VOLATILITY LINKAGES BETWEEN GOLD FUTURES IN EMERGING MARKETS
Hasan F. BAKLACI (Izmir University of Economics-TURKEY)
Ömür SÜER (Galatasaray University-TURKEY)
Tezer YELKENCI (Izmir University of Economics-TURKEY)
13th EBES Conference - Istanbul
WHY GOLD FUTURES IN EMERGING MARKETS?
The most ancient financial asset.
One of the most popular instrument during turmoil periods.
The introduction of gold derivatives has enhanced the importance of gold futures.
WHY GOLD FUTURES IN EMERGING MARKETS?
There is an upward trend in the trading volume of gold futures in emerging markets. In 2012, India’s derivative exchange had the highest silver
and gold future transactions in the world.
10 of the top 20 derivative exchanges are located in emerging markets.
Untapped area in finance literature.
OBJECTIVE
To examine the volatility linkages and transmissions among gold futures exchanges of four emerging markets:
Turkey, India, China, Taiwan.
WHY THESE COUNTRIES?
(I) Size of their gold markets
(II) The trading activity in their derivative markets
WHY THESE COUNTRIES?
(I) Size of their gold markets China, India and Turkey are the top three gold demanders in
the world (World Gold Council Report, November 2013).
Turkey is the second largest gold exporter in the world behind Italy.
Turkey and India are among the top three gold producers after Italy.
In the ranking of the top 40 countries based on their official gold holdings worldwide, China, Turkey, India and Taiwan are placed among top 15 countries.
WHY THESE COUNTRIES?
(II) The trading activity in their derivative markets
Top 30 derivatives exchanges list contained Indian, Chinese, Taiwanese and Turkish derivatives exchanges.
The trading volume in these countries’ derivatives exchanges corresponds to approximately one-fourth of total trading volume in top 30 derivatives exchanges globally.
The trading volume of gold futures traded in these markets represent more than 55% of total gold futures trading listed in the top 20 metals future contracts. (Annual Volume Survey 2012, published by Futures Industry Association).
RELATED STUDIES
Previous studies related with emerging gold futures markets have investigated predominantly the price discovery function of gold futures for spot market transactions [Pavabutr and Chaihetphon, 2010; Thenmozhi and Priya, 2011; Ho, Wang and Liou, 2010].
Volatility relationships are tested only among developed gold futures markets [Dhillon, Lasser and Watanabe, 1997; Xu and Fung, 2005; Lin, Chiang and Chen, 2008 ].
RELATED STUDIES Pavabutr and Chaihetphon (2010) : investigate the price
discovery process of the gold futures contracts in the Multi Commodity Exchange of India (MCX) for the period 2003 to 2007. The findings indicate that futures price leads spot price, which means that price discovery occurs in the futures market.
Thenmozhi and Priya (2011) : investigate whether gold, silver and crude oil futures markets serve as price discovery vehicle for spot market transactions, for the period 2005-2007, in top three developed and emerging commodity markets: Multi Commodity Exchange, India (MCX), New York Mercantile Exchange, U.S. (NYMEX) and Tokyo Commodity Exchange, Japan (TOCOM). The findings indicate that there is causality and information flow between futures and spot markets at MCX, COMEX and TOCOM for all the commodities and that gold, silver and crude oil futures and spot prices are cointegrated.
RELATED STUDIES Dhillon, Lasser and Watanabe (1997) : analyze volatility in US and
Japanese commodity futures markets (COMEX and TOCOM respectively) for the period from July 1987 to May 1992. Their findings demonstrate that more information seems to be released in gold markets during US trading hours compared to Japan. The information flow is estimated to be greater from COMEX to TOCOM.
Xu and Fung (2005) : analyze the patterns of information flows for gold, platinum, and silver futures contracts traded in US and Japanese markets, for the period 1994-2001 by using a bivariate asymmetric ARMA-GARCH model. Their results demonstrate a strong volatility transmission and feedback effects across both markets.
Lin, Chiang and Chen (2008) : investigate the dynamic relationships between two major gold futures markets, namely, COMEX and TOCOM by using bivariate GARCH model. Volatility spillover effects are observed in both COMEX and TOCOM before and during the uptrend.
DATA & METHODOLOGY
DATA:
Daily gold futures settlement prices
September 2, 2008 - December 20, 2012 (covers the crisis period)
1075 observations
DATA & METHODOLOGY
METHODOLOGY:
Multivariate GARCH
VECH - MGARCH Model
DATA & METHODOLOGY
DESCRIPTIVE STATISTICS:
Median values for all countries’ price series are close to the mean values indicating that the data are evenly distributed around the mean.
However, as revealed by Jarque-Berra statistics, the series do not fully conform to a normal distribution.
The kurtosis parameters point out that the price series conform to a platykurtic distribution with wider peak and shorter tail.
