An analysis of Stock Markets Integration and Dynamics of Volatility Spillover

Introduction

As globalization, liberalization, and technological progress continue to reshape the interconnected global economy, the significance of understanding the integration of stock markets and the dynamics of volatility spillover become increasingly evident. The stock market integration helps to know the extent to which securities markets from different countries move together, reflecting the growing interdependence of financial markets worldwide. Volatility spillover, on the other hand, is the transmission of volatility shocks across markets, influencing price movements and risk perceptions beyond their originating market. The significance of studying stock market integration lies in its implications for portfolio diversification, risk management, and the allocation of capital across borders. Additionally, portfolio managers see emerging markets as promising destinations for diversifying funds (Kumar et al., 2017) and reaping the benefits of portfolio diversification (Vo et al., 2018). However, international investors often lack adequate understanding of emerging market dynamics, leading to herding behavior that can spike stock return volatility (Watson et al., 2012). Moreover, volatility in the stock market can stem from national and global events such as the 2008 global financial crisis or the 2019 COVID-19 pandemic, resulting in spillover effects observed in returns and volatility, typically associated with increased risk (Diebold et al., 2009). These risks prompt investors to shift their investments from financial markets to safer alternatives, exerting downward pressure on the stock market. Nevertheless, in high-volatility scenarios, investors unwilling to move their funds anticipate higher returns to compensate for the additional risk.

As markets become more integrated, the traditional benefits of diversification may diminish, as relationships between assets increase, potentially exposing investors to higher levels of systemic risk. Moreover, understanding the drivers of stock market integration can shed light on the mechanisms through which economic and financial connections are formed and evolve over time. Volatility spillover, meanwhile, presents challenges and opportunities for market participants. On one hand, increased volatility spillovers can amplify market confusion and contagion, leading to rapid and widespread fluctuations in asset prices. Considering these economic dynamics, this research is driven by the following three questions:

1. Do changes in global financial markets move the direction of the Indian stock market?
2. Is the growth of the Indian financial market driven solely by its past performance, or are there external financial factors involved?
3. Does volatility in Indian stock market returns spill over to other global markets?

If the volatility remains limited within India and doesn\'t influence other chosen markets, these selected global financial markets could serve as alternatives for portfolio diversification. Hence, knowing the importance of understanding volatility spillover and associated investment risks in financial markets, an attempt has been made to investigate the dynamics of stock market integration and volatility spillover among selected global stock markets.

Research Methodology

Data and Sample

This study examined the impact of asymmetric volatility spillover from the Indian stock market to selected global markets. The analysis focused on the NSE-500 Index, representing the Indian stock market, alongside four global market indices, namely NASDAQ (USA), SSE (China), DAX (Germany), and N225 (Japan). The daily time series data, spanning from March 2006 to March 2024, was sourced from Yahoo Finance.

Analytical Technique

To achieve the research objective, the study employed several econometric models:

• Granger Causality Test: To determine the causality between the markets.
• Vector Autoregression (VAR): To analyse the dynamic relationship among the markets.
• Dynamic Conditional Correlation (DCC): To examine the time-varying correlation between the Indian market and the selected global markets.

Limitations of the Study

This study acknowledges several limitations that may influence the interpretation and generalizability of the results:

1. The study merely focused on the volatility spillover from the Indian stock market to selected global markets, without exploring spillover in the opposite direction.
2. The research is limited to four specific global indices, which might not represent the full spectrum of diversification opportunities.
3. The chosen econometric models are robust but might not have captured all possible relationships or patterns.
4. Broader economic or geopolitical factors influencing volatility spillover are not considered.
5. The influence of regulatory changes during the study period was not explicitly analyzed.

Data Analysis

Table 1: Summary statistics of selected indices

Note: ***, ** and * denote statistical significance at 0.01%, 1%, and 5% level, respectively.

