سرریز نوسانات روزانه بین نقطه و شاخص معاملات آتی: شواهدی از بازار سهام کره ای

This study provides empirical evidence of the relationship between spot and futures markets in Korea. In particular, the study focuses on the volatility spillover relationship between spot and futures markets by using three high-frequency (10 min, 30 min, and 1 h time-scales) intraday data sets of KOSPI 200 spot and futures contracts. The results indicate a strong bi-directional causal relationship between futures and spot markets, suggesting that return volatility in the spot market can influence that in the futures market and vice versa. Thus, the results indicate that new information is reflected in futures and spot markets simultaneously. This bi-directional causal relationship provides market participants with important guidance on understanding the intraday information transmission between the two markets. Thus, on a given trading day, there may be sudden and sharp increases or decreases in return volatility in the Korean stock market as a result of positive feedback and synchronization of spot and futures markets.

Since the introduction of futures markets, a large number of studies have examined the effects of futures trading on underlying spot prices, including the lead–lag relationship between spot and futures markets [1], [2], [3] and [4]. The lead–lag relationship between these two markers indicates how much they are closely related each other and how fast one market reflects new information from the other. According to the efficient market hypothesis, any new information is quickly reflected in the underlying spot market and its futures market simultaneously, so that investors cannot make any profits using currently available information in these markets [5] and [6].1 In reality, however, information can be disseminated in one market first and then transmitted to the other later as a result of market frictions such as transactions costs or market microstructure effects [13]. Some recent studies have suggested that futures markets play an important price discovery role for spot markets because of low transaction costs, the ready availability of short positions, low margins, and rapid execution [14]. Thus, futures prices may contain useful lead information on subsequent spot prices, including information not yet embedded in current spot prices. In terms of the returns of these two markets, some studies have also provided similar results, indicating that, although futures returns have a bi-directional relationship with spot market returns, futures markets tends to lead the underlying spot markets, because the relationship from futures markets to spot markets is stronger [4], [6], [15], [16], [17], [18] and [19]. On the other hand, more recent studies have focused on the volatility interaction between spot and futures markets and provided evidence of strong cross-market dependence in the volatility process. Understanding volatility spillovers is important for predicting future volatility in both markets. There are three streams of research on volatility spillover effects. The first stream proposes bi-directional volatility spillovers between the two markets [3], [20], [21], [22], [23] and [24]. The second stream suggests unidirectional volatility spillovers from futures to spot markets, suggesting that the arrival of new information disseminates faster in derivatives markets [19], [21], [25] and [26].2 The final stream proposes no volatility spillover effects [28] and [29]. In this paper, we focus on the issue of volatility spillovers between the KOSPI 200 spot market and its futures market by using bivariate GARCH models, in order to provide an important insight on the mechanism of information transmission between the two markets. It is well known that the two indices related to stock markets exhibit fat tails in the distributions of their returns. In the literature of econophysics, the non-Gaussianity is often described as αα-Levy stable distributions, given that the fat-tail distributions have power-law decay. Mantegna and Stanley [30] and Podobnik et al. [31] suggested the use of ARCH–GARCH type models to describe the power-law stability in the distributions of the returns in these kinds of the variables. Our major concern in this study is to examine the lead–lag relationships between spot and futures indices, which exhibit power-law tails in the distributions of their returns. To investigate the correlations of the return volatilities with this characteristic, we use bivariate GARCH models, which are useful in simultaneously exploiting the possible linkages of volatility spillovers between the two non-Gaussian indices, rather than separate univariate models [32]. In this study, we examine the intraday volatility spillover effect by using three intraday data sets (10 min, 30 min, and 1 h intervals). Although many empirical studies have documented the volatility spillover effect in these markets, few have examined it for the cases of intraday high frequencies data of the Korean stock market, which has one of the largest futures derivatives markets in the world. The reason for using different time scales is because we are interested in measuring a fractal structure with self-similarity [33]. Although many econophysics studies analyzed the high frequency data of financial markets [34], [35] and [36], they mainly focused on a single market rather than cross-market relationships. Daly [37] introduces cross-market issues and techniques in the econophysics field. In particular, the interaction relationship of econophysics variables can be examined by a multivariate context. Lee, Chiu and Lee [4] investigated the lead–lag price jump relationship between spot and futures markets using a bivariate Granger-causality test. Chiang, Yu and Wu [38] employed a multivariate model, dynamic conditional correlation (DCC), in estimating the difference intraday time scale relationship between the US spot and futures markets. We also investigate the market opening effect of overnight information flow on intraday returns and consider the volatility spillover effect by using the bivariate GARCH model. This investigation of the intraday volatility spillover effect provides arbitrageurs, hedgers, and speculators with a better understanding of short-term dynamics of return volatility. The rest of this paper is organized as follows. Section 2 provides the descriptive statistics of 10 min high-frequency data. Section 3 discusses the econometric methodology used in this study. Section 4 provides the results, and several conclusions are discussed in Section 5.

This empirical research has investigated volatility spillovers between KOSPI 200 spot and futures contracts, using a bivariate GARCH–BEKK model. In the study, we used three high-frequency intraday data sets (10 min, 30 min, and 1 h intervals). This investigation of intraday volatility spillover effects would provide arbitrageurs, hedgers, and speculators in KOSPI 200 spot and futures markets with a deep understanding of short-term dynamics in both markets. Our study shows that, while past information shocks in futures markets influence current volatility in spot markets and vice versa, there is also evidence for the existence of bi-directional volatility spillovers between futures and spot markets, suggesting that new information is reflected in spot and futures markets simultaneously. These shock and volatility transmissions between KOSPI 200 futures and spot markets seem to indicate a strong bi-directional causality, suggesting that market volatility in the spot market can influence that in the futures market and vice versa. This bi-directional causal relationship between the two markets may provide important guidance on understanding intraday information transmission to market participants. For example, on a given trading day, there may be sudden and sharp increases or decreases in return volatility in the stock market as a result of positive feedback from KOSPI 200 futures and spot markets. In other words, self-organized criticality can emerge unexpectedly in intraday trading in the case of Korean stock markets. In the sense of physics, one implication of volatility spillovers we found in this study is that spot and future markets are integrated and synchronized. By regarding the two markets as individual network nodes, respectively, the synchronization can be understood in the line of network theory. According to Newman [46] and [47], synchronization is a network phenomenon where individual nodes tend to exhibit similar behavior. In this context, the interrelationships between nodes can be classified into two types of patterns. One is assortative mixing, with which nodes tend to connect tighter to other nodes with similar network properties. The other is disassortative mixing, with which each node has a tendency to evolve, unless otherwise constrained, towards its maximum entropy state. The latter can be often found in technological or biological networks [48]. In the sense of this network theory in physics, the volatility spillovers we have uncovered in this study could be regarded as the result of assortative mixing between spot and futures markets.