Annals of “Dunarea de Jos” University of Galati

Fascicle I. Economics and Applied Informatics

Years XXX – no3/2024
ISSN-L 1584-0409 ISSN-Online 2344-441X
[Link]

DOI [Link]

Examining Volatility Spillover between India and Global

Emerging Stock Markets during COVID 19 and Russia-
Ukraine War Crisis
Veerendra Anchan , Harshita Maurya, Aastha Panchamia, Aakriti Lakhanpal

A R T I C L E I N F O A B S T R A C T

Article history: The objective of this research is to examine the dynamic volatility spillover effects that
Accepted November 2024 occurred during two significant worldwide crises—the COVID-19 pandemic and the Russia-
Available online December 2024 Ukraine war—between the Indian stock market and six growing global stock markets:
JEL Classification Brazil, Argentina, Indonesia, South Africa, Saudi Arabia, and Taiwan. The data for each index
D53, E44, G41
has been taken for 10 years from 2014-2024 and has been divided into three parts: before
Keywords: COVID-19 and Russia-Ukraine war (Nov 2014-March 2020), during COVID-19 (Mar 2020-
Volatility spillover, DCC-GARCH, Dec 2021), and during Russia-Ukraine war (Feb 2022-Sep 2024). The Granger causality test
emerging markets, COVID-19, and DCC-GARCH model are used to analyse the time-varying correlations. The results reveal
Russia-Ukraine war, India, portfolio significant spillover effects, with Brazil and South Africa consistently influencing India,
diversification particularly during periods of heightened uncertainty. Brazil's continuous volatility
spillover highlights its crucial role in influencing India's market behaviour during crises,
even though markets like Argentina and Taiwan showed weaker connections. The study
provides valuable insights for portfolio managers and policymakers to optimize global
diversification strategies amid emerging market volatility and crisis-driven spillovers.

© 2024 EAI. All rights reserved.

1. Introduction
The interdependence of the global financial markets has been demonstrated in recent years, since
shocks in one region can spread to others through various channels of information transmission. With the
world becoming more interconnected, international trade and finance are growing increasingly complex, which
presents multiple challenges for academics, business professionals, and governing bodies. Understanding
global flows, their networks and evolution, and future scenarios in greater detail is necessary to navigate what
might be an increasingly complex and challenging era (Jeongmin et al. (2022)). Cross-border equity investment
has emerged as a key means of international diversification for industrialized nations and a channel for capital
flows from industrialized to developing nations since the liberalization and growth of markets in developing
nations (Feldstein (1999)). However, international portfolio investments are often unpredictable due to their
association with several risk factors, including currency rates fluctuations, political uncertainty, diverse
regulatory frameworks, and economic volatility in overseas markets, which increases their susceptibility to
adverse events in the case of a crisis. Amidst political instability such as the Russia-Ukraine conflict crisis since
early 2022 and global uncertainties like the pandemic, there was a steep decrease of nearly 15% in the
worldwide portfolio investment assets holdings (IMF 2023). The unavoidable precariousness of COVID-19
brought everyone and everything to a standstill. When discussing geopolitical issues such as the conflict
between Israel and Palestine, Russia and Ukraine, an investor's top objective is to protect and maintain the
equilibrium of their assets. It has thus become increasingly important for investors to diversify their holdings,
in lieu of the global stock markets becoming highly volatile amidst these tensions. One of the popular avenues
for investors to diversify their portfolio is to invest in both developed economies like USA, Europe, Japan and
in emerging economies like the BRIC countries (Brazil, India, Russia and China). Compared to developed
countries, emerging economies have less developed capital markets and lower family incomes,
notwithstanding their rapid development. Their economy is growing quickly, but their household incomes and
infrastructure haven't kept up with it yet.
Due to its huge reliance on imported crude oil—it is the second-largest oil importer after China—India
is extremely sensitive to interruptions in international trade and supply chains. Global oil prices rise due to
geopolitical events like the conflict between Russia and Ukraine, which increases India's import costs and local
, , ,
   
Anil Surendra Modi School of Commerce, NMIMS University, Mumbai. E-mail addresses: veerendraanchan@[Link] (V. Anchan), harrshita39maurya@[Link] (H.
Maurya – Corresponding author), aasthapanchamia19@[Link] (A. Panchamia), [Link]@[Link] (A. Lakhanpal).
inflation. Not only oil prices, but also the prices of commodities like energy, wheat and grains are affected,
which contributes to the rise in inflation of the economy. India is also dependent on imports of goods like
fertilizers, particularly from Russia and Ukraine, raw materials and intermediary goods for its textile,
automotive, and electronics manufacturing industries, and crises like these lead to disruptions in supply, which
can again lead to an increase in the prices of these commodities, thus affecting the sectors and industries as a
whole. Similarly, emerging markets like Brazil and Argentina are major exporters of agricultural commodities,
Saudi Arabia relies heavily on the exports of oil, and South Africa is a key exporter of metals like gold and
platinum. The war crisis causing global supply chain disruptions, while benefitting some countries like Saudi
Arabia, has negatively impacted emerging markets like Indonesia, South Africa and Taiwan to name a few.
These countries are also dependent on their imports of essential goods like fuel, food, and even tourism, which
was negatively affected during the pandemic. Taiwan is one of the many of these nations that is deeply
entwined in international supply networks. As one of the world's leading semiconductor producers, Taiwan is
directly impacted by trade interruptions, and crises may lower demand for its high-tech exports. India's
expanding IT industry is immediately impacted by any disruption in Taiwan's manufacturing caused by
problems with the global supply chain, as they have close trade and technological ties, particularly in the
electronics and semiconductor industries. This study is aimed at analysing the dynamic linkages and spillover
effects between the Indian market, and 6 global emerging markets like Brazil, Argentina, Indonesia, South
Africa, Saudi Arabia and Taiwan, especially during periods of major global crises like the health pandemic
COVID-19 and geopolitical instability like the Russia-Ukraine war. This paper adds to the corpus of existing
research in this field by focusing on the particular connections between India and six emerging markets. The
purpose of this study is to help global investors and policymakers navigate the intricacies of the global financial
landscape by examining the dynamic nature of these linkages.