Table 1 presents the results of the descriptive and correlation analyses, as well as the unit root test. Looking at the minimum (Min) and maximum (Max) values of daily returns across chosen global financial markets, we observe that NSE-500 exhibits the varied range, reflecting a wide range of returns. The variability for all the indices are notably larger compared to their respective means, suggesting a high level of volatility in these markets. With the exception of DAX, the value of skewness is negative for all of the selected indices\' return series. This implies that the distributions are tilted toward the left side of the curve, making them asymmetrical. On the other hand, the fact that the Kurtosis value exceeds three for all the indices suggests that the data has thinner tails than a normal distribution. These results are further confirmed by the significant p-value of the Jarque-Bera test. Almost all the return series are either weakly or moderately associated with each other, and none of them are neither strongly nor negatively associated. Lastly, the results of the Automatic Data Format (ADF) test confirm that the return series for all the indices exhibit stationarity after taking the first difference.

Table 2: Granger causality test

Note: ***, ** and * denote statistical significance at 0.01%, 1%, and 5% level, respectively.

The results of the Granger causality test conducted on various pairs of stock market indices are depicted in Table 2. These tests aim at assessing whether the returns from Indian financial market (NSE-500) are useful in predicting the returns of chosen global markets or vice versa (Khan, 2023). It is evident that around 80% of the total pairs show no causality between each other, implying that there is no significant directional influence between those particular indices. However, N225 and NSE-500 confirm the existence of bi-directional causality between NASDAQ and DAX.

Table 3: Estimated results of Multivariate Vector Auto Regression (VAR) model on NSE-500

Note: ***, ** and * denote statistical significance at 0.01%, 1%, and 5% level, respectively.

Table 4: Variance decomposition results on NSE-500

Source: Author\'s calculation.

The multivariate vector autoregression model is used to examine if the returns of the Indian market are influenced by its own past returns or the past returns of selected global indices. The VAR analysis involves selecting a suitable lag order as a requirement. To determine the optimum lag order, the Akaike Information Criterion (AIC) values are used, identifying a lag order of five as most appropriate. The results from Table 3 show that the current return of the NSE-500 index is significantly influenced by its own past two returns (lag one and two), as well as the past returns of all the selected global indices at lag one. Notably, the immediate past returns of SSE and N225 have a negative effect on NSE-500. With the exception of lag order three, at least one other index\'s past returns are significant at different orders, irrespective of their signs.

The fluctuations in past values of the NSE-500 index have a significant impact on its current value, whereas the variance in past values of other financial markets has a negligible effect on the current values of NSE-500. This suggests that it is primarily the past variance within the Indian market that influences its current value, rather than external market forces from other nations. These findings are in line with previous studies by Sharma et al. (2019) and Khan (2023).

Table 5: Estimated results of Dynamic Conditional Correlation (DCC) model

Note: ***, ** and * denote statistical significance at 0.01%, 1%, and 5% level, respectively.

The dynamic conditional correlation analysis is applied to examine the integration between the selected financial markets and to measure the volatility spillover from India to other markets. The results of this analysis are presented in Table 5. The alpha1 values, signifying short-term volatility, and the beta1 values, representing long-term volatility spillover, are both positive and statistically significant. This suggests that volatility persists across all the indices. Additionally, when we sum the short and long-term volatility values for each index, the result is less than 1. This implies that volatility remains persistent and tends to worsen over time. The positive and significant values for dcca1 and dccb1 indicate that there is integration between the Indian and other global financial markets. This suggests that fluctuations in the Indian market impact and spill over into the global financial markets. These results align with previous findings from Bonga-Bonga (2018) and Khan (2023). Therefore, when there is a decline in the Indian financial market, it may trigger decays in other markets as well. So, investors might consider avoiding investments in these markets and explore diversification options in other financial markets for their portfolios.

Conclusion

In this research study, we examined the integration and volatility spillover between the Indian stock market and selected global stock markets using a daily time series spanning 2006 to 2024. The results of the Granger causality test revealed no significant causality among majority of pairs. Subsequently, the findings of VAR model show that the Indian financial market is primarily influenced by its own recent movements, as well as recent movements in selected indices. On the other hand, the past variance within the Indian market influences its current value, rather than external market forces from other nations. The significant alpha, beta and joint coefficients suggest that volatility in the Indian financial market spills over to the selected global markets. This confirms the presence of volatility spillover from the Indian financial market to global financial markets, indicating integration between these financial markets.