Table 1: Descriptive Statistics for Gold Future Price Series
Turkey Taiwan India China
Mean 69.86 4888.72 21022.53 280.84
Median 62.93 4914.51 19672 293.37
Maximum 107.75 6722.51 32636 394.09
Minimum 30.915 2846 11260 155
Standard Dev. 21.86 993.37 6115.05 58.72
Skewness 0.229 -0.123 0.352 -0.314
Kurtosis 1.576 1.846 1.742 1.816
Jarque-Bera 100.14 62.31 93.15 80.41
Probability 0 0 0 0
Observations 1075 1075 1075 1075
DATA & METHODOLOGY
DIAGNOSTIC TESTS:Unit root test results (ADF, PP, KPSS)
indicate that all the price series are non-stationary.
The series are first differenced to transform into return series.
The return series turn out to be stationary as documented in earlier studies.
DATA & METHODOLOGY
Conditional mean equation:
wherei : Turkey, India, China & Taiwanj : # of lags
DATA & METHODOLOGY
{ 𝝐𝒕=𝑯𝒕
𝟏𝟐𝜼𝒕 ,𝒘𝒉𝒆𝒓𝒆 𝑯𝒕 𝒊𝒔 𝒑𝒐𝒔𝒊𝒕𝒊𝒗𝒆𝒅𝒆𝒇𝒊𝒏𝒊𝒕𝒆 𝒔𝒖𝒄𝒉𝒕𝒉𝒂𝒕
𝒗𝒆𝒄𝒉 (𝑯𝒕 )=𝝎+∑𝒊=𝟏
𝒒
𝑨 (𝒊 )𝒗𝒆𝒄𝒉 (𝝐𝒕− 𝒊𝝐𝒕− 𝒊′ )+∑
𝒋=𝟏
𝒑
𝑩( 𝒋 )𝒗𝒆𝒄𝒉 (𝑯 𝒕− 𝒋)
𝒘𝒉𝒆𝒓𝒆𝝎𝒊𝒔 𝒂𝒗𝒆𝒄𝒕𝒐𝒓 𝒐𝒇 𝒔𝒊𝒛𝒆{𝒎 (𝒎+𝟏 )/𝟐}×𝟏 ,𝒂𝒏𝒅𝒕𝒉𝒆 𝑨 (𝒊 )𝒂𝒏𝒅 𝑩( 𝒋 )𝒎𝒂𝒕𝒓𝒊𝒄𝒆𝒔𝒐𝒇 𝒅𝒊𝒎𝒆𝒏𝒔𝒊𝒐𝒏𝒎(𝒎+𝟏)/𝟐×𝒎(𝒎+𝟏) /𝟐
Conditional variance equation:
RESULTS
VECH-MGARCH Estimates – Student-t Distribution
Coefficients
Conditional variance-covariance equation
rChina (i = 4) rIndia (i = 3) rTaiwan (i = 2) rTurkey (i = 1)
𝛼4𝑖
0.377166*
[2.633104]
𝛼3𝑖
0.122611
[0.664133]
0.327418*
[2.810226]
𝛼2𝑖
0.307862*
[2.416842]
0.268214*
[2.615038]
0.295523*
[2.685018]
𝛼1𝑖
0.100422
[0.435533]
0.187406*
[2.327549]
0.153628*
[1.900756]
0.270489*
[2.493751]
𝛽4𝑖
0.387527*
[9.110853]
𝛽3𝑖
0.684206
[1.265564]
0.835434*
[34.30973]
𝛽2𝑖
0.472267*
[2.195101]
0.796705*
[17.71359]
0.814408*
[32.49917]
𝛽1𝑖
0.559861
[0.735958]
0.872846*
[0.735958]
0.852483*
[0.735958]
0.876901*
[24.06252]
* denotes significance at 5% level. Numbers in square brackets correspond to t-statistics.
RESULTSThe results imply a volatility transmission and
spillover effect among the gold futures markets in the selected countries.
There is bi-directional short-run (ARCH effect) and long-run (GARCH effect) volatility transmission among the sample countries except China which has a siginificant volatility linkage with only Taiwan .
The MGARH estimates point out to a strong degree of volatility spillovar effect and volatility persistence.
RESULTSThe volatility interdependence observed
may stem from the existence of homogeneous investors to hedge against ascending inflation in sample countries during the period examined.
The risk diversification as well as cross-hedging opportunities among these countries gold futures is limited.
RESULTSWhy is China isolated?
Trading in Chinese gold futures market is available only for domestic investors.
China had banned foreign banks to import gold since January 2014 and has recently granted gold import licenses to foreign banks.
CONCLUDING REMARKS This paper is the first study to examine the volatility
transmissions and interdependencies among emerging gold futures markets.
Presence of bidirectional volatility spillover effect
among majority of the sample countries’ gold futures markets. (India,Turkey and Taiwan)
China is an exception which seems to be a relatively isolated market due to restrictions on gold trading particularly for foreign investors.
Limited risk diversification opportunities between the gold futures markets in majority of these countries.
CONCLUDING REMARKS
The policy making issue is particularly important for China, which does not yet allow foreign investors to engage in gold trading in futures markets.
Ascending gold trading volume coupled with the presence of tight cross-market volatility linkages among the sample countries point out the heightened role of emerging gold futures markets in global perspective.
Thanks for your patience