCES

 Acemoglu, D., & Guerrieri, V. (2008). Capital deepening and nonbalanced economic growth. Journal of political Economy, 116(3), 467-498. Available at: https://doi.org/10.1086/589523.

Billio, M., Getmansky, M., Lo, A. W., & Pelizzon, L. (2012). Econometric measures of connectedness and systemic risk in the finance and insurance sectors. Journal of Financial Economics, 104(3), 535-559. Available at: https://doi.org/10.1016/j.jfineco.2011.12.010.

Campbell, J. Y., Lettau, M., Malkiel, B. G., & Xu, Y. (2001). Have individual stocks become more volatile? An empirical exploration of idiosyncratic risk. The Journal of Finance, 56(1), 1-43. Available at: https://doi.org/10.1111/0022-1082.00318.

Diebold, F. X., & Yilmaz, K. (2012). Better to give than to receive: Predictive directional measurement of volatility spillovers. International Journal of Forecasting, 28(1), 57-66. Available at: https://doi.org/10.1016/j.ijforecast.2011.02.006.

Engle, R. F., & Kroner, K. F. (1995). Multivariate simultaneous generalized ARCH. Econometric Theory, 11(1), 122-150.

Ewing, B. T., Forbes, S. M., & Payne, J. E. (2003). The effects of macroeconomic shocks on sector-specific returns. Applied Economics, 35(2), 201-207.

Feng, S., Huang, S., Qi, Y., Liu, X., Sun, Q., & Wen, S. (2018). Network features of sector indexes spillover effects in China: A multi-scale view. Physica A: Statistical Mechanics and its Applications, 496, 461-473. Available at: https://doi.org/10.1016/j.physa.2017.12.091.

Garas, A., Schweitzer, F., & Havlin, S. (2012). A k-shell decomposition method for weighted networks. New Journal of Physics, 14(8), 083030. Available at: https://doi.org/10.1088/1367-2630/14/8/083030.

Gonzalez-Navarro, M., & Quintana-Domeque, C. (2016). Paving streets for the poor: Experimental analysis of infrastructure effects. Review of Economics and Statistics, 98(2), 254-267. Available at: https://doi.org/10.1162/rest_a_00553.

Greenwood, J., Hercowitz, Z., & Krusell, P. (2000). The role of investment-specific technological change in the business cycle. European Economic Review, 44(1), 91-115. Available at: https://doi.org/10.1016/s0014-2921(98)00058-0.

Gupta, K., & Chatterjee, N. (2020). Examining lead-lag relationships in-depth, with focus on FX market as Covid-19 crises unfolds. arXiv preprint arXiv: 2004.10560v2.

Herrendorf, B., & Fang, L. (2019). High-skilled services and development in China (pp. 1-35). 2019 RBF Working Paper. Federal Reserve Bank of Atlanta: Atlanta,USA.

Huang, J., Wang, H., Xiong, H., Fan, M., Zhuo, A., Li, Y., & Dou, D. (2020). Quantifying the economic impact of COVID-19 in mainland China using human mobility data. arXiv preprint arXiv: 2005.03010.

Jiang, Y., Jiang, C., Nie, H., & Mo, B. (2019). The time-varying linkages between global oil market and China's commodity sectors: Evidence from DCC-GJR-GARCH analyses. Energy, 166, 577-586. Available at: https://doi.org/10.1016/j.energy.2018.10.116.

Justiniano, A., Primiceri, G. E., & Tambalotti, A. (2010). Investment shocks and business cycles. Journal of Monetary Economics, 57(2), 132-145.

Kang, S. H., Cheong, C., & Yoon, S.-M. (2013). Intraday volatility spillovers between spot and futures indices: Evidence from the Korean stock market. Physica A: Statistical Mechanics and its Applications, 392(8), 1795-1802. Available at: https://doi.org/10.1016/j.physa.2013.01.017.

Kitsak, M., Gallos, L. K., Havlin, S., Liljeros, F., Muchnik, L., Stanley, H. E., & Makse, H. A. (2010). Identification of influential spreaders in complex networks. Nature Physics, 6(11), 888-893.

Lai, S., Ruktanonchai, N. W., Zhou, L., Prosper, O., Luo, W., Floyd, J. R., . . . Tatem, A. J. (2020). Effect of non-pharmaceutical interventions to contain COVID-19 in China. Nature, 585(7825), 410–413. Available at: https://doi.org/10.1038/s41586-020-2293-x.

Lewis, A. L. (2020). US equity risk premiums during the COVID-19 pandemic. arXiv preprint arXiv: 2004.13871.

Lin, B., Wesseh Jr, P. K., & Appiah, M. O. (2014). Oil price fluctuation, volatility spillover and the Ghanaian equity market: Implication for portfolio management and hedging effectiveness. Energy Economics, 42, 172-182. Available at: https://doi.org/10.1016/j.eneco.2013.12.017.

Liu, X., An, H., Li, H., Chen, Z., Feng, S., & Wen, S. (2017). Features of spillover networks in international financial markets: Evidence from the G20 countries. Physica A: Statistical Mechanics and its Applications, 479, 265-278. Available at: https://doi.org/10.1016/j.physa.2017.03.016.

Mateus, C., Chinthalapati, R., & Mateus, I. B. (2017). Intraday industry-specific spillover effect in European equity markets. The Quarterly Review of Economics and Finance, 63, 278-298. Available at: https://doi.org/10.1016/j.qref.2016.04.011.

Opsahl, T., Agneessens, F., & Skvoretz, J. (2010). Node centrality in weighted networks: Generalizing degree and shortest paths. Social Networks, 32(3), 245-251. Available at: https://doi.org/10.1016/j.socnet.2010.03.006.

Procacci, P. F., Phelan, C. E., & Aste, T. (2020). Market structure dynamics during COVID-19 outbreak. arXiv preprint arXiv:2003.10922.

Rubner, Y., Tomasi, C., & Guibas, L. J. (2000). The earth mover's distance as a metric for image retrieval. International Journal of Computer Vision, 40(2), 99-121.

Singh, P., Kumar, B., & Pandey, A. (2010). Price and volatility spillovers across North American, European and Asian stock markets. International Review of Financial Analysis, 19(1), 55-64. Available at: https://doi.org/10.1016/j.irfa.2009.11.001.

Singh, V. K., Nishant, S., & Kumar, P. (2018). Dynamic and directional network connectedness of crude oil and currencies: Evidence from implied volatility. Energy Economics, 76, 48-63. Available at: https://doi.org/10.1016/j.eneco.2018.09.018.

Wang, Z. (2010). Dynamics and causality in industry-specific volatility. Journal of Banking & Finance, 34(7), 1688-1699.

Xu, C., & Zhang, B. (2020). The expansion of domestic demand during the transition of the economic structure: A perspective based on supply restraints in the service sector. Social Sciences in China, 2020(1), 64-83.

Yang, Z., Chen, Y., & Zhang, P. (2020). Macroeconomic shock, financial risk transmission and governance response to major public emergencies. Management World, 36(5), 13-35.

Yarovaya, L., Brzeszczyński, J., & Lau, C. K. M. (2016). Intra-and inter-regional return and volatility spillovers across emerging and developed markets: Evidence from stock indices and stock index futures. International Review of Financial Analysis, 43, 96-114. Available at: https://doi.org/10.1016/j.irfa.2015.09.004.

Yin, K., Liu, Z., & Jin, X. (2020). Interindustry volatility spillover effects in China’s stock market. Physica A: Statistical Mechanics and its Applications, 539, 1-14. Available at: https://doi.org/10.1016/j.physa.2019.122936.

Footnotes: