Sección Abierta

DOI: http://dx.doi.org/10.25115/eea.v39i1.3016

Volumen:39-2 // ISSN: 1133-3197

Linkages between Gold and Emerging Asian Stock Markets:

New Evidence from the Chinese Stock Market Crash
IMRAN YOUSAF1, SHOAIB ALI2
AIR UNIVERSITY SCHOOL OF MANAGEMENT, AIR UNIVERSITY, ISLAMABAD, PAKISTAN
E-mail: imranyousaf_12@pide.edu.pk
Email: ShoaibAli@mail.au.edu.pk

ABSTRACT

This study examines the return and volatility transmission between gold and nine emerging Asian Stock Markets
during the global financial crisis and the Chinese stock market crash. We use the VAR-AGARCH model to estimate
return and volatility spillovers over the period from January 2000 through June 30, 2018. The results reveal the
substantial return and volatility spillovers between the gold and emerging Asian stock markets during the global
financial crisis and the Chinese stock market crash. However, these return and volatility transmissions vary across
the pairs of stock markets and the financial crises. Besides, we analyze the optimal portfolios and hedge ratios
between gold and emerging Asian stock markets during all sample periods. Our findings have important
implications for effective hedging and diversification strategies, asset pricing and risk management.
Key Words: Gold markets, Stock markets, optimal weights, hedge ratios, Chinese stock market crash, VAR-
AGARCH
JEL Classification: G10, G11, G13, G15

Recibido: 6 de Enero de 2020

Aceptado: 13 de Agosto de 2020
 Imran Yousaf and Shoaib Ali

1. Introduction

The stock markets play a vital role in economic development. It facilitates the economic activities
through efficient mobilisation and allocation of financial resources (Enisan & Olufisayo, 2009; Carp,
2012). Likewise, gold is also play important role in all economies, because gold is the one of the oldest
ways to store value and to create wealth. The unique qualities of gold have made it most desirable in
almost every civilization. There have been active gold markets for over 6000 years (Green, 2007). The
stock and gold use as critical assets in portfolio allocation and diversification. According to (Chkili,
2016), risk adjusted returns can be increased by adding gold in stock portfolio. The link across financial
markets in terms of return and volatility transmission is important for designing optimal portfolio
(Arouri et al., 2015), asset pricing and risk management.
The abrupt changes in volatility of certain market can lead to changes in the related market’s
volatility (Yousaf and Ahmed, 2018). Therefore, the enough knowledge of association between
different markets is important to get maximum benefit from portfolio, especially during market crisis.
During stock market crisis, normally there is an overall downward trend in stock markets, while upward
trend is observed in gold prices. According to (Frank and Hesse, 2009), many stock markets are
severely affected after collapse of Lehman Brothers during the global financial crisis. However,
(Baur and McDermott, 2010) show that the nominal gold prices rose by 42% during the global financial
crisis of 2008. As gold and many stock markets were moving in opposite direction during global
financial crisis of 2008, therefore the return and volatility spillover between gold and stock markets
can provide useful insights to portfolio managers during crisis.
Most of the previous studies focus on relationship between gold and developed stock markets. This
study focuses on examining the relation gold and emerging Asian stock markets. Because, the biggest
gold users of the world are China and India and USA respectively*. The top two biggest users of gold
belong to emerging Asian countries. The purpose of his study is twofold. First objective is to examine
the return and volatility transmission across gold and emerging Asian stock markets during two major
episodes of crisis; the global financial crisis and the Chinese stock market crash of 2015. This study
distinguishes itself based upon the Chinese stock market crash, because none of the previous study
has examined the spillover between gold and stock markets during the Chinese stock market crash of
2015.
The Chinese stock market crashed during 2015 (Han and Liang, 2017; Lieo and Ziemba, 2016; Yousaf
and Hassan, 2019; Ahmed and Huo, 2018; Yousaf et al., 2020). The CSI 300 index has reached up to
5178 points until mid-June of 2015. Then it toke roller-coaster ride and decline up to 34% in just 20
days; also losing 1000 points within just one week. After rapid decline, there were many attempts
made by Chinese authorities to overcome this problem. Around 50% of the stocks lost more than half
of their pre-crash market value. Some stocks lost more than 77% of their pre-crash market value. This
was the most dramatic stock market crisis in the history of crashes (Han and Liang, 2017). This crash
affects the many other markets of the world as well. Second objective is to estimate optimal weights
and hedge ratios for gold-Asian stocks portfolios during the global financial crisis and the Chinese stock
market crash. To estimate spillover, this study uses VAR AGARCH Model proposed by the (McAleer et
al., 2009). The rest of the paper is organized as follows: Section 1 consists of introduction. Section 2
provides the literature review. Section 3 describes the data and methodology. Section 4 consist of the
findings of this study. Finally, Section 5 concludes the whole discussion.

2. Literature Review

2.1. Gold and stock markets

There is a considerable amount of literature on relationship between gold and stock. Gold has been
gaining much attention due to its unique set of qualities. Gold is the good asset to make portfolio with

* https://www.gold.org/about-gold/gold-demand/geographical-diversity

2
Linkages between gold and emerging Asian stock markets: New evidence from the Chinese stock market crash

stocks, because gold has different volatilities, returns, and have lower correlation with stocks
(Daskalaki and Skiadopoulos, 2011). According to modern portfolio theory, the gains of international
portfolio diversification are higher, when the correlation between security returns become lower.
Michaud et al., (1996) discussed the advantages of low correlation between assets with respect to
international portfolio diversification. Therefore, the relationship between gold and stock is important
to investigate for portfolio managers.
Some studies have investigated the relationship between gold and stocks during normal period.
Arouri et al. (2015) analyse volatility transmission between Chinese stock and international gold
markets by using VAR-GARCH framework. This study finds that return and volatility transmission
between gold and Chinese stocks is significant. Moreover, the risk adjusted returns increases by adding
gold to Chinese stocks portfolio. Mensi et al. (2013) investigates the spillover between gold, oil, wheat,
and beverages prices by using VAR GARCH mode. It finds that past volatility and shocks of S&P 500
have a significant effect on gold and oil markets. Charlot and Marimoutou (2014) also find that past
shocks of S&P 500 significantly affect the volatility of silver, gold and oil markets. Mensi et al. (2017)
examine the volatility transmission between precious metals and stocks markets of US, Japan, Europe
and Asia. It finds significant volatility transmission between stocks and precious metals.
Huang et al. (2016) investigate the causality between Gold, oil and Chinese stock markets. The
results revealed that Gold, Oil and Chinese stock markets are significantly integrated. Further, causality
between three markets is heterogenous among various bands of frequency. Raza et al. (2016)
examines the effect of Gold prices on stock prices. It finds that gold prices positively and significantly
affect the BRICS stocks prices, while gold prices negatively and significantly affect the stock prices of
Mexico, Chile, Indonesia, Malaysia and Thailand. Gokmenoglu and Fazlollahi (2015) study the
relationship between gold, crude oil and stock prices of USA. It finds that US stock market prices are
significantly affected by the gold prices in the short and long run, thus stocks can be substituted by
gold.
Bouri et al. (2017) find that implied volatility of gold produces a positive and nonlinear effect on the
implied volatility of stock market of India. It also provides an evidence of strong linkages between gold,
oil and Indian markets. Tursoy and Faisal (2018) examine the effect of crude oil and gold prices on
Turkish stock market and find that gold inversely affect the stock prices. However, (Ciner et al., 2013)
find that there is a no evidence of association between gold and stock returns of US and UK between
the time frame of 1990 to 2010. The spillover between the gold and stock market is important for risk
management, portfolio design and for asset pricing. Majority studies focus on relationship between
gold and developed stock markets. This study focuses on examining the return and volatility
transmission between gold and emerging Asian stock markets.

2.2. Gold and stock markets during Crisis

Few studies have examined the link between gold and stocks during crisis or turmoil period. Baur
and Mcdermott (2010) analyses the relationship between gold and stock markets of Australia,
Switzerland, BRIC and G7 countries during two crises: Asian crisis of 1997 and Subprime crisis of 2008.
Creti et al. (2013) investigate the spillover between S&P 500 and gold returns during global crisis period
of 2007-2009. Gurgun and Unalmis (2014) analyses the spillover between gold and equity markets of
emerging and developing countries during the global financial crisis. Arouri et al. (2015) examine the
spillover between gold and Chinese stock market during recent global financial crisis 2007-2009. It
finds that gold serve as a safe haven for Chinese stock market during global financial crisis. Chouhdhary
et al. (2015) examine the mean and volatility spillover between gold and stock markets of US, UK and
JAPAN during global financial crisis. Chkili (2016) examine spillover between gold and stocks during
subprime US crisis. It finds that correlation is not stable and drop significantly during the global
financial crisis.
Bekiros et al. (2017) investigated the spillover between gold and stock during global financial crisis
of 2008-2009. It concludes that gold is a diversifier, but not a hedge or safe haven for BRICS stocks

3
 Imran Yousaf and Shoaib Ali

during normal and bearish market. Dey et al. (2018) examine the spillover between gold and stocks
during global financial crisis. It finds that gold serve as safe haven for stocks of Indian Technology
companies during global financial crisis. Wen and Cheng (2018) investigate the relation between gold,
US dollar and emerging markets stocks during different crises periods including global financial crisis
of 2008. It finds that gold act as safe haven for stocks of Brazil, Chile, the Czech Republic, Russia, South
Africa, China, India, Malaysia, and Thailand. Previous work has only focused on examining the spillover
between gold and stocks during US financial crisis of 2008 and Asian crises of 1997. Despite this
interest, no one to the best of our knowledge has studied the spillover between gold and stock markets
during Chinese market crash of 2015.

3. Data and Methodology

3.1. Data
This study uses the daily data of accepted benchmark stock indices of the nine Asian countries. The
Emerging Asian stock markets include China, India, Korea, Indonesia, Pakistan, Malaysia, Philippine,
Thailand, and Taiwan. This study selects emerging economies of Asia by using list of countries including
in MSCI emerging market index. To cover both Subprime crisis and the Chinese stock market crash,
this study uses a sample period from January 2000 to June 30, 2018. This study uses the three-time
periods including full sample (January 2000 to June 2018), the global financial crisis (August 2007 to
July 2010) and Chinese Stock market crash (June 2015 to May 2018). This study uses three-year data
for each crisis, as used by Yousaf and Hassan (2019). The index assumes to be same on non-trading
days as on the previous trading day (Ali et al. 2020). The data of stock indices has taken from Data
Stream database. To examine spillover between Gold and Stock returns, the data of daily London gold
spot prices (in the US dollars per troy ounce) has taken from the London Bullion Market Association
Homepage. The data of stock indices, trading hours, opening closing time and difference between the
time zones of countries is given in Table 1.
Table 1. Trading Hours of US, China, and Other Asian Markets
Code Index Country Open Close GMT
INDEXCSI SSE Composite Index China 09:30 15:00 (UTC +8 Hours)
SENSEX BSE SENSEX India 09:15 15:30 (UTC +5:30 Hours)
KOSPI KOSPI Index Korea 09:00 15:30 (UTC +9 Hours)
JCI Jakarta Composite Index Indonesia 09:00 16:00 (UTC +7 Hours)
FBMKLCI Kuala Lumpur Composite Index Malaysia 09:00 17:00 (UTC +8 Hours)
KSE100 KSE-100 index Pakistan 09:30 15:30 (UTC +5 Hours)
PCOMP Philip Composite (or PSEi) Index Phillipine 09:30 15:30 (UTC +8 Hours)
SET Index SET Index Thailand 10:00 16:30 (UTC +7 Hours)
Taiwan Stock Exchange Weighted
TWSE Index Taiwan 09:00 13:30 (UTC +8 Hours)

3.2. Methodology
Ling and McAleer (2003) proposed the multivariate VAR-GARCH Model to estimate return and
volatility spillover between different financial markets. For two series, the VAR GARCH model has
following specifications for the conditional mean equation:

𝑅𝑡 = 𝜇 + ∅ 𝑅𝑡−1 + 𝑒𝑡 (1)
1/2
𝑒𝑡 = 𝐷𝑡 𝜂𝑡
Rt represents a 2 × 1 vector of daily returns on the implied volatility indexes at time t, 𝜇 denotes
to a 2 × 1 vector of constants, ∅ is a 2 × 2 matrix of parameters measuring the impacts of own lagged

4
Linkages between gold and emerging Asian stock markets: New evidence from the Chinese stock market crash

and cross mean transmissions between two series, 𝑒𝑡 is the residual of the mean equation for the
stock and gold returns series at time t, 𝜂𝑡 indicates a 2 × 1 vector of independently and identically
1/2
distributed random vectors and 𝐷𝑡 = diag (√ℎ𝑡𝑆 .√ℎ𝑡𝐺 ), where, ℎ𝑡𝑠 and ℎ𝑡𝑜 representing the
conditional variances of stock and gold returns respectively, are given as
2 2 (2)
ℎ𝑡𝑆 = 𝐶𝑆2 + 𝑎11
2 𝑆
(𝑒𝑡−1 2
) + 𝑎21 𝐺
(𝑒𝑡−1 2 𝑆
) + 𝑏11 2 𝐺
ℎ𝑡−1 + 𝑏21 ℎ𝑡−1
2 2 (3)
ℎ𝑡𝐺 = 𝐶𝐺2 + 𝑎12
2 𝑆
(𝑒𝑡−1 2
) + 𝑎22 𝐺
(𝑒𝑡−1 2 𝑆
) + 𝑏12 2 𝐺
ℎ𝑡−1 + 𝑏22 ℎ𝑡−1
Equations (2) and (3) clearly reveal how shock and volatility are transmitted over time and across
the markets under investigation. Furthermore, the conditional covariance between stock and gold
returns can be estimated as follows:

(4)
ℎ𝑡𝑆,𝐺 = 𝑝𝑡 . √ℎ𝑡𝑆 .√ℎ𝑡𝐺

In above equation, 𝑝𝑡 indicates the constant conditional correlation between the stock and gold
returns. The VAR GARCH model assumes that the positive or negative shocks have same impact on
the conditional variance. To estimate the spillover between different markets, we will estimate
spillover between gold and stock markets by using the VAR AGARCH Model proposed by the (McAleer
et al., 2009). The VAR AGARCH model incorporates asymmetry in model as well. Specifically, the
conditional variance is defined as follows instead of Eq. (2) & (3):
2 2 (5)
ℎ𝑡𝑆 = 𝐶𝑆2 + 𝑎11
2 𝑆
𝐴(𝑒𝑡−1 2
) + 𝑎21 𝐺
𝐴(𝑒𝑡−1 2 𝑆
) + 𝑏11 2 𝐺
ℎ𝑡−1 + 𝑏21 ℎ𝑡−1
2 𝑆 𝑆
+ 𝑎11 𝐵[(𝑒𝑡−1 )((𝑒𝑡−1 ) < 0))]
2 2 (6)
ℎ𝑡𝐺 = 𝐶𝐺2 + 𝑎12
2 𝑆
𝐴(𝑒𝑡−1 2
) + 𝑎22 𝐺
𝐴(𝑒𝑡−1 ) + 𝑏122 𝑆 2 𝐺
ℎ𝑡−1 + 𝑏22 ℎ𝑡−1
2 𝐺 𝐺
+ 𝑎22 𝐵[(𝑒𝑡−1 )((𝑒𝑡−1 ) < 0))]
𝑆 2 𝑆 𝑆 𝐺 2
In the above equations, A(𝑒𝑡−1 ) and 𝐵[(𝑒𝑡−1 )((𝑒𝑡−1 ) < 0))] as well as 𝐴(𝑒𝑡−1 ) and
𝐺 𝐺
𝐵[(𝑒𝑡−1 )((𝑒𝑡−1 ) < 0))] reveal the association between a market's volatility and own lagged positive
as well as negative returns respectively (Lin el al., 2014). The above equation (5) & 6 show that
conditional variance of each market depends upon its own past shock and past volatility, as well as on
𝑆 2 𝐺 2
the past shock and past volatility of other(cross) markets. In equation (5), (𝑒𝑡−1 ) and (𝑒𝑡−1 ) explain
how own past shocks and gold market shocks affect the current conditional volatility of stock returns
𝑆 𝐺
respectively. Moreover ℎ𝑡−1 and ℎ𝑡−1 measure how own past volatility and gold market volatility affect
the current conditional volatility of stock market.
The estimates of VAR-AGARCH model can used to calculate optimal weights and hedge ratio. This
study follows the Kroner and Ng (1998) to calculate the optimal portfolio weights of stock and gold.
ℎ𝑔,𝑡 − ℎ𝑠𝑔,𝑡 (7)
𝑤𝑠𝑔,𝑡 =
ℎ𝑠,𝑡 − 2ℎ𝑠𝑔,𝑡 + ℎ𝑔,𝑡
0, 𝐼𝑓 𝑊𝑠𝑔,𝑡 < 0
𝑤𝑠𝑔,𝑡 = {𝑤𝑠𝑔,𝑡 , 𝐼𝑓 0 ≤ 𝑤𝑠𝑔,𝑡 ≤ 1
1, 𝐼𝑓 𝑤𝑠𝑔,𝑡 > 1
𝑤𝑠𝑔,𝑡 is the weight of stock in a one-dollar portfolio of stock and gold at time t. ℎ𝑠𝑔,𝑡 is the
conditional covariance between stock and gold, whereas ℎ𝑠,𝑡 and ℎ𝑔,𝑡 represent the conditional
variance of stocks and gold respectively. Moreover, 1-𝑤𝑠𝑔,𝑡 is the weight of gold in a one-dollar
portfoilio of stock and gold.
ℎ𝑠𝑔,𝑡 (8)
𝛽𝑠𝑔,𝑡 =
ℎ𝑔,𝑡

5
 Imran Yousaf and Shoaib Ali

𝛽𝑠𝑔,𝑡 represents the hedge ratio, suggested by the (Kroner and Sultan, 1993). A long position in
stock can be hedged with a short position in gold asset. RATS 10.0 software is used for estimation of
mean and volatility spillover between markets.

4. Empirical Results

4.1 Descriptive Analysis

The summary statistics of all Stock index and Gold returns is reported in Table 2. The mean return
of Pakistani stock market is highest, while lowest for Taiwan stock market during full sample period.
The standard deviation of Malaysia stock returns is lowest, while highest in Chinese stock market. The
standard deviation of China, India, Indonesia, Korea, Philippine, Taiwan and Thailand is almost equal
or greater than the Pakistan stock market, but the returns of Pakistani stock market is much higher
than all these emerging Asian markets during full sample period. Skewness is negative in most cases
and kurtosis is higher than 3 in all cases. Jarque-Bera statistics does not accept the hypothesis of
normality of all series. Moreover, there is significant evidence of autocorrelation for all series. Lastly,
there is strong evidence of ARCH effect in all series. Table 3 presents the results of unit root test for all
series. This study applies Augmented Dickey Fuller (ADF) test and Phillip -Perron test to check
stationarity of all series. Each test is applied three times, with ‘no constant and trend’, ‘constant’ and
‘constant and trend’. The results depict that all series are significant while applying ADF and Phillip-
Perron test, so there is a no evidence of stationarity in any series.

4.2 Return and Volatility Spillover Analysis

4.2.1 Linkages between Gold and Asian stock markets during the full Sample period
Table 4 represents the results of return and volatility spillover between Gold and Asian emerging
stock market during full sample period. The current stock returns of all Asian markets are significantly
affected by their lag returns. This shows that change in past stock price helps to explain the current
stock prices for emerging Asian stock markets. Furthermore, lagged gold returns significantly affect the
current gold returns.
The return spillover from gold to China, Indonesia, Malaysia, Taiwan, and Thailand stock markets is
found to be significant. This reveals that mean effect of gold market is significantly transmitted to the
majority of Asian stock markets. Moreover, the return spillover from all Asian stock markets to Gold
market is not significant except India, Korea, and Malaysia.
ARCH coefficient captures the shock dependence, while GARCH coefficient Captures the
persistence of volatility in conditional variance equations. The results reveal that the sensitivity of past
own shocks (ARCH term) is significantly positive for all Asian stock markets in short run. Furthermore,
the sensitivity of past own volatility (GARCH term) is also significant for all Asian emerging stock
markets, therefore ARCH (1) model is appropriate for estimation purpose as well. The coefficient of
past own shock is smaller as compared to the coefficient of the past own volatility in all Asian emerging
stock markets, suggesting that past own volatility is more important factor in explaining the current
volatility as compared to past own shocks. In addition, sensitivity of past own shocks and volatility is
significantly affecting the conditional volatility of the Gold markets.
The Gold market Shocks significantly affect the conditional volatility of Indonesia, Malaysia, and
Philippine stock markets. Therefore, gold market shocks lead to decrease the volatility of Malaysia and
Philippine stock markets. Conditional volatility of gold market is significantly affected by the shocks in
India, Indonesia, Korea, Philippine, and Taiwan stock markets.
The volatility spillover is not significant from the gold to the majority of Asian emerging stock
markets (except Malaysia). However, there is a presence of volatility spillover from gold to Malaysian
stock market. Furthermore, the volatility transmission is not significant from the majority of Asian stock
markets (Except Korea, Philippine, Thailand, and Taiwan) to the gold market. Asymmetric coefficients
of all Asian stock markets are significant and positive, showing that negative news (or unexpected

6
Linkages between gold and emerging Asian stock markets: New evidence from the Chinese stock market crash

shock) of gold market has more ability to increase the volatility of all Asian Stock markets as compared
to positive news. In addition, asymmetric coefficient of gold market is predominantly negatively
significant, showing that negative unexpected shock of emerging Asian Stock markets (except Pakistan)
will decrease the volatility more in gold markets as compared to the positive shock.
Constant conditional correlation (CCC) is positively significant for the majority of pairs of gold and
stock markets. The diagnostic tests on residuals are performed to check the adequacy of the model.
AIC and SIC test are used to measure the relative goodness of fit of estimated model. Jarque -Bera test
is used for the normality based on kurtosis and skewness. Q-statistics test is used to check the
autocorrelation in standardized residuals and squared standardized residuals respectively, results
show that there is a no evidence of autocorrelation in the majority of estimations.

4.2.2 Linkages between Gold and Asian stock markets during the global financial Crisis
Table 5 reports the results of mean and volatility spillover between gold and Asian emerging stock
market during the global financial crisis. The lag returns significantly affect the current returns of all
emerging Asian stock markets. Moreover, current gold returns are significantly affected by their lag
returns during the global financial crisis. These findings reveal that lag return is the important factor in
explaining the current returns of the gold and emerging Asian stock markets.
Mean spillover is not significant from gold to the majority of emerging Asian stock markets (except
Indonesia and Malaysia). In addition, the mean transmission is not significant from the majority of
emerging Asian stock markets (except India) to gold markets during the global financial crisis. The past
own shocks of India, Malaysia, Philippine, and Thailand significantly affect the conditional volatility.
Furthermore, the conditional volatility of gold market is significantly affected by their own past shocks
during subprime financial crisis.
The sensitivity of past own volatility is significant for all emerging Asian stock markets. Moreover,
coefficient of past own volatility is also significant for gold market during the global financial crisis. The
coefficient of past own volatility is higher than the coefficient of the past own shock in all Asian stock
markets, suggesting that past volatility is the more important factor in explaining the current volatility
in Asian stock markets as compared to the their past shocks.
The shocks in gold markets significantly affect the conditional volatility of Indonesia, Malaysia, and
Taiwan stock markets. Moreover, the conditional volatility of gold market is significantly affected by
shocks in China, Korea, Philippine, Taiwan, and Thailand stock markets during the global financial crisis.
The past volatility of Indonesia, Malaysia and Taiwan stock markets are significantly affected by the
volatility in gold markets. The conditional volatility of gold markets is significantly affected by the past
volatility of Indonesia, Korea, Philippine, and Thailand stock markets during the global financial crisis.
Asymmetric coefficients of all Asian stock markets are significant and positive, showing that negative
news (or unexpected shock) of gold market has more ability to increase the volatility of all Asian Stock
markets as compared to positive news. Constant conditional correlation (CCC) is positively significant
for the majority of pairs of gold and stock markets except Pakistan and Philippine.

4.2.3 Linkages between Gold and Asian stock markets during the Chinese Stock Market Crash
Table 6 represents the mean and volatility transmission between gold and Asian stock markets
during the Chinese stock market crash. The one period lagged stock returns significantly affect the
current stock returns of the majority of Asian stock markets (except Korea) during the Chinese stock
market crash. Thus, past returns explain the current returns of the majority of Asian stock markets
during the Chinese stock market crash. Moreover, past returns of gold market do not significantly
affect the current returns during the Chinese stock market crash.
The mean transmission from Gold to China, India, Pakistan, and Thailand stock markets is significant
during the Chinese stock market crash. However, India, Indonesia, and Malaysia stock markets transmit
mean effect to gold market. The coefficient of past own shock is significant in China, India, Indonesia,
Malaysia, and Taiwan stock markets. Moreover, the sensitivity of past own volatility is significant in all

7
 Imran Yousaf and Shoaib Ali

Asian stock markets. The coefficient of past own volatility is higher than the coefficient of past own
shocks, so past own volatility plays important role in estimating the current volatility in Asian Stock
markets. The past shocks of gold market significantly affect the conditional volatility of China,
Indonesia, Korea, Malaysia, and Taiwan stock markets during the Chinese stock market crash. The past
shocks of the majority of Asian stock markets (except Malaysia) do not significantly affect the
conditional volatility of gold market during Chinese Stock market crash.
There is a significant evidence of volatility transmission from gold to China, Korea, and Malaysia
stock markets. However, there is a significant evidence of volatility spillover from India, Indonesia,
Malaysia, and Thailand stock markets to gold Market during the Chinese stock market crash.
Asymmetric coefficient of the majority of emerging Asian stock markets is significant and positive
except China stock market. Furthermore, the asymmetric coefficient of gold market is also significant
and positive in all cases. Constant conditional correlation is negatively significant, when Gold asset is
paired with India and Taiwan stock market.

4.3 Optimal weights and hedge ratios of stock and gold portfolio
Table 7 presents the optimal weights and hedge ratios for stock-gold portfolio. The average optimal
weight range is 0.38 for CHN/Gold to 0.68 for MYS/Gold during full sample period, indicating that for
portfolio of $1, 38 cents should be invested in Chinese stocks and remaining 62 cents in gold market.
In addition, in $1 Malaysian stock-gold portfolio, 68 cents should be invested in Malaysian stocks and
reaming 32 cents in gold market during full sample period. However, average optimal portfolio weight
ranges from 0.31 for CHN/Gold to 0.72 for MYS/Gold during US sub-prime crisis. Majority Asian optimal
stock weights are found to be lower during the global financial crisis as compared to stocks weights
during full sample period. For the portfolio of gold-Asian stocks, investors are suggested to reduce
their investment in the majority of Asian stocks and increase their investment in gold during the global
financial crisis. Moreover, the average optimal portfolios vary from 0.42 for PHL/Gold to 0.73 for
MYS/Gold during the Chinese stock market crash. However, the average optimal weights of stocks in
stock-gold portfolio are found to be higher during the Chinese stock market crash as compared to the
optimal stock weights during full sample. For the portfolio of gold-Asian stock, investors are suggested
to increase their investment in Asian stock and decrease their investment in gold during the Chinese
stock market crash.
The average hedge ratio range is 0.05 for THA/Gold to 0.10 for INDO/Gold during full sample period,
indicating that a $1 long position in Thailand stocks can be hedged for 5 cents with short position in
Gold. In addition, a $1 long position in Indonesia can be hedged for 10 cents with short position in gold
during full sample period. Furthermore, average hedge ratio range is -0.02 for PAK/GOLD to 0.20 for
IND/Gold during the global financial crisis. The average hedge ratio of Stock/Gold is found to be higher
for the majority of Asian markets during the global financial crisis as compared to the average hedge
ratios during full sample period. However, the average hedge ratio range is -0.07 for IND/Gold to 0.06
for CHN/Gold during the Chinese stock market crash. Majority of the Asian markets has negative hedge
ratio during the Chinese stock market crash, indicating that one dollar invested in Asian stocks (except
China, Indonesia, and Philippine) can be hedged by buying gold. The lowest hedging ratio is -0.07 for
IND/Gold pair, indicating that one-dollar investment in Indian stock can be hedged by buying 7 cents
of gold during Chinese stock market Crash. Overall, optimal weights and hedge ratios are different
during the global financial crisis and the Chinese stock market crash. These optimal weights and hedge
ratios have important insights for portfolio managers to reduce their risk and to maintain returns
during crisis in Chinese or US markets.

5. Conclusion

This paper examines the return and volatility transmission between gold and nine emerging Asian
Stock Markets: China, India, Indonesia, Korea, Malaysia, Pakistan, Philippine, Taiwan, and Thailand.
This paper investigates the spillover during three periods, full sample, the global financial crisis and

8
Linkages between gold and emerging Asian stock markets: New evidence from the Chinese stock market crash

Chinese Stock market crash. To estimate spillover, it uses the VAR-AGARCH Model (McAleer et al.,
2009).
The results indicate that the return spillover is unidirectional from gold to Indonesia and Malaysia
stock markets, and from India stock market to gold market during the global financial crisis. However,
returns transmission is unidirectional from gold to China, Pakistan, and Thailand stock markets, and
from Indonesia and Malaysia stock markets to gold market during the Chinese stock market crash. In
addition, return spillover is bidirectional between the gold and India stock market during the Chinese
stock market crash.
Moreover, the volatility spillover analysis reveals the unidirectional transmission from gold to
Malaysia and Taiwan stock markets, and from Korea, Philippine, and Thailand stock market to gold
market during the global financial crisis. The volatility spillover is bidirectional between the gold and
Indonesia stock market during the global financial crisis. It implies that investor can get maximum
benefit of diversification by making a portfolio pairs of gold with the China, India, and Pakistan stock
markets during the global financial crisis. On the other hand, volatility spillover is unidirectional from
gold to China and Korea stock markets, and from India, Indonesia, and Thailand stock markets to gold
market during the Chinese stock market crash. In addition, the volatility spillover is bidirectional
between gold and Malaysia stock market during the Chinese stock market crash. It implies that
portfolio managers can get maximum benefit of diversification by investing in making a portfolio pairs
of gold with the Pakistan Philippine, Taiwan stock markets during the Chinese stock market crash.
The majority of the optimal weights of the emerging Asian stocks are found to be lower during the
global financial crisis as compared to stocks weights during full sample period. For the portfolio of gold-
Asian stocks, investors are suggested to reduce their investment in the majority of Asian stocks and
increase their investment in gold during the global financial crisis. Moreover, the average optimal
weights of stocks in stock-gold portfolio are found to be higher during the Chinese stock market crash
as compared to the optimal stock weights during full sample period. For the portfolio of gold-Asian
stock, investors are suggested to increase their investment in Asian stock and decrease their
investment in gold during the Chinese stock market crash. These findings have useful insights for
portfolio diversification, asset pricing and risk management.

References

1. Ahmed, A. D., & Huo, R. (2018). Impacts of China's crash on Asia-Pacific financial integration:
Volatility interdependence, information transmission and market co-movement. Economic
Modelling. 79, 28-46.
2. Akbar, M., Iqbal, F., & Noor, F. (2019). Bayesian analysis of dynamic linkages among gold price, stock
prices, exchange rate and interest rate in Pakistan. Resources Policy, 62, 154-164.
3. Ali, S., Yousaf, I., & Naveed, M. (2020). Role of credit rating in determining capital structure:
Evidence from non-financial sector of Pakistan. Studies of Applied Economics, 38(3).
4. Arouri, M. E. H., Lahiani, A., & Nguyen, D. K. (2015). World gold prices and stock returns in China:
insights for hedging and diversification strategies. Economic Modelling, 44, 273-282.
5. Basher, S. A., & Sadorsky, P. (2016). Hedging emerging market stock prices with oil, gold, VIX, and
bonds: A comparison between DCC, ADCC and GO-GARCH. Energy Economics, 54, 235-247.
6. Baur, D. G., & Lucey, B. M. (2010). Is gold a hedge or a safe haven? An analysis of stocks, bonds and
gold. Financial Review, 45(2), 217-229.
7. Baur, D. G., & McDermott, T. K. (2010). Is gold a safe haven? International evidence. Journal of
Banking & Finance, 34(8), 1886-1898.
8. Bekiros, S., Boubaker, S., Nguyen, D. K., & Uddin, G. S. (2017). Black swan events and safe havens:
The role of gold in globally integrated emerging markets. Journal of International Money and
Finance, 73, 317-334.
9. Bollerslev, T. (1986). Generalized autoregressive conditional heteroskedasticity. Journal of
econometrics, 31(3), 307-327.

9
 Imran Yousaf and Shoaib Ali

10.Bollerslev, T., Engle, R. F., & Wooldridge, J. M. (1988). A capital asset pricing model with time-
varying covariances. Journal of political Economy, 96(1), 116-131.
11.Bouri, E., Jain, A., Biswal, P. C., & Roubaud, D. (2017). Cointegration and nonlinear causality amongst
gold, oil, and the Indian stock market: Evidence from implied volatility indices. Resources Policy, 52,
201-206.
12.Carp, L. (2012). Can stock market development boost economic growth? Empirical evidence from
emerging markets in Central and Eastern Europe. Procedia Economics and Finance, 3, 438-444.
13.Charlot, P., & Marimoutou, V. (2014). On the relationship between the prices of oil and the precious
metals: Revisiting with a multivariate regime-switching decision tree. Energy Economics, 44, 456-
467.
14.Chkili, W. (2016). Dynamic correlations and hedging effectiveness between gold and stock markets:
Evidence for BRICS countries. Research in International Business and Finance, 38, 22-34.
15.Choudhry, T., Hassan, S. S., & Shabi, S. (2015). Relationship between gold and stock markets during
the global financial crisis: Evidence from nonlinear causality tests. International Review of Financial
Analysis, 41, 247-256.
16.Ciner, C., Gurdgiev, C., & Lucey, B. M. (2013). Hedges and safe havens: An examination of stocks,
bonds, gold, oil and exchange rates. International Review of Financial Analysis, 29, 202-211.
17.Creti, A., Joëts, M., & Mignon, V. (2013). On the links between stock and commodity markets'
volatility. Energy Economics, 37, 16-28.
18.Daskalaki, C., & Skiadopoulos, G. (2011). Should investors include commodities in their portfolios
after all? New evidence. Journal of Banking & Finance, 35(10), 2606-2626.
19.Dey, S., & Sampath, A. (2018). Dynamic linkages between gold and equity prices: Evidence from
Indian financial services and information technology companies. Finance Research Letters, 25, 41-
46.
20.Engle, R. F. (1982). Autoregressive conditional heteroscedasticity with estimates of the variance of
United Kingdom inflation. Econometrica: Journal of the Econometric Society, 987-1007.
21.Engle, R. F., & Kroner, K. F. (1995). Multivariate simultaneous generalized ARCH. Econometric
theory, 11(1), 122-150.
22.Enisan, A. A., & Olufisayo, A. O. (2009). Stock market development and economic growth: Evidence
from seven sub-Sahara African countries. Journal of Economics and Business, 61(2), 162-171.
23.Frank, N., & Hesse, H. (2009). Financial spillovers to emerging markets during the global financial
crisis (No. 9-104). International Monetary Fund.
24.Gokmenoglu, K. K., & Fazlollahi, N. (2015). The interactions among gold, oil, and stock market:
evidence from S&P500. Procedia Economics and Finance, 25, 478-488.
25.Green, T. (2007). The ages of gold. London: Gold Fields Minerals Services Ltd.
26.Gürgün, G., & Ünalmış, İ. (2014). Is gold a safe haven against equity market investment in emerging
and developing countries?. Finance Research Letters, 11(4), 341-348.
27.Han, Q., & Liang, J. (2017). Index futures trading restrictions and spot market quality: Evidence from
the recent the Chinese stock market crash. Journal of Futures Markets, 37(4), 411-428.
28.He, L. T. (2001). Time variation paths of international transmission of stock volatility—US vs. Hong
Kong and South Korea. Global Finance Journal, 12(1), 79-93.
29.He, Z., O'Connor, F., & Thijssen, J. (2018). Is gold a Sometime Safe Haven or an Always Hedge for
equity investors? A Markov-Switching CAPM approach for US and UK stock indices. International
Review of Financial Analysis, 60, 30-37.
30.Hood, M., & Malik, F. (2013). Is gold the best hedge and a safe haven under changing stock market
volatility?. Review of Financial Economics, 22(2), 47-52.
31.Huang, S., An, H., Gao, X., & Huang, X. (2016). Time–frequency featured co-movement between the
stock and prices of crude oil and gold. Physica A: Statistical Mechanics and its Applications, 444,
985-995.
32.Kroner, K. F., & Ng, V. K. (1998). Modeling asymmetric comovements of asset returns. The review
of financial studies, 11(4), 817-844.