2. Literature Review
The modern portfolio theory by Markowitz (1952) states that an investment portfolio should be
selected and constructed in such a way that it maximizes returns while minimizing the investor’s risk.
Markowitz had considered the portfolio’s variance of returns as a measure of risk, which must be minimized
using diversification. Investors will usually diversify away systematic risks by investing in different industries
and asset classes that are uncorrelated with each other. Another avenue to mitigate risk is international
portfolio diversification. Driessen & Laeven (2007) have determined that local investors in both developed and
developing nations can benefit significantly from regional and global diversification. Bekaert et al. (2011)
document the effect of financial openness and liberalization, especially in the stock market, on the growth of a
country’s economy. With an increase in liberalization, the financial markets of the world have become highly
interconnected. This leads to cross-country spillover of financial shocks in the local market, leading to a global
event like the 2008 financial crisis. Raddant & Kenett (2021) analyzed the dependencies between almost 4000
stocks and 15 countries, the results of which shows the existence of a significant amount of volatility and
interdependence between the stocks of various countries. This makes it really important for investors to
analyze these dependencies, in order to allocate their resources in the global stock market in such a way that
minimizes such volatility of returns There is an extant amount of research available analyzing the dynamic
spillover volatility of financial markets, especially the equity markets, between different developed countries
like US, Japan and major European markets. (Baele (2005); Diebold et al. (2009); Wang & Wang (2010);
Miyakoshi (2003); Kouki et al. (2011)). Due to their tremendous expansion, emerging nations account for a
sizable portion of the portfolios of international investors when considering portfolio diversification.
According to Ben Rejeb & Boughrara (2013), the increased financial liberalization policies in emerging markets
led to greater informational efficiency and reduced the likelihood of financial crises. Caporale et al. (2013)
investigated the impact of volatility spillovers from developed to emerging markets during financial shocks, as
well as the conditional correlations between them. Their findings indicated that the returns of many emerging
markets are influenced not only by volatility spillovers from developed markets but also by turbulent periods
in the mature markets. Using the GARCH-BEKK (Engle & Kroner, 1995) approach, Hung (2019) evaluated the
spillover volatility between China and four emerging stock markets in Southeast Asia. They also looked at the
returns and volatility transmissions by splitting their dataset into pre- and post-2008 global financial crises.
The analysis's findings demonstrated that, both during and after the financial crisis, the Chinese market had a
major influence on the markets that were chosen.
In recent years, the global financial markets have experienced a number of crises. With a high degree
of financial interconnectedness between nations, it becomes important to analyze the way these markets are
connected with each other, and how the shocks spills over and affects the returns and volatility of the financial
markets of these nations. This kind of analysis becomes very important for investors who are looking to
diversify their portfolio by investing in the markets of countries that are seemingly uncorrelated. However, the
advantages of global diversification lessen if there is proof of volatility spillovers during a crisis. Bensaïda et al.
(2018) analysed the volatility indices of 8 developed stock markets, the findings of which show that the
spillovers are significantly higher during times of crisis, whereas the volatility spillovers are generally mild
during normal times. Attia et al. (2023) compared portfolio diversification benefits for US investors globally

103
during different times periods, the findings of which suggest that the diversification benefits are large for
shorter time horizons, and that the COVID-19 crisis created an adverse opportunity for diversification for them.
Existing literature has examined the effects of COVID-19 on financial markets. Thangamuthu et al. (2022)
studied the effects of volatility spillover from five major international stock markets to the stock market in
India before and after the COVID-19 pandemic. They found evidence of significant volatility spillover from the
US markets to Indian stock market in the post-COVID period. Moreover, Yadav et al. (2023) also analyzed the
spillover effects with China as the dominant economy, to other emerging markets. The results showed the
unidirectional and bidirectional causality between China and the selected emerging markets. However, there
was no evidence of volatility spillover during subperiods of pre and post global financial crisis. They thus
concluded that the selected emerging economies (India, Brazil and Mexico) have attractive diversification
opportunities for investors. Sahoo and Kumar (2023) examined the dynamic volatility spillover among
developing market sustainable stock indices from countries like Malaysia, Indonesia, Thailand, South Korea,
China, and India. For this study, pre- and post-COVID-19 eras were taken into account, and ideal portfolio
weights were computed. Furthermore, the findings demonstrated a notable rise in volatility spillovers in the
aftermath of the COVID-19 pandemic. The optimization of portfolio weights was designed to provide
preference to Malaysia, a nation that significantly influenced the volatility of other Asian countries during the
crisis. Yousaf et al. (2020) examined the 2015 Chinese stock market crash and its impact on the stock markets
of Argentina, Brazil, Chile, and Mexico, finding evidence of one-way return transmission from China to the Latin
American markets during the world financial crisis. Additionally, they observed two-way volatility
transmission between the US and the stock markets of Chile and Mexico, as well as volatility spillovers between
China and Brazil during the Chinese market meltdown, indicating interconnectedness between these markets.
Green & Figlewski's (1999) analysis of eight BSE sectoral indices revealed increased market volatility during
the crisis, with only the healthcare industry showing positive returns. The banking sector had the highest
coefficient during the pandemic. The study concluded that risk-averse investors should avoid investments
during economic downturns and that geopolitical tensions negatively impact global financial markets, leading
to a loss in investor confidence and volatility spillover. (He et al. (2017); Charfeddine and Refai (2019); Mnasri
and Nechi (2016); Nikkinen et al. (2008)). During these times of uncertainty, investors tend to shift their
investments from risky global markets to safer investments, which creates a selling pressure, leading to a
downward shift in the prices of these assets (Ashraf et al. (2022); Griffith-Jones (2005)). The recent conflict
between Russia and Ukraine has significantly contributed to global financial unrest and volatility spillovers
between nations. Russia and Ukraine are both major players in the international trade segment, and the conflict
has caused disruptions in the global supply chain. This thus has global economic implications, contributing to
the spillover of volatility in markets between countries. (Yaser Almansour et al. (2023)). Raavinuthala et al.
(2024) examined the behaviour of stock markets worldwide, before and after the declaration of the Russia-
Ukraine war. The transmission of returns between the world's major stock markets was examined using DCC-
GARCH. The results showed that all asymmetric transmissions were negative and that, once word of the
declaration of war reached the markets, there was a significant increase in volatility spillovers between the
markets. Yaser Almansour et al. conducted a study on the dynamic spillover between US, Russia, and Ukraine
during the war and COVID-19 pandemic. They examined how crises affect risk management and financial
institutions' ability to make wise investment decisions. The study used empirical data from five key indices,
including the DJIA, Nasdaq, S&P 500, FTS, and MOEX. Results showed that at a 45% level of the total
connectedness index, the connectivity between these indices could explain the interconnectedness of these
economies.
Regulators, legislators, investors, and portfolio managers have long had a keen interest in the
interconnection of the financial asset markets. Amidst the geopolitical tensions between Russia and Ukraine,
Biswas et al. (2024) studied its impact on the volatility linkage of 5 key commodities and 18 international stock
exchanges. It becomes imperative to not only study the linkages between global stock markets, but also the
linkages between stock markets and different other asset classes like commodities, leading towards volatility
spillovers between these markets (Ahmed and Huo, 2021; Al-Yahyaee et al., 2019; Boubaker and Raza, 2017;
Creti et al., 2013). Because commodities are utilized as raw materials in production, a link with the stock
markets is established. Biswas et al. analysed the volatility spillover globally of both commodities as well as
equities as Russia and Ukraine are few of the major exporters of commodities like oil and natural gas, metals
like copper, iron and platinum, and agricultural commodity like wheat, which is what is considered by the
authors in this study. The study's findings demonstrate a shift in volatility spillover among commodities and
equities resulting from the Russia-Ukraine war. Following the war declaration, crude oil switched from being
a net shock transmitter to a net shock receiver, while both net exporters and importers experienced volatility
shocks in wheat and platinum. Leveraging these findings, the researchers constructed an optimized portfolio
of equities and commodities, assigning appropriate weights based on correlations and risk to create a "safe
investment" strategy. This paper aims to enhance understanding of volatility spillovers by investigating the
impact of the COVID-19 crisis and the Russia-Ukraine war on dynamic volatility spillover effects between the
Indian stock market and six emerging stock markets (Brazil, Argentina, Indonesia, South Africa, Saudi Arabia,
and Taiwan) using advanced econometric models such as the DCC-GARCH technique. Furthermore, this

