Hiệu ứng lan truyền quốc tế: Thực chứng từ Asean 5

Nội dung text: Hiệu ứng lan truyền quốc tế: Thực chứng từ Asean 5

1. INTERNATIONAL TRANSMISSION EFFECTS: EVIDENCE FROM ASEAN 5 HIỆU ỨNG LAN TRUYỀN QUỐC TẾ: THỰC CHỨNG TỪ ASEAN 5 Hsin-Fu Chen Lunghwa University of Science and Technology Shu-Mei Chiang Lunghwa University of Science and Technology, Taiwan Abstract This paper applies the ARJI model to proceed empirical analysis on the influence of financial market owing to the panic. In addition to examining the variation in the stock index caused by the financial crisis from the US in ASEAN FIVE to check if panic could infect, we could also further explore if the panic among ASEAN FIVE would result in varying influences and regional spillover effects. The empirical results show that the investors fear gauge of the US can affect the stock market of Malaysia and the Philippines, not those of Indonesia, Thailand and Singapore; that is, there is partial international spillover effect between the stock markets of US and ASEAN FIVE. Besides, there are also unidirectional or bidirectional influence effect which indicates that there existing not only strong spillover among the stock indices of ASEAN FIVE, but also highly regional effects among them. Although the returns are higher in the stock markets of Indonesia and Philippines, their risks are also lower. On the contrary, the risks are higher and the returns are lower in Thailand and Singapore, which are relatively worse market for investment. Further, the total variance is mainly influenced by the GARCH variance in the stock indices of Indonesia, Thailand, Malaysia, Philippines and Singapore. Finally, whenever there are important events occurring in the markets, the total variance, GARCH variance and jump variance will increase. Keywords： ARJI Model, VIX, ASEAN, Spillover Effects JEL Classification：C4, G1, G15,G34 Tóm tắt Bài báo áp dụng mô hình ARJI để thực hiện phân tích thực nghiệm về tác động của thị trường tài chính do tâm lý lo sợ. Bên cạnh việc xem xét các biến động của chỉ số chứng khoán gây ra do khủng hoảng tài chính của Mỹ và 5 nước ASEAN (ASEAN 5) để kiểm tra xem liệu tâm lý lo sợ có tác động không, chúng tôi cũng tìm hiểu thêm xem liệu tâm lý lo sợ trong ASEAN 5 này có gây ra các ảnh hưởng khác nhau và có hiệu ứng lan truyền trong cả khu vực hay không. Nghiên cứu thực nghiệm chỉ ra rằng việc các nhà đầu tư e ngại về thị trường Mỹ có tác động tới thị trường chứng khoán của Malaysia và Phillipines nhưng không ảnh hưởng đến thị trường chứng khoán của Indonesia, Thailand và Singapore, có nghĩa là có tác động lan truyền một phần giữa thị trường chứng khoán Mỹ và thị trường chứng khoán của ASEAN 5. Bên cạnh đó, cũng có tác động đơn hướng và nhị hướng, cho thấy rằng không chỉ tồn tại hiệu ứng lan truyền mạnh giữa chỉ số chứng khoán của ASEAN 5 mà còn có tác động khu vực mạnh giữa các chỉ số này. Mặc dù kết quả hoạt động của thị trường chứng khoán Indonesia và Philippines cao hơn nhưng mức độ rủi ro của hai thị 201
2. trường này lại thấp hơn. Ngược lại, mức độ rủi ro cao hơn và kết quả hoạt động thấp hơn ở thị trường Thái lan và Singapore, đây cũng là hai thị trường tương đối không hiệu quả cho đầu tư. Ngoài ra, tổng dao động chịu ảnh hưởng của dao động GARCH trong các chỉ số chứng khoán của Indonesia, Thailand, Malaysia, Philippines và Singapore. Cuối cùng, mỗi khi có sự kiện quan trọng gì xảy ra trên thị trường, tổng dao động, dao động GARCh và dao động nhảy sẽ tăng lên. Từ khoá: Mô hình ARJI, VIX, ASEAN, hiệu ứng lan truyền 1. Introduction The US economy is the financial center of the world, so-called the world economic “locomotive”, all the events happened in the United States have an impact on countries, for instance: the dot-com bubble in 2000, the 911 attack in 2001, the subprime mortgage crisis in 2007-2008, to the recent European debt crisis, the financial volatilities generated by the US quantitative easing (QE1, QE2, QE3) monetary policy, the fiscal cliff and other important events, have a dramatic impact on the globally political, economic and financial markets, almost no country is immune. Especially, such as the technology innovation and internet development in recent years, any related events occurred in the US and their effects can spread very quickly to every corner of the world. As far as the developed countries are concerned, the emerging markets “ASEAN Five” especially suffer investors to favor. Referred to the Association of Southeast Asian Nations as ASEAN, a group consisting of five countries: Indonesia, Thailand, Malaysia, the Philippines and Singapore in order to prevent the spread of communism and promote the regionally economic and trade exchanges and cooperation thereof. According to the Declaration of ASEAN (ASEAN Declaration) 1967 signed by these five countries, its aims and objectives are to accelerate this region’s economic growth, social progress and cultural development, while respecting each country’s law regulations and sticking to the principles of Charter of the United Nations, to promote regional peace and stability. The ASEAN endowed with affluent resources has been gradually conspicuous emerging internationally in recent years. In order to balance the trend of China’s rising, ASEAN has become the United States eagerly to entice, the key to “return to Asia”, and therefore ASEAN takes Chinese orders on the left-hand way and accepts the United States’ court on the right-hand way, benefiting from both ways politically. Economically, ASEAN also jumps quickly, not only actively negotiates with FTA externally, but also targets to establish the “ASEAN Economic Community” (AEC) in 2015, ascending to the global second largest single market, second to EU. Economic and Cooperation Development (OECD) predicts the 2012-2016 annual GDP growth rate of ASEAN countries expected to reach 5.6%, higher than EU and NAFTA. Therefore, with the slowdown in economic growth of the Asian Four Tigers, ASEAN attracts the international attention on the political and strategic status, as well as economically has gradually risen to become the potential stocks the investors being eager to comprehend. Moreover, ASEAN FIVE countries have become the axis of Asian economic growth. 202
3. Since the ASEAN FIVE countries not only possess with high degree of similarity in the political, economic and national development, but also continue to fast grow on the financial and economic side, thus being the investors’ good investment targets. In the past, these five countries attract substantial foreign capital inflows through the factors of the market liberalization, the related reconstructing systems of stock market, and the deregulation and international integration etc., in addition to the rapid growth of these five countries stock markets, and therefore the importance of the of ASEAN FIVE countries to the world economy is second to none. Nonetheless, whether the global financial crisis will affect the ASEAN FIVE stock markets and furthermore will affect each other between these ASEAN FIVE stock markets? Fig. 1.1 US VIX Index Trend, 2005.1.2 to 2014.12.31 Fig. 1.2 Indonesia Stock Index Trend, 2005.1.2 to 2014.12.31 Fig. 1.3 Thailand Stock Index Trend, 2005.1.2 to 2014.12.31 203
4. Fig. 1.4 Malaysia Stock Index Trend, 2005.1.2 to 2014.12.31 Fig. 1.5 Philippine Stock Index Trend, 2005.1.2 to 2014.12.31 Fig. 1.6 Singapore Stock Index Trend, 2005.1.2 to 2014.12.31 From Fig. 1.1 to Fig. 1.6 presented here, we find that: When the global information transmits far-reaching, thus an important event occurring will very rapidly spread around the world. Moreover, the US, being the world financial center, its economically dynamic impact on the countries is self-evident, the spillover effects produced seem to become an inevitable trend. VIX is a trademarked ticker symbol for the CBOE Volatility Index, a popular measure of the implied volatility of S&P 500 index options; the VIX is calculated by the Chicago Board Options Exchange (CBOE). Often referred to as the fear index or the fear gauge, the VIX represents one measure of the market's expectation of stock market volatility over the next 30-day period. In this way, we use the ASEAN FIVE viewpoint to 204
5. investigate when the US market investors’ psychological panic, what the spillover effect of this panic is; besides, we want to inquire whether the ASEAN FIVE stock markets will be affected by the US panic psychology. Furthermore, due to the ASEAN FIVE high degree of similarity, we would analyze the correlation between the ASEAN FIVE stock markets, especially encountering the major events impact, whether the changes will be the same between the ASEAN FIVE stock markets? Spillover effects1 are defined as externalities of economic activity or processes that are not directly involved in it. Therefore, when the financial market of some country has the crisis; it will usually result in some degree of influence in the financial market of other countries. Especially those countries which have highly economic growth, such as the countries whose economic activities may be closely related to the crisis country or those markets which have gained the attention from the world, will face greater shock. Given this relevance, when financial crisis happens, investors with risk aversion will promptly transfer their funds to other safe haven (e.g. gold or precious metal markets) or reserve cash position. Therefore, under the shocks from exogenous events, crisis country will more or less spills-over the related countries. During the past decade, international stock markets have experienced sustained- growing interaction with one another. Volatility and return spillover effects have been concurred across national stock markets owing to economic integration and the development of stock markets. As emerging markets are also one part of global finance, they could also be influenced by the positive/negative systematic risk. The sub-prime mortgage crisis, which originated in the dominant US market, ultimately gave rise to a global financial meltdown throughout the 2007-2008 period, a period which has also had enormous impacts on the ASEAN FIVE economies; unlike many of the other recent crises, such as the Asian financial crisis (1997), the Russian economic crisis (1998) and the Brazilian real crisis (1999), it is clear that the current crisis stemmed directly from the US market. Since the US market continues to represent the world’s most influential economic system, the sub-prime mortgage crisis has triggered a worldwide scare and spillover effects for both emerging and developed markets (Cheung et al., 2010). With the economies of the developed world suddenly finding themselves faced with raging recession, growth in the ASEAN FIVE was also set to experience a significant slowdown; this has inevitably resulted in severe falls in the ASEAN FIVE stock markets. As a result, amid the devastation of the US sub-prime mortgage crisis, the hopes and expectations of many investors with particular focus on the emerging markets were quickly undermined. Given that most of the emerging markets are small-scale, with very little depth, prices can tend to be extremely volatile; for example, the return distribution of the financial assets is generally found to be high-peaked and fat-tailed, and although their rates of return 1 Contagion refers to a scenario in which the economies have highly relevance without the US subprime crisis and have causal/ simultaneous relationship. Thus, if there is contagion effect between two countries, then reciprocal effect would exist in the stock market. Consequently, spillover effect is not the same as contagion effect. 205
6. tend to be higher, they are also often accompanied by greater hazards. When new information flows into the financial markets, asset prices react to such news, ultimately leading to changes in the expectations of investors; however, various news events, such as financial crises, macroeconomic declarations, market crashes and political news, can be the source of jump innovations on returns. Kim and Mei (2001) suggest that an important impact of any unexpected event will be the occurrence of discontinuous jumps. In addition to the spillover effect from the US stock markets, given that the ability to digest jump innovations is of considerable importance to the financial markets, an issue of particular interest to the present study centers on which of the BRICV markets have the lowest risks and greatest information efficiency. As a result of such considerations, from a perspective of asset allocation, the question also arises as to which of these countries should receive the greatest capital allocation when investors are considering BRICV stock markets as their investment objectives. Although it is quite clear that there is already a significant wealth of research available on mean and volatility spillover effects in the financial markets2, there are relatively few examples of any examination of the contagious effects of the sub-prime mortgage crisis on the stock markets of the BRICVs; furthermore, despite Cheung et al. (2010) having provided some evidence relating to the impact of the financial crisis on the relationships between the US and other countries, their analysis was confined to the informational role of TED spread. However, in addition to the spillover effects, since price fluctuations are easily identifiable in the emerging markets, it is also important to understand and compare the risks and efficiency of information transmission within the BRICV markets. This study therefore sets out to examine the potential infection of the US stock markets on those of the ASEAN FIVE s originating from the sub-prime mortgage crisis, along with the related risk and information efficiency levels among the five countries. Accordingly, we adopt the ARJI model (Chan and Maheu, 2002) to explore the stochastic return and volatility processes within the ASEAN FIVE markets, with consideration of the sub-prime mortgage crisis also being included in the model construction. Such analysis may help to provide investors with a better understanding of the inherent risks and efficiency of these stock markets, and also assist them in their selection of the most effective asset allocation. While our results should assist the governments of the ASEAN FIVE economies, with regard to the adoption of the most appropriate supervisory strategies, it is further anticipated that such a model - with the inclusion of spillover effects - should also help to capture the impact of mean and volatility spillover effects in the ASEAN FIVE stock markets following the outbreak of the sub-prime mortgage crisis, as well as the associated responses of each of these markets to news events. 2 See for example, Liu and Pan (1997), Ng (2000), Christiansen (2007), Vrugt (2009), Bhar and Nikolova (2009), Bowman, Chan, and Comer (2010) and Cheung et al. (2010). 206
7. The remainder of this paper is organized as follows. Descriptions of the related data and the empirical methodology adopted for this study are provided in Section 2, followed in the penultimate section by presentation of our empirical results. Finally, the conclusions drawn from this study are summarized in the closing section. 2. Data and methodology 2.1. Data Our analysis in this study is based upon the daily indices of the US Volatility Index (VIX) and the ASEAN FIVE economies of Indonesia, Thailand, Malaysia, the Philippines and Singapore; the data used for this analysis were obtained from Taiwan Economic Journal (TEJ). The sample period, which begins from the date, 1 January 2005 of each of the stock markets, runs until 31 December 2014; this ultimately provides a ten-year period, with all of the subsequent analyses being conducted on the return data. Although there are overlaps between the trading hours of the ASEAN FIVE markets, the US market is closed when the non-US markets are operating; thus, the ASEAN FIVE markets, which start to open three to four hours after the US market has closed, can be significantly influenced by any information flows coming out of the US market. In order to accommodate such a condition, we use lagged returns and volatility data to examine the spillover effects of the US sub-prime mortgage crisis. 2.2. Methodology The primary aim of the present study is to investigate and compare the impacts of the sub-prime mortgage crisis on risk and efficiency levels within the BRICV stock markets. We follow Chan and Maheu (2002) and Maheu and McCurdy (2004) to apply the ARJI model for our examination of the related influences; the ARJI models postulate thatjump intensity obeys an ARMA process while also incorporating the GARCH effect of the return series. Furthermore, in order to explore the effects of the sub-prime mortgage crisis, we include one dummy variable within both the mean equations and the variance equation in an attempt to capture the potential contagious effects of the event. The ARJI model can be expressed as: p q nt Ri,t = µ + ∑φi,Ri,t −i +ψ l At−1 + ∑ς jO j,t,i≠ j + ∑π t,k (1) i=1 j=1 k =1 q p 2 ht = ω + ∑α iε t−i + ∑ βiht−i (2) i=1 i=1 r s λt = λ0 + ∑∑ρiλt−i + γ iξt−i (3) i=11i= Z~(0,1) NID , , π ~ N(θ ,δ2 ), n~ Poissonλ dt t t,k t t tt( ) where Ri,t and O j,t are the return series of the stock indices of Indonesia, Thailand, Malaysia, the Philippines and Singapore; At is the changing rate of US Volatility Index (VIX), Zt ~ NID(0,1) is the standardized Winner Process. We assume the jump size (π t ,k ) 207