10
Linkages between gold and emerging Asian stock markets: New evidence from the Chinese stock market crash

33.Kroner, K. F., & Ng, V. K. (1998). Modeling asymmetric comovements of asset returns. The review
of financial studies, 11(4), 817-844.
34.Kroner, K. F., & Sultan, J. (1993). Time-varying distributions and dynamic hedging with foreign
currency futures. Journal of financial and quantitative analysis, 28(4), 535-551.
35.Li, Y., & Giles, D. E. (2015). Modelling volatility spillover effects between developed stock markets
and Asian emerging stock markets. International Journal of Finance & Economics, 20(2), 155-177.
36.Ling, S., & McAleer, M. (2003). Asymptotic theory for a vector ARMA-GARCH model. Econometric
theory, 19(2), 280-310.
37.Lleo, S., & Ziemba, W. T. (2016). The bond-stock earnings yield differential model: additional
applications and other models for stock market crash prediction. Quantitative Finance Letters, 4(1),
26-34.
38.Mandelbrot, B. (1963). The variation of certain speculative prices. Journal of Business 36, 394–419.
39.McAleer, M., Hoti, S., & Chan, F. (2009). Structure and asymptotic theory for multivariate
asymmetric conditional volatility. Econometric Reviews, 28(5), 422-440.
40.Mensi, W., Al-Yahyaee, K. H., & Kang, S. H. (2017). Time-varying volatility spillovers between stock
and precious metal markets with portfolio implications. Resources Policy, 53, 88-102.
41.Mensi, W., Beljid, M., Boubaker, A., & Managi, S. (2013). Correlations and volatility spillovers across
commodity and stock markets: Linking energies, food, and gold. Economic Modelling, 32, 15-22.
42.Mensi, W., Hammoudeh, S., Reboredo, J. C., & Nguyen, D. K. (2015). Are Sharia stocks, gold and US
Treasury hedges and/or safe havens for the oil-based GCC markets?. Emerging Markets Review, 24,
101-121.
43.Michaud, R. O., Bergstrom, G. L., Frashure, R. D., & Wolahan, B. K. (1996). Twenty years of
international equity investing. The Journal of Portfolio Management, 23(1), 9-22.
44.Raza, N., Shahzad, S. J. H., Tiwari, A. K., & Shahbaz, M. (2016). Asymmetric impact of gold, oil prices
and their volatilities on stock prices of emerging markets. Resources Policy, 49, 290-301.
45.Shahzad, S. J. H., Raza, N., Shahbaz, M., & Ali, A. (2017). Dependence of stock markets with gold
and bonds under bullish and bearish market states. Resources Policy, 52, 308-319.
46.Tursoy, T., & Faisal, F. (2018). The impact of gold and crude oil prices on stock market in Turkey:
Empirical evidences from ARDL bounds test and combined cointegration. Resources Policy, 55, 49-
54.
47.Tuysuz, S. (2013). Conditional correlations between stock index, investment grade yield, high yield
and commodities (gold and oil) during stable and crisis periods. International Journal of Economics
and Finance, 5(9), 28-44.
48.Wen, X., & Cheng, H. (2018). Which is the safe haven for emerging stock markets, gold or the US
dollar?. Emerging Markets Review, 35, 69-90.
49.Yousaf, I., & Ahmed, J. (2018). Mean and Volatility Spillover of the Latin American Stock Markets.
Journal of Business & Economics, 10(1), 51-63.
50.Yousaf, I., Ali, S., & Wong, W. K. (2020). Return and Volatility Transmission between World-Leading
and Latin American Stock Markets: Portfolio Implications. Journal of Risk and Financial
Management, 13(7), 148.
51.Yousaf, I., & Hassan, A. (2019). Linkages between crude oil and emerging Asian stock markets: New
evidence from the Chinese stock market crash. Finance Research Letters, 31(4), 207-217.

11
 Table 2. Summary statistics
Mean Std. Dev. Skewness Kurtosis J-B Q-stat (6) ARCH
GOLD 0.00035 0.011 -0.243 8.218 5519.3 a
22.88 a
109.13a
CHN 0.00045 0.016 -0.317 8.215 5547.5a 54.64 a 180.13a
IND 0.00050 0.015 -0.222 10.542 11474.2a 84.62 a 283.82a
KOR 0.00028 0.015 -0.573 9.649 9149.3a 24.06 a 210.03a
INDO 0.00046 0.014 -0.854 10.924 13206.4a 154.0 a 457.64a
MYS 0.00020 0.008 -0.855 13.331 22039.0a 226.83a 267.31a
PAK 0.00081 0.014 -0.349 6.838 3058.0a 165.52a 594.67a
PHL 0.00038 0.013 0.230 19.783 56658.4a 96.41a 161.13a
THA 0.00044 0.013 -0.705 12.862 19948.6a 70.20a 656.23a
TAIW 0.00018 0.014 -0.275 6.596 2659.7a 77.69a 201.58a
Notes. CHN-China, IND-India, KOR-Korea, INDO-Indonesia, MYS-Malaysia, PAK-Pakistan, PHL-Philippine, THA-Thailand, TAIW-Taiwan. J-B, Q-stat denote Jarque Bera Ttest and Ljung-Box Q statistic respectively. a,b,c
indicate the statistical significance at 1%, 5% and 10% respectively.

Table 3. Unit Root Tests

ADF (t-test) Phillips-Perron test
None Constant Constant and Trend None Constant Constant and Trend
GOLD -66.741 -66.8016 -66.8110 -66.743 -66.8028 -66.8142
CHN -30.991 -31.038 -31.047 -66.482 -66.401 -66.394
IND -61.457 -61.511 -61.512 -61.336 -61.327 -61.326
KOR -64.997 -65.010 -65.004 -64.899 -64.908 -64.902
INDO -58.437 -58.486 -58.481 -58.461 -58.535 -58.529
MYS -56.326 -56.349 -56.343 -56.611 -56.607 -56.601
PAK -34.616 -34.793 -34.791 -62.501 -62.258 -62.254
PHL -60.840 -60.878 -60.877 -60.659 -60.760 -60.759
THA -44.099 -44.160 -44.159 -64.137 -64.163 -64.157
TAIW -33.808 -33.815 -33.819 -63.935 -63.937 -63.935
Notes. CHN-China, IND-India, KOR-Korea, INDO-Indonesia, MYS-Malaysia, PAK-Pakistan, PHL-Philippine, THA-Thailand, TAIW-Taiwan.

12
 Table 4. Estimates of bivariate VAR-AGARCH for Gold and Emerging Asian stock markets during the full sample period

CHN GOLD IND GOLD KOR GOLD INDO GOLD MYS GOLD PAK GOLD PHL GOLD THA GOLD TAIW GOLD
Panel A: Mean Equation
Constant 0.000b 0.000a 0.000a 0.000a 0.000b 0.000a 0.001a 0.000a 0.000b 0.000a 0.001a 0.000 0.000b 0.000a 0.001a 0.000b 0.000c 0.000a
𝒓𝑺𝒕−𝟏 0.056a 0.007 0.138a 0.016c 0.082a 0.020b 0.151a 0.013 0.193a 0.037b 0.182a 0.011 0.148a 0.012 0.118a 0.019 0.101a 0.000
𝒓𝑮𝒕−𝟏 -0.027c 0.051a 0.016 0.050a -0.002 0.053a 0.043a 0.049a 0.022a 0.052a 0.012 0.035 0.015 0.050a 0.054a 0.052a 0.041a 0.053a
Panel B: variance Equation
Constant 0.000a 0.000a 0.000a 0.000a 0.000a 0.000a 0.000a 0.000a 0.000a 0.000a 0.000 0.000a 0.000a 0.000a 0.000 0.000c 0.000a 0.000a
(𝒆𝑺𝒕−𝟏 )𝟐 0.066a -0.002 0.052a 0.023a 0.026a 0.008c 0.070a 0.015b 0.082a 0.000 0.044a 0.004 0.053a 0.020b 0.072a 0.003 0.032a 0.009a
(𝒆𝑮𝒕−𝟏 )𝟐 0.001 0.077a -0.002 0.078a 0.002 0.077a 0.006c 0.076a -0.011a 0.082a 0.014 0.020 -0.001b 0.081a 0.007 0.044a -0.002 0.080a
𝒉𝑺𝒕−𝟏 0.920a 0.006 0.869a -0.004 0.931a -0.013b 0.830a 0.012 0.873a 0.001 0.795a 0.027 0.798a -0.019c 0.835a 0.057a 0.920a -0.007c
𝒉𝑮𝒕−𝟏 0.001 0.915a 0.002 0.917a -0.003 0.919a -0.005 0.920a 0.015a 0.911a 0.006 0.903a 0.004 0.915a -0.006 0.959a 0.001 0.915a
Asymmetry 0.022a -0.021b 0.113a -0.022b 0.075a -0.023a 0.112a -0.022b 0.075a -0.021b 0.253a 0.058b 0.107a -0.025a 0.123a -0.019b 0.079a -0.022b
Panel C: Constant Conditional Correlation

𝝆𝑺,𝑮 0.064 0.044a 0.052a 0.080a 0.043a -0.010 0.049a 0.039a 0.038a
Panel D: Diagnostic Tests
LogL 29264.4 29827.4 29804.9 29972.4 32600.4 10718.6 29848.0 30087.9 30093.5
AIC -9.852 -9.801 -10.199 -10.179 -11.356 -12.232 -10.542 -10.390 -10.396
SIC -9.458 -9.407 -10.150 -9.987 -10.962 -12.062 -10.148 -9.965 -10.001
JB 1518.3a 5714.1a 705.2a 5606.6a 622.0a 5261.4a 1254.5a 5716.1a 1518.2a 4836.2a 694.9a 3658.2a 1167.2a 5400.0a 10399.8a 7411.6a 460.7a 5459.3a

𝑸(𝟏𝟐) 53.297a 7.353 18.165 6.823 9.706 6.922 10.290 6.457 15.808 7.237 53.183a 7.398 14.006 6.835 34.730a 6.269 25.856b 6.776
𝑸𝟐 (𝟏𝟐) 10.748 10.671 11.793 9.954 5.603 10.277 18.588c 9.578 20.236c 10.337 9.709 3.707 5.554 10.509 1.155 16.903 11.582 10.400
Notes: S and G indicate the stock and gold markets, respectively. LogL refers to the Log Likelihood. The number of lags for VAR is decided using AIC and SIC criteria. JB, Q (12) and Q2(12) refer to the empirical statistics
of Jarque -Bera test for normality, Ljung-Box Q statistics of order 12 for autocorrelation applied to the standardized residuals and squared standardized residuals, respectively. CHN, China; IND, India; INDO, Indonesia;
KOR, Korea; MYS, Malaysia; PAK, Pakistan; PHL, Philippine; TAIW, Taiwan; THA, Thailand. a,b,c indicate the statistical significance at 1%, 5% and 10% respectively.

13
 Table 5. Estimates of bivariate VAR-AGARCH for Gold and Emerging Asian stock markets during the global financial crisis

CHN GOLD IND GOLD KOR GOLD INDO GOLD MYS GOLD PAK GOLD PHL GOLD THA GOLD TAIW GOLD

Panel A: Mean Equation

Constant 0.000 0.001b 0.001a 0.001a 0.000 0.001 0.000 0.001a 0.000b 0.001a 0.002a 0.001c 0.000 0.001a 0.001c 0.001a 0.000 0.001b
𝑆
𝑟𝑡−1 0.068c 0.029 0.154a 0.038b 0.067c -0.030 0.144a 0.004 0.205a 0.039 0.085b 0.018 0.132a -0.004 0.116a 0.011 0.120a -0.037
𝐺
𝑟𝑡−1 -0.007 0.083b 0.005 0.077b -0.052 0.078b 0.106b 0.068b 0.043a 0.077a -0.051 0.091b 0.030 0.076c 0.046 0.091b 0.041 0.041
Panel B: variance Equation
Constant 0.000a 0.000 0.000a 0.000 0.000b 0.000c 0.000a 0.000 0.000b 0.000 0.000c 0.000 0.000a 0.000b 0.000 0.000 0.000 0.000
𝑆
(𝑒𝑡−1 )2
0.000 0.066c 0.094a 0.029 -0.022 0.055a -0.017 0.051 0.102a 0.000 0.050 0.050 0.075a 0.062c 0.046c -0.032a 0.013c 0.048a
𝐺
(𝑒𝑡−1 )2
-0.001 0.059a 0.004 0.071a 0.029 -0.037 -0.011c 0.052a -0.016c 0.068a -0.013 0.068a -0.002 0.078a 0.003 0.077a -0.033a -0.034a
𝑆
ℎ𝑡−1 0.854a -0.018 0.836a -0.024 0.956a -0.088b 0.493a 0.232a 0.834a 0.003 0.552a 0.123 0.665a -0.079b 0.849a 0.055a 1.000a -0.102
𝐺
ℎ𝑡−1 0.015 0.926a -0.001 0.925a 0.069 0.767a 0.031b 0.925a 0.023c 0.947a 0.023 0.953a -0.005 0.926a 0.000 0.938a 1.026a -0.375
Asymmetry 0.147a -0.013 0.120a -0.017 0.122a 0.172a 0.464a -0.017 0.088a -0.037a 0.352a 0.058b 0.166b -0.017 0.128a -0.033 0.060a 0.150b
Panel C: Constant Conditional Correlation
𝜌 𝑆,𝐺 0.112a 0.152a 0.093a 0.099a 0.122a 0.049 0.041 0.093a 0.114a
Panel D: Diagnostic Tests
LogL 4192.7 7362.5a 4398.5 4374.1 12036.4 3121.2 4376.9 4420.4 4378.2
AIC -10.153 -10.769a -10.629 -10.606 -12.553 -11.079 -10.799 -10.870 -10.832
SIC -10.027 -10.573a -10.336 -10.313 -12.403 -10.679 -10.505 -10.577 -10.538
JB 379.8a 642.1a 445.3a 42.9a 353.6a 683.0a 427.3a 637.3a 415.9a 101.3a 2047.0a 3314.7a 2689.6a 672.2a 342.4a 672.5a 357.4a 638.4a
𝑄(12) 12.903 5.456 17.395 12.223 12.143 7.154 11.272 6.130 13.302 10.274 40.220a 5.915 10.557 6.558 12.868 6.222 13.507 7.510
𝑄 2 (12) 10.369 15.108 8.179 5.259 8.306 16.515 6.718 15.086 8.344 5.680 3.182 7.554 25.364b 15.234 7.740 16.923 6.396 17.196
Notes: S and G indicate the stock and gold markets, respectively. LogL refers to the Log Likelihood. The number of lags for VAR is decided using AIC and SIC criteria. JB, Q (12) and Q2(12) refer to the empirical statistics
of Jarque -Bera test for normality, Ljung-Box Q statistics of order 12 for autocorrelation applied to the standardized residuals and squared standardized residuals, respectively. CHN, China; IND, India; INDO, Indonesia;
KOR, Korea; MYS, Malaysia; PAK, Pakistan; PHL, Philippine; TAIW, Taiwan; THA, Thailand. a,b,c indicate the statistical significance at 1%, 5% and 10% respectively.

14
 Table 6. Estimates of bivariate VAR-AGARCH for Gold and Emerging Asian stock markets during the Chinese Stock Market Crash

CHN GOLD IND GOLD KOR GOLD INDO GOLD MYS GOLD PAK GOLD PHL GOLD THA GOLD TAIW GOLD

Panel A: Mean Equation

Constant 0.000 0.000 0.000c 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
𝑆 0.065a 0.000 0.159 a 0.056c 0.027 -0.028 0.152 a 0.053b 0.153 a 0.079c 0.304a 0.033 0.056c 0.028 0.108 a 0.039 0.116 a 0.007
𝑟𝑡−1
𝐺 -0.038c 0.041 0.043c 0.035 -0.006 0.023 0.039 0.033 0.019 0.031 0.054b 0.037 0.017 0.057 0.080 a 0.022 0.021 0.029
𝑟𝑡−1
Panel B: variance Equation
Constant 0.000 0.000 a 0.000b 0.000 0.000 a 0.000 a 0.000 a 0.000 0.000 a 0.000 0.000 a 0.000 0.000c 0.000 0.000 0.000b 0.000 a 0.000
𝑆
(𝑒𝑡−1 )2 0.057 a -0.005 -0.023b 0.013 0.011 -0.002 0.102 a -0.010 0.072 a -0.007 a -0.019 0.000 0.020 0.009 -0.017 0.013 -0.031c 0.008
𝐺
(𝑒𝑡−1 )2 -0.007 a 0.034 a 0.002 0.000 0.045 a 0.007c 0.017b 0.010c 0.049 a -0.005 0.001 0.018b 0.007 0.003 0.007 0.010c 0.018a -0.005
𝑆 0.940 a 0.007 0.858 a 0.039c 0.928 a 0.001 0.776 a 0.046c 0.848 a 0.009b 0.842 a -0.019 0.870a -0.013 0.803 a 0.085b 0.846a 0.006
ℎ𝑡−1
𝐺 0.011 a 0.938 a 0.014 0.978 a -0.041b 0.963 a -0.013 0.977a -0.040c 0.991 a -0.002 0.981 a -0.002 0.983a 0.028 0.949 a -0.013 0.992 a
ℎ𝑡−1
Asymmetry 0.003 0.025b 0.166 a 0.015 a 0.080a 0.019b 0.096 a 0.010c 0.108a 0.021 a 0.258 a -0.003 0.128 a 0.019b 0.227 a -0.006 0.206a 0.017 a
Panel C: Constant Conditional Correlation
𝜌 𝑆,𝐺 0.043b -0.074 a 0.031 -0.003 -0.014 -0.044 0.001 -0.017 -0.041c
Panel D: Diagnostic Tests
LogL 11641.5 8888.8 05129.1 8795.3 3762.4 5466.4 3564.2 9046.6 9032.5
AIC -11.997 -12.983 -12.770 -12.823 -13.958 -13.072 -13.183 -12.962 -13.270
SIC -11.847 -12.786 -12.636 -12.626 -13.761 -12.778 -12.948 -12.765 -13.073
JB 546.8 a 3967.2 a 176.0 a 1235.0 a 185.7 a 3518.1 a 159.2 a 2998.8a 435.8 a 3584.9 a 416.7 a 3908.5 a 688.0 a 3970.1 a 516.4 a 3233.8a 540.5a 3830.7 a
𝑄(12) 19.851c 6.973 12.774 7.324 12.871 5.206 12.665 7.292 11.341 4.897 19.617c 4.757 19.194c 4.255 16.224 5.009 13.522 5.497
𝑄 2 (12) 17.069 3.532 6.119 6.228 13.008 3.765 8.116 3.748 18.025 5.456 17.627 4.944 14.026 5.175 12.277 3.323 12.157 7.173
Notes: S and G indicate the stock and gold markets respectively. LogL refers to the Log Likelihood. The number of lags for VAR is decided using AIC and SIC criteria. JB, Q (12) and Q 2(12) refer to the empirical statistics
of Jarque -Bera test for normality, Ljung-Box Q statistics of order 12 for autocorrelation applied to the standardized residuals and squared standardized residuals, respectively. CHN, China; IND, India; INDO, Indonesia;
KOR, Korea; MYS, Malaysia; PAK, Pakistan; PHL, Philippine; TAIW, Taiwan; THA, Thailand. a,b,c indicate the statistical significance at 1%, 5% and 10% respectively.

15
 Table 7. Optimal Weights and Hedge Ratios for Stock-Gold Portfolio
CHN/GOLD IND/GOLD KOR/GOLD INDO/GOLD MYS/GOLD PAK/GOLD PHL/GOLD THA/GOLD TAIW/GOLD
Full Sample Period
𝑤𝑡𝑆𝐺 0.38 0.43 0.44 0.44 0.68 0.45 0.43 0.46 0.46
𝛽𝑡𝑆𝐺 0.09 0.06 0.07 0.10 0.03 0.01 0.06 0.05 0.05
Global Financial Crisis
𝑤𝑡𝑆𝐺 0.31 0.39 0.45 0.43 0.72 0.46 0.49 0.38 0.42
𝛽𝑡𝑆𝐺 0.17 0.20 0.10 0.12 0.08 -0.02 0.03 0.07 0.14
Chinese Stock Market Crash
𝑤𝑡𝑆𝐺 0.45 0.52 0.61 0.49 0.73 0.50 0.42 0.58 0.57
𝛽𝑡𝑆𝐺 0.06 -0.07 -0.07 0.00 -0.01 -0.05 0.01 -0.01 -0.04
Note: 𝑤𝑡𝑆𝐺 and 𝛽𝑡𝑆𝐺 indicates the optimal weights and optimal hedge ratios, respectively.

16