104
research seeks to provide valuable data to aid investors and policymakers in navigating the complexities of the
global financial landscape and supporting economic recovery in developing nations.

3. Methodology
3.1. Data description
In this paper, the main aim is to understand the effects of dynamic volatility spillovers on the Indian
stock market during periods of crises like COVID-19 and the Russia-Ukraine war, by examining its relation with
6 global emerging countries- Brazil, Argentina, Indonesia, South Africa, Saudi Arabia and Taiwan. For this
purpose, data of daily prices of the selected countries’ respective stock exchanges (India: NIFTY 50; Brazil:
BOVESPA; Argentina: MERVAL; Indonesia: IDX Composite; South Africa: JSE; Saudi Arabia: Tadawul, and
Taiwan: TSEC), has been collected. The data has been taken from [Link]. Daily returns for every index
are computed for a ten-year period, from 2014 to 2024, following the collection and cleaning of the data. The
returns were computed using the log of each index's daily closing prices.
𝑃𝑃
𝑅𝑅𝑖𝑖 = 𝑙𝑙𝑙𝑙 � 𝑖𝑖+1�…………………….…………….(1)
𝑃𝑃𝑖𝑖
The impact of COVID-19 and the Russia-Ukraine war crisis will be analysed by segmenting our dataset
into three main time periods. Each of these sub-periods will be examined and evaluated. The three periods will
be:
1. Before COVID-19 and Russia-Ukraine war (Nov 2014- March 2020)
2. During COVID-19 (Mar 2020- Dec 2021)
3. During Russia-Ukraine war (Feb 2022- Sep 2024)

3.2. Test for stationarity

The Augmented Dickey-Fuller (ADF) test is used to examine the stationarity of the time series data.
Since non-stationary data might result in misleading regressions, stationarity is essential for time series
investigation, particularly when examining long-term correlations between variables. Stationarity is an
important assumption for many time series analysis techniques, such as forecasting, because it simplifies the
data and makes it easier to model and analyse. This will be done for each sub-period that the data has been
divided into.
𝑝𝑝
𝛥𝛥𝑦𝑦𝑡𝑡 = 𝛼𝛼 + 𝛽𝛽𝑡𝑡 + 𝛾𝛾𝑦𝑦𝑡𝑡−1 + ∑𝑖𝑖=1 𝛿𝛿𝑖𝑖 ∆𝑦𝑦𝑡𝑡−𝑖𝑖 + 𝜖𝜖𝑡𝑡 …………………(2)
Where,
∆𝑦𝑦𝑡𝑡 represents the first difference of time series, 𝛾𝛾 is the coefficient of the lagged term 𝑦𝑦𝑡𝑡 , and 𝜖𝜖𝑡𝑡 represents the
error term. For a time-series to be considered stationary 𝛾𝛾 < 0 has to be proven true. Hence, the following
hypothesis will be tested:
𝐻𝐻0 : 𝛾𝛾 = 0
𝐻𝐻1 : 𝛾𝛾 < 0

3.3. Granger Causality test

The Granger Causality test is a statistical method used to evaluate the predictability of two time series,
based on the concept that if event X1 "Granger-causes" event X2, then knowledge from past values of X1 and
past values of X2 alone should help predict X2. This test can determine the direction of volatility transmission
between different economies for each sub-period of the data. The OLS regression model that forms the basis of
the test is as follows:
𝑝𝑝 𝑝𝑝
𝑦𝑦𝑡𝑡 = 𝛼𝛼0 + ∑𝑖𝑖=1 𝛼𝛼𝑖𝑖 𝑌𝑌𝑡𝑡−𝑖𝑖 + ∑𝑖𝑖=1 𝛽𝛽𝑖𝑖 𝑋𝑋𝑡𝑡−𝑖𝑖 + 𝜖𝜖𝑡𝑡 …………………….(3)

Where, 𝛼𝛼𝑖𝑖 and 𝛽𝛽𝑖𝑖 are regression coefficients and 𝜖𝜖𝑡𝑡 is the error term. The test is based on the following null
hypothesis:
𝐻𝐻0 : 𝛽𝛽1 = 𝛽𝛽2 = ⋯ = 𝛽𝛽𝑝𝑝 = 0

When the null hypothesis is rejected, it can be concluded that event X “Granger causes” event Y.

3.4. DCC-GARCH model

To examine the spillover of volatility effects from one market to another, the GARCH model has been
proven to be mostly successful. (Smolović et al. (2017)). However, by extending univariate GARCH models to
multivariate datasets by using multivariate GARCH models like Dynamic Conditional Correlation GARCH
models (Engle & Kroner (1995)), we can have a more robust understanding of the volatility and co-movements
of the underlying indices, especially in our case where we have the Indian stock market and 6 other global
emerging markets in consideration. DCC-GARCH will help us study the time-varying correlations between
multiple stock markets.
In order to apply this model, univariate GARCH models will be first fitted for each market's return
series to capture the volatility. Thus, for each market, a univariate GARCH (1,1) model will be estimated and
the conditional variances will be found.

105
 2 2
𝜗𝜗𝑡𝑡2 = 𝜔𝜔 + 𝛼𝛼𝜖𝜖𝑡𝑡−1 + 𝛽𝛽𝜗𝜗𝑡𝑡−1 ………………………...(4)
2 2
𝜖𝜖𝑡𝑡 = 𝜗𝜗𝑡𝑡 + 𝜇𝜇𝑡𝑡 ………………………………(5)

Where,
𝜗𝜗𝑡𝑡2 represents the conditional volatilities, 𝜔𝜔 is the unconditional volatility, 𝜖𝜖𝑡𝑡2 represents the actual realized
2
volatilities, 𝜖𝜖𝑡𝑡−1 represents the lagged volatility residuals squared, and 𝜇𝜇𝑡𝑡 is the error term.
𝛼𝛼 represents the ARCH component, which explains the immediate or the short-term impact of disturbance of
conditional volatilities. The 𝛽𝛽 term represents the GARCH component, which explains the persistence of
conditional volatilities across time (long-term).

The fitted residuals will be used to estimate the time-varying correlations using the DCC model. The
multivariate time-series will be analysed with the help of the model. Once the univariate GARCH models
provide estimates of the individual conditional volatilities, the DCC model dynamically estimates the
correlation matrix. The dynamic correlation matrix 𝑅𝑅𝑡𝑡 is modelled using a GARCH type structure for the
correlations where:
𝑄𝑄𝑡𝑡 = (1 − 𝑎𝑎 − 𝑏𝑏)𝑄𝑄� + 𝑎𝑎(𝜖𝜖𝑡𝑡−1 𝜖𝜖𝑡𝑡−1
𝑇𝑇 )
+ 𝑏𝑏𝑄𝑄𝑡𝑡−1 ……………….…….(6)
Where,
𝑄𝑄𝑡𝑡 is the intermediate correlation matrix (before standardization) at time t,
𝑄𝑄� is the unconditional correlation matrix of the standardized residuals ∈𝑡𝑡 ,
𝑇𝑇
𝜖𝜖𝑡𝑡 = �𝜖𝜖1,𝑡𝑡 , 𝜖𝜖2,𝑡𝑡 , … , 𝜖𝜖𝑁𝑁,𝑡𝑡 � is the vector of standardized residuals from the univariate GARCH models.
To obtain the final dynamic correlation matrix 𝑅𝑅𝑡𝑡 , we standardize 𝑄𝑄𝑡𝑡 by its diagonal elements:
1 1
𝑅𝑅𝑡𝑡 = 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑(𝑄𝑄𝑡𝑡 )−2 𝑄𝑄𝑡𝑡 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑(𝑄𝑄𝑡𝑡 )−2 ………………………..(7)

The covariance matrix 𝐻𝐻𝑡𝑡 at time t is given by:

𝐻𝐻𝑡𝑡 = 𝐷𝐷𝑡𝑡 𝑅𝑅𝑡𝑡 𝐷𝐷𝑡𝑡 ……………………………...…….(8)

Where, 𝐻𝐻𝑡𝑡 is the conditional covariance matrix at time t, 𝐷𝐷𝑡𝑡 is 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑(𝜗𝜗𝑡𝑡 ), and 𝑅𝑅𝑡𝑡 is the time varying correlation
matrix at time t. The matrix 𝐷𝐷𝑡𝑡 captures the volatility dynamics of each individual time series, while 𝑅𝑅𝑡𝑡 models
the time-varying correlations between them.

4. Empirical results and analysis

The obtained DCC-GARCH results will be analysed for each sub-period and compared. The main point
of analysis will be to look for shifts in correlation dynamics across the crisis periods, and how different selected
economies behave with the Indian market. This section will thus follow the results and interpretation of the
models applied. R and Excel were used for all model computation and application.

4.1. Descriptive Statistics:

This study aims to examine the spillover effects from the Indian market (NIFTY50) to six selected
emerging markets, namely Brazil (BOVESPA), Argentina (MERVAL), Indonesia (IDX Composite), South Africa
(JSE), Saudi Arabia (TADAWUL), and Taiwan (TSEC). The dataset has been divided into three periods, and an
analysis has been conducted for each of the given periods, particularly during times of crises like COVID-19 and
the Russia-Ukraine war. A broad summary of the condition of the chosen markets prior to COVID-19 and the
Russia-Ukraine conflict, or the years 2014–2020, is provided in Table 1(a). Given that MERVAL (Argentina) has
the highest mean (0.103%) daily returns and the highest volatility (2.62% standard deviation), it is evident
that MERVAL stands out. It also has extreme kurtosis (85.19), suggesting very fat tails and a high likelihood of
extreme values (outliers) in the return distribution. Indices like NIFTY50, IDX and TSEC have relatively lower
mean and median returns, and also have lower standard deviations, suggesting more stability compared to
MERVAL. All indices show indications of negative skewness, implying that periods of negative returns are more
frequent than positive returns. Moreover, to apply the GARCH models, stationarity of each index was tested
using the Augmented Dickey-Fuller test. The test result gave a probability value of approximately 0.01 for each
index, suggesting that the time series of returns are indeed stationary.

106
 Table 1(a). Descriptive Statistics for the 1st period

Source: Authors’ own computation

Figure 1 and 2. Returns plots for each selected stock market for the 1st period
Source: Author generated

Similarly, the descriptive statistics for the 2nd (COVID-19) and the 3rd period (Russia-Ukraine war)
were examined as shown in Table 1(b) and Table 1(c).

Table 1(b). Descriptive Statistics for the 2nd period

Source: Authors’ own computation

107
 Table 1(c). Descriptive Statistics for the 3rd period

Source: Authors’ own computation

Figure 3 and 4. Plot of market returns of all indices for the 2nd and 3rd period respectively
Source: Author generated

Table 1(b) shows us the state of the markets during the COVID-19 period, where, in comparison to the
pre-COVID period, the majority of indexes saw lower minimum values, which suggests extremely unfavourable
market conditions during the pandemic. The countries with the largest negative minimum returns, BOVESPA
(Brazil) and TADAWUL (Saudi Arabia) (-15.99% and -8.68%, respectively), are indicative of how severe the
economic crisis is. The markets with the highest mean return throughout this crisis were MERVAL at 0.15%
and NIFTY50 at 0.1%, suggesting that some markets saw positive average performance. MERVAL has the
highest standard deviation of 2.81%, suggesting higher level of risk, when volatility increased during the
epidemic. Comparing the pre-COVID period to the present, BOVESPA, JSE, and IDX all experienced increased
volatility. When compared to the other indices, TADAWUL and TSEC have shown less volatility in this period.
The majority of indexes exhibited extreme negative skewness, with the most negatively skewed being
TADAWUL (-2.2955) and NIFTY 50 (-1.8761). This implies that there is a greater chance of extremely negative
returns during COVID. The kurtosis of the majority of indices is very high, suggesting fat tails and extreme
values. The exceptionally high kurtosis of the NIFTY50 (17.325) and BOVESPA (15.822) indicates notable
outliers and market stress during COVID. Although MERVAL's kurtosis is comparatively lower (4.635), it still
points to some extreme returns. Moving on to Table 1(c) that gives us the summary statistics of all the selected
indices during the 3rd period (Russia-Ukraine war), in comparison to the COVID period, the minimum values
are often less negative, indicating less drastic downward market movement. With a comparatively high mean
return (0.466%) and a more negative minimum return (-13.282%), MERVAL stands out once more, indicating
greater volatility in Argentina's stock market during the conflict. The mean of the NIFTY50, BOVESPA, and JSE
exhibits little fluctuation, suggesting that average returns were comparatively steady throughout the conflict.
While IDX (0.74%) and MERVAL (2.89%) remained high, most indexes saw a drop in volatility. The NIFTY 50,
TADAWUL, and BOVESPA all displayed comparatively little volatility, pointing to more stable market
circumstances in such regions amid the ongoing conflict between Russia and Ukraine. Most indexes see a minor
improvement in skewness; the JSE (0.2724) and BOVESPA (0.0574) both exhibit positive skewness, which
suggests a better likelihood of positive returns. Though not as much as during the COVID era, TADAWUL (-
0.5636) and MERVAL (-0.7825) are nonetheless negatively biased. Kurtosis sharply declines for the majority
of indices during the conflict, suggesting a decline in the frequency of severe returns. While there is still some
excess kurtosis, it is far less than during COVID in NIFTY50 (5.70) and TADAWUL (2.10). In comparison to the

108
COVID period, MERVAL (4.87) and IDX (3.18) exhibit comparable kurtosis, indicating some volatility but not
as much as before. In general, MERVAL and NIFTY50 seem to stand out as the markets that were the most
susceptible to high risk during these two periods of crises. The ADF test performed for both the periods
suggested stationarity in the time series of each index’s returns.

Figure 5 and 6. Plot of rolling covariance and correlation for the 1st period
Source: Author generated

Covariance and correlation are important statistical measures that can tell us how and to what degree
can two assets move together or against each other. To analyse the relationships in markets and economies
that are often dynamic, we use the plots of rolling covariances and rolling correlations, between the 6 selected
emerging economies and the Indian market for each sub-period under study. The relationship between all of
the indices during the first period is largely fluctuating, and the abrupt changes and variances in correlation
strength can be linked to several global macroeconomic factors, perhaps including the build-up to COVID-19 in
late 2019 and early 2020.

Figure 7 and 8. Plot of rolling covariance and correlation for the 2nd period
Source: Author generated

The plots in Figures 7 and 8 give us a better understanding of the relationships between India and
these countries during the global pandemic. The covariances between the Indian market and the markets of
Brazil, Indonesia, South Africa and Taiwan were largely stable throughout the second period (being around 0),
except when it spiked up suddenly during the end of 2021, indicating that as the initial shocks of the pandemic
were absorbed and economies began to recover, the global markets, including India, started moving more in

109
sync. The COVID-19 pandemic may have caused considerable uncertainty in global markets, as evidenced by
the large shifts in covariance and correlation toward the conclusion of the period and the numerous markets
that moved independently of each other during the crisis' peak. The covariance and correlation between India
and the Argentinian and the Saudi markets remained largely negative, with a steep drop at the end of the period,
indicating a weakening of relationships between the markets during the pandemic. The Figures 8 and 9 give us
an insight into the type of relationship between India and the chosen countries during the third sub-period, i.e.,
the Russia-Ukraine war crisis. An upward trend is observed in the plots for BOVESPA and TADAWUL, indicating
that energy prices, inflation and commodity-driven economies like Brazil and Saudi Arabia may lead towards
more robust ties with the Indian market. Whereas, countries like Argentina (MERVAL) and Indonesia (IDX)
show a weaker relationship with India, indicating that that there was minimal alignment between them over
the examined period, presumably as a result of differing economic structures or market sensitivities to the war.