8. and standardized Winner Process ( Zt ) is mutually independent, where the jump size is 2 assumed to be independent and normally distributed with mean θt and variance δ t . ht denotes the conditional volatility dynamics of the return R which follows a GARCH (p, p q ntt q) process; εµφt =−−RRtit∑ −i -ψ l A t−1 + ∑ς jOj,t,t≠ j + ∑π t,k is the mean equation residual. i=1 j=1 k=1 nt is the discrete counting process governing the number of jumps arriving between t −1 and t , which is distributed as a Poisson random variable. Chan and Maheu (2002) define λttt≡ΩEn[]| −1 is the conditional expectation of the counting process, which is rs assumed to follow an ARMA(,) r s process, and denote λt =+λρλγξ0 ∑∑iti−− + iti , where ii==11 λ 0 > 0, ρi ≥ γ i , γi ≥ 0. ξti− is the jump intensity residual, then defined as: ∞ ξt−−−−itititititit≡Φ−==Φ−En[| ]λλ∑ jPn ( − j | −− ) i i=0 En[|ti−−−Ω ti1 ]is the expected number of jumps as measured ex post between ti−−1 and ti− , λti− is the conditional expected number of jumps conditional on the information set Ωt−i−1 . As far as the information set Ωt−i−1 is known, ξti− is the unpredictable impact factor inferring on the conditional mean, and follows the Martingale difference sequence, EEEn[|ξti−−−Ω ti11 ][[| = ti −− ΩΩ ti ]|] ti −−− −=−=λλλ ti ti −− ti 0, therefore, the property of the jump intensity residual is serially uncorrelated. From Eq. (3)， the unconditional jump intensity is equal to: λ0 E[]λt = r 1− ∑ ρi i=1 Given the Poisson Distribution, λt must be positive value. In the case of rs= , the ARJI model can be re-expressed as follows: rr λλt =+0 ∑∑() ργλiiti −− + γ ititiEn [|]−− Ω ii==11 Where in order for λt > 0 for all t , the sufficient conditions are: λ0 > 0, ρi ≥ γ i , γ i ≥ 0 . Let fR(|tt n= j ,Ω t−1 ) denote the probability density function of returns given that j jumps occur and the information set Ωt−1 , Chan and Maheu (2002) infer ex post the probability of the occurrence of j at time t based on time t −1 information: fR(|tt n=Ω j , t−−11 ).( Pn t = j | Ω t ) Pn(tt=Ω= j | ) , j = 0,1,2,3, fR(|ttΩ −1 ) Where Pn(|)tt=Ω j −1 is so-called filter, this filter is the predicted value of probability for nt at time t −1. Therefore, the conditional density of returns is: 208
9. ∞ PR(|ttΩ=−−−111 )∑ f (| R tt n =Ω j , t ).( Pn t =Ω j | t ) j=1 In this way, the normal conditional probability density function of returns conditional on j jumps occurring at unit time interval is as follows: p ⎛⎞2 ()Rtiti−−µϕRj− − θ −1 ⎜⎟∑ 2 2 ⎜⎟i=1 fR(|tt n=Ω j , t−1 ;)(2( Ψ=πδ h t + j ))exp − 2 ⎜⎟2(hjt + δ ) ⎜⎟ ⎝⎠ Based on the above specification, the log-likelihood function can be constructed as: T L(Ψ) = ∑log f (Rt | Ωt−1;Ψ) t=1 Where Ψ = (µ,φi ,ψ l ,ς j ,ω,α i , β i ,θ ,δ ,λ0 , ρ i ,γ i ) are the parameters vector to be estimated. The estimation in above equation involves an infinite summation. In order to make our estimation feasible, the maximum number of jumps ( nt ) takes the large value of 5 as the truncation point for the distribution determining the number of jumps. We set the number of jumps at 5 essentially because, in practice, the conditional Poisson distribution constructed above has zero probability in the tail for values of nt > 5. 3. Estimation results The descriptive statistics of the ASEAN FIVE stock indices are presented in Table 1. The results reveal that the Indonesia (0.0807%) is the highest among the ASEAN FIVE, the Philippines (0.0664%) is second, but the Singapore (0.0230%) is the smallest. As far as the risk (the volatility of return) is concerned, the Malaysia (0.8662) is the smallest, the Singapore (1.3209) is second, the Indonesia (1.6782) is the highest. Thus, among the five stock markets, the returns of Indonesia and the Philippines are higher, but risks are relatively higher; although the risk of Singapore stock is lower, the return of this area is the lowest, not seeming a good investment area. As the result, the ASEAN FIVE belong to the same economic community, whereas owing to the market characteristics and the different technology and industry levels in different countries, so that there are great differences in the returns and risks of five countries stock markets. In addition, as showed in the Table 1 results, the return distributions of Thailand and Malaysia stock indices are left-skewed, the return distributions of Indonesia, the Philippines and Singapore are right-skewed; Moreover, the return distributions of these five countries stock indices are non-normal and fat-tailed and high-peaked than normal distribution, consistent with the GARCH effect. The ADF and PP unit-root testing for the US VIX and the ASEAN FIVE stock indexes suggests evidence of the possible presence of a unit root, provided on request. Table 1 Descriptive statistics of returns for the ASEAN FIVE stock indices. 209
10. Variables Mean S.D. Maximum Minimum Skewness Kurtosis Jarque-Bera Indonesia 0.0807 1.6782 10.3200 -18.4970 16.1425 16.2711 22685.3517 Thailand 0.0373 1.5032 10.5770 -17.8484 -1.6457 21.2211 39410.8176 Malaysia 0.0321 0.8662 5.7160 -10.2370 -1.0533 16.1425 22647.9987 Philippines 0.0664 1.5140 8.8330 -15.7890 1.9866 -0.8875 8257.6255 Singapore 0.0230 1.3209 10.3440 -8.6960 0.0792 8.5976 6128.0538 ＊＊＊