Figure 8 and 9. Plot of rolling covariance and correlation for the 3rd period
Source: Author generated

4.2. Granger Causality test

After analysing the descriptive statistics for each period, we move on to analysing the Granger-
causality test results. The test will help us in determining the direction in which volatility is being transmitted
from one country to another. Tables 2(a), 2(b) and 2(c) show us the Granger-causality test results for each sub-
period.
Table 2(a). Granger-causality test results for the 1st period

Source: Authors’ own computation

110
 Table 2(b). Granger-causality test results for the 2nd period

Source: Authors’ own computation

Table 2(c). Granger-causality test results for the 3rd period

Source: Authors’ own computation

The test results show significant evidence of information flows from Brazil (BOVESPA) to India
(NIFTY50) across all three periods, suggesting that the Brazilian market may lead the Indian market, especially
during times of crises like COVID-19 and Russia-Ukraine war. However, there is no evidence that the Indian
stock market influences the Brazilian stock market. For Argentina, given that there is no discernible causality
flows between MERVAL and NIFTY50 throughout any of the periods, it is likely that information flows between
these two markets are independent of one another. Observing the Indonesian market, NIFTY50 Granger-caused
IDX during the COVID-19 timeframe, suggesting that the Indian market was leading the Indonesian market at
this time. Nonetheless, IDX roughly influences NIFTY50 (p = 0.091) during the Russia-Ukraine war, suggesting
a possible change in market dynamics. During the COVID-19 period, analysis of the Indian market in relation
to the South African stock market revealed a strong bidirectional Granger causality between NIFTY50 and JSE.
However, during the Russia-Ukraine conflict, only JSE was found to Granger-cause NIFTY50, indicating that the
South African market had a greater influence on the Indian market during that period. Regarding the Saudi
Arabian stock market, a significant bidirectional causal relationship was observed between NIFTY50 and
TADAWUL during the COVID-19 period, but this causality seemed to have disappeared during the Russia-
Ukraine conflict, suggesting a breakdown in their relationship. In the case of Taiwan, NIFTY50 was found to
Granger-cause TSEC before the pandemic, but this influence weakened during both the COVID-19 period and
the Russia-Ukraine conflict, indicating a decrease in their interdependence. In general, during uncertain times,
certain markets, such as BOVESPA and JSE, appear to have gained greater influence over NIFTY50, while other
markets, such as TADAWUL and TSEC, have shown less interdependence.

111
4.3. DCC-GARCH model
Tables 3(a), 3(b), 3(c) capture the results of the Dynamic Conditional Correlation-Generalised
Autoregressive Conditional Heteroskedasticity model (DCC-GARCH) for each sub-period for the selected 6
emerging stock markets, with the Indian stock market.

Table 3(a). DCC-GARCH output for the 1st period

Source: Authors’ own computation

Table 3(b). DCC-GARCH output for the 2nd period

Source: Authors’ own computation

112
 Table 3(c). DCC-GARCH output for the 3rd period

Source: Authors’ own computation

Our main objective is to analyse the time-varying correlations between the Indian stock market
(NIFTY50) and the other 6 emerging stock markets of the world. To this purpose, we must inspect the estimates
given by the DCC-GARCH model output. The above tables contain the following estimates: the overall mean (μ),
unconditional volatility (ω), both of which are constants, the ARCH term (α), the GARCH term (β), and the DCC
estimates- a and b. The ARCH term (α) is calculated based on the lagged volatility residuals squared, and it
represents immediate (short term) impact of disturbance of conditional volatility. The GARCH term (β) is
calculated based on the lagged conditional volatilities and it represents the persistence of conditional
volatilities over time (long term). It is important to note that the summation of α and β should always be less
than 1 for the model to be stationary, and that the closer it is to 1, there exists more persistence in the volatility
process. The spillover parameters DCC a and DCC b are the correlation estimates, where DCC a explains the
short-term perseverance, by evaluating effect of lagged volatility residuals on the correlation at hand. A larger
DCC a indicates that recent shocks have a major impact on correlations in the future. Whereas, DCC b explains
the persistence in the correlation between assets over time, a higher DCC b indicating that correlations are
more persistent over time.
Upon analysing the above outputs, we find that the α and β of the Brazilian market (BOVESPA) across
the three time periods are significant. In fact, there was a strong persistence of volatility before and during the
COVID-19 period (β = 0.869 and β = 0.859 respectively), which became even stronger during the third period
of Russia-Ukraine war crisis (β = 0.968). The short-term effect of volatility increases a bit during the 2nd period,
but reduces significantly during the 3rd period. On the other hand, another notable market- MERVAL
(Argentina) shows a high volatility clustering (α = 0.256) and a moderate persistence of volatility over time (β
= 0.713) during the period before both the crises. However, only the β estimate was significant during the
period of COVID-19, with a sharp increase in the long-term effect of volatility (β = 0.947) and a significant
decrease in the short-term effect of volatility (α = 0.051). For indices IDX, JSE, TADAWUL and TSEC, the α and
β are mostly significant (with only the α of IDX being statistically insignificant). These 4 indices are seen to be
following a similar pattern in terms of their volatilities, where they have low to moderate short-term effects of
volatility, and a high persistence in volatility in the long run, indicating stable but prolonged volatility during
the period before both the crises. During the 2nd period, a larger response to market shocks is shown by increase
in volatility clustering (α), but the long-term persistence of volatility remains high, indicating that the volatility
impacts took longer to fade. For the third period, the short-term effects of volatility are higher, especially in IDX
and TADAWUL, however, persistence somewhat declines (lower β), indicating that while markets respond
rapidly to shocks, they fade more quickly.
Looking further into the spillover parameters- DCC a and DCC b, for the 1st period as observed in Table
3(a), all the values of DCC a are insignificant, while all the values of DCC b are significant. In line with this, there
is a long-term spillover impact from the Indian stock market to the other stock markets, but not a short-term
one, thus indicating that information cannot be transferred in the short term, but it can be transferred in the
long term. The DCC b coefficients are significantly high, implying long term persistence of conditional
correlations between the Indian stock market with the other countries. However, TSEC still had a lower DCC b
(DCC b = 0.557) than the other indices. For the 2nd period, i.e., during COVID-19, there are no short-term
volatility spillover effects between NIFTY50 and the other indexes, owing to the fact that all the DCC b values
are statistically insignificant, whereas the DCC b values for all the indexes are significant, indicating long term

113
transmission of information between these countries. The DCC b values for almost every index has increased
substantially during this period, indicating that correlations between NIFTY 50 and global selected indices
became more volatile and reactive. In the Table 3(c), we observe a similar situation of all the DCC a value and
all the DCC b values being statistically insignificant and significant respectively. However, there appears to be
a dynamic shift in the correlations owing to the conflict between Russia and Ukraine. The short-term
information spillover effects have considerably increased from the previous period, whereas the long-term
persistence of correlations have substantially decreased for most of the indexes.

Figure 10. Plot of Dynamic Conditional Correlation for the 1st period
Source: Author generated

114
Figure 11. Plot of Dynamic Conditional Correlation for the 2nd period
Source: Author generated

115
 Figure 12. Plot of Dynamic Conditional Correlation for the 3rd period
Source: Author generated

The Figures 10, 11 and 12 are plots for the dynamic conditional correlations between the Indian stock
market (NIFTY50) and the selected 6 emerging global stock markets namely- Brazil (BOVESPA), Argentina
(MERVAL), Indonesia (IDX Composite), South Africa (JSE), Saudi Arabia (TADAWUL) and Taiwan (TSEC), for
all the 3 sub-periods. As depicted in Figure 10, the conditional correlation between Indian and the other
markets range from -0.02 and 0.04, -0.04 and 0.03, -0.02329 and -0.02327, -0.01 and 0.06, -0.06 and 0.04, and
-0.02 and 0.02 respectively for the selected countries. However, the pattern of correlations for the same
countries are significantly different from the 1st period, with a slight increase in values and change in trend
during both the periods of crises, as seen from Figures 11 and 12.

5. Conclusion and Discussion

The volatility spillover between the Indian market and 6 other global emerging stock markets, i.e.,
Brazil, Argentina, Indonesia, South Africa, Saudi Arabia, and Taiwan, was analysed in order to understand the
directions and magnitude of information flows between these countries. The analysis was done in three parts
corresponding to the three kinds of time periods under study: before COVID-19 (2014-2020), during COVID-
19 (2020-2021) and during the Russia-Ukraine war (2022-2024), using the multivariate DCC-GARCH model.
The findings for the three time periods are significantly different from each other, confirming the role of the
crises on the volatility spillovers between the selected countries and the Indian market. Before these crises,
Argentina's MERVAL had high returns and volatility, but weak correlation with India. Brazil, South Africa, Saudi
Arabia, and Taiwan influenced India during the pandemic, with Brazil showing unidirectional spillover and
South Africa showing strong bidirectional causality. Throughout the crises, there was consistent long-term

116
volatility spillover from Brazil and South Africa to India, as demonstrated by the DCC-GARCH model, which
supported these findings. Remarkably, Brazil and India's correlation grew stronger during the
conflict between Russia and Ukraine, most likely as a result of their shared exposure to the world's commodity
and energy markets. Even though Argentina's MERVAL was very volatile all along, its correlation
with India was weaker, suggesting that despite their volatility, the two markets are not as interconnected.
During the Russia-Ukraine war, Indonesia (IDX) and Taiwan showed relatively low but increasing short-term
volatility spillover, indicating that specific market alignments might be forming in response to global
geopolitical events. The findings underscore the significance of Brazil and South Africa as principal catalysts
for volatility spillover into the Indian market, particularly in times of crisis.
For portfolio managers, especially those looking to maximize global diversification strategies
in light of emerging market volatility and crisis-driven spillovers, the study's conclusions have a number of
significant ramifications. The findings of the study by Singhal & Ghosh (2016) demonstrate that, in order to
optimize returns and reduce risk, investors who are trying to diversify their holdings should constantly take
dynamic volatility and correlation links into account. Portfolio managers should be cautious about relying too
heavily on emerging markets like Brazil and South Africa for risk mitigation during global uncertainty. Instead,
they may need to adjust portfolio allocations dynamically to limit downside risks. Markets like Argentina and
Taiwan offer better diversification benefits during crises, but also carry volatility risks. In
order to adapt to shifting correlations and volatility spillovers during crises, portfolio managers must
implement dynamic allocation strategies. In times of crisis, for example, managers may need to lower exposure
to Brazilian assets in order to mitigate increased risks, given Brazil's enduring influence on the Indian market.
On the other hand, markets with lessened or negligible spillover effects, such as Taiwan and Saudi
Arabia, could be deliberately used as a means of diversifying. Portfolio managers should constantly monitor
the time-varying nature of volatility transmission in order to optimize their holdings, as the DCC-GARCH model
findings suggest that correlations are not static. In times of market volatility brought on by shocks linked to
politics or pandemics, employing hedging techniques, such as using options or futures, may also aid in reducing
risks. The COVID-19 pandemic and the Russia-Ukraine war are the study's primary foci, which,
while noteworthy, do not fully account for the possible volatility spillover effects from other ongoing global
crises. Due to a lack of data on the conflict's effects on the economy and markets, for example, the study does
not take the Israel-Iran conflict into consideration. More volatility in the global financial markets, especially
those reliant on oil and commodities, is expected to be brought about by the escalating geopolitical tensions in
the Middle East, particularly in relation to vital industries like energy. Moreover, this study only considers the
bivariate relationship between India and 6 emerging markets, and not the full network of interdependencies
between the 6 markets themselves. For a more comprehensive understanding of financial interconnectedness,
future research could therefore enlarge the dataset to encompass a wider range of emerging markets and asset
classes, like bonds and commodities. Further insights into volatility transmission, particularly during times of
market stress, may be obtained by using alternative econometric models, such as copula-based models, which
take into account nonlinear dependencies and tail risks. Lastly, a more detailed examination of the volatility
spillovers unique to a given sector may aid regulators and investors in better controlling portfolio risk during
financial emergencies.

Disclosure of Potential Conflicts

The authors have no competing interests to declare that are relevant to the content of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.

References
1. Ashraf, D., Rizwan, M. S., Adiwibowo, D. H., & Yusan, R. I. (2022). LOAN PORTFOLIO COMPOSITION OF ISLAMIC AND CONVENTIONAL
BANKS PRE- AND POST-COVID-19 PANDEMIC? CASE OF INDONESIA. Journal of Islamic Monetary Economics and Finance, 8(3), 407–428.
[Link]
2. Attia, E. F., Aly, S. M., ElRawas, A. said, & Awad, E. O. (2023). Portfolio diversification benefits before and during the times of COVID-19:
evidence from USA. Future Business Journal, 9(1). [Link]
3. Baele, L. (2005). Volatility Spillover Effects in European Equity Markets. In JOURNAL OF FINANCIAL AND QUANTITATIVE ANALYSIS (Vol.
40, Issue 2).
4. Behera, C. (2023). THE CRUDE OIL PRICE–STOCK RETURN CONNECTEDNESS AND THE IMPACT OF THE RUSSIAN-UKRAINE WAR ON
STOCK RETURNS IN EAST ASIAN COUNTRIES. Bulletin of Monetary Economics and Banking, 26(0), 97–110.
[Link]
5. Bekaert, G., Harvey, C. R., & Lundblad, C. (2011). Financial openness and productivity. World Development, 39(1), 1–19.
[Link]
6. Ben Rejeb, A., & Boughrara, A. (2013). Financial liberalization and stock markets efficiency: New evidence from emerging economies. In
Emerging Markets Review (Vol. 17, pp. 186–208). [Link]
7. Brana, S., Djibenou, M., & Prat, S. (2012). Global excess liquidity and asset prices in emerging countries: A PVAR approach. Emerging
Markets Review, 13(3), 256–267. [Link]

117
8. Busse, M., Dacorogna, M., & Kratz, M. (2014). The impact of systemic risk on the diversification benefits of a risk portfolio. Risks, 2(3),
260–276. [Link]
9. Caporale, G. M., Beirne, J., Schulze-Ghattas, M., & Spagnolo, N. (2013). Volatility spillovers and contagion from mature to emerging stock
markets. Review of International Economics, 21(5), 1060–1075. [Link]
10. Chittedi, K. R. (2015). Financial crisis and contagion effects to Indian stock market: ‘DCC–GARCH’ analysis. Global Business Review, 16(1),
50–60. [Link]
11. Chudik, A., & Fratzscher, M. (2010). Identifying the global transmission of the 2007–2009 financial crisis in a GVAR model. European
Economic Review, 55(3), 325–339. [Link]
12. Diebold, F. X., Yilmaz, K., Binder, M., Dominguez, K., Frankel, J., Giavazzi, F., Leeper, E., Reichlin, L., & West, K. (n.d.). MEASURING
FINANCIAL ASSET RETURN AND VOLATILITY SPILLOVERS, WITH APPLICATION TO GLOBAL EQUITY MARKETS*.
[Link]
13. Driessen, J., & Laeven, L. (2007). International portfolio diversification benefits: Cross-country evidence from a local perspective. Journal
of Banking and Finance, 31(6), 1693–1712. [Link]
14. Engle, R. F., & Kroner, K. F. (1995). Multivariate Simultaneous Generalized ARCH. In Econometric Theory (Vol. 11).
15. Feldstein, M. S. . (1999). International capital flows. University of Chicago Press.
16. Hung, N. T. (2019). Return and volatility spillover across equity markets between China and Southeast Asian countries. Journal of
Economics, Finance and Administrative Science, 24(47), 66–81. [Link]
17. Jeongmin, S., Olivia, W., Lola, W., & Sven, S. (2022, November 15). Global flows: The ties that bind in an interconnected world. McKinsey &
Company.
18. Kouki, I., Harrathi, N., & Haque, M. (2011). A Volatility Spillover among Sector Index of International Stock Markets. In Journal of Money,
Investment and Banking. [Link]
19. Khan, M., Khan, M., Kayani, U. N., Mughal, K. S., & Mumtaz, R. (2023). Unveiling Market connectedness: Dynamic returns Spillovers in
Asian emerging stock markets. International Journal of Financial Studies, 11(3), 112. [Link]
20. Khan, M. N. (2024). Market volatility and crisis dynamics: a comprehensive analysis of U.S., China, India, and Pakistan stock markets with
oil and gold interconnections during COVID-19 and Russia–Ukraine war periods. Future Business Journal, 10(1).
[Link]
21. Li, Y., & Giles, D. E. (2014). Modelling volatility spillover effects between developed stock markets and Asian emerging stock markets.
International Journal of Finance & Economics, 20(2), 155–177. [Link]
22. Markowitz, H. (1952). Portfolio Selection. In The Journal of Finance (Vol. 7, Issue 1).
23. Miyakoshi, T. (2003). Spillovers of stock return volatility to Asian equity markets from Japan and the US. Journal of International Financial
Markets, Institutions and Money, 13(4), 383–399. [Link]
24. R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
URL [Link]
25. Raavinuthala, S. K. S., Jain, G., & Patel, G. (2024). Spillovers across global stock markets before and after the declaration of Russia’s
invasion of Ukraine. Investment Management and Financial Innovations, 21(2), 130–143. [Link]
26. Raddant, M., & Kenett, D. Y. (2021). Interconnectedness in the global financial market. Journal of International Money and Finance, 110.
[Link]
27. Singhal, S., & Ghosh, S. (2016). Returns and volatility linkages between international crude oil price, metal and other stock indices in
India: Evidence from VAR-DCC-GARCH models. Resources Policy, 50, 276–288. [Link]
28. Smolović, J. C., Božović, M. L., & Vujošević, S. (2017). GARCH models in value at risk estimation: Empirical evidence from the montenegrin
stock exchange. Economic Research-Ekonomska Istrazivanja, 30(1), 477–498. [Link]
29. Thangamuthu, M., Maheshwari, S., & Naik, D. R. (2022). Volatility Spillover Effects during Pre-and-Post COVID-19 Outbreak on Indian
Market from the USA, China, Japan, Germany, and Australia. Journal of Risk and Financial Management, 15(9).
[Link]
30. Wani, S. H., & Mir, M. A. (2021). Globalisation and Economic Growth in India: an ARDL approach. The Indian Economic Journal, 69(1),
51–65. [Link]
31. Wang, P., & Wang, P. (2010). Price and volatility spillovers between the Greater China Markets and the developed markets of US and
Japan. Global Finance Journal, 21(3), 304–317. [Link]
32. Withanage, Y., & Jayasinghe, P. (2017). Volatility spillovers between south Asian stock markets: evidence from Sri Lanka, India and
Pakistan. Sri Lanka Journal of Economic Research, 5(1), 79–94. [Link]
33. Yadav, M. P., Sharma, S., & Bhardwaj, I. (2023). Volatility Spillover Between Chinese Stock Market and Selected Emerging Economies: A
Dynamic Conditional Correlation and Portfolio Optimization Perspective. Asia-Pacific Financial Markets, 30(2), 427–444.
[Link]
34. Yaser Almansour, B., Elkrghli, S., Tellez Gaytan, J. C., & Mohnot, R. (2023). Interconnectedness dynamic spillover among US, Russian, and
Ukrainian equity indices during the COVID-19 pandemic and the Russian–Ukrainian war. Heliyon, 9(12).
[Link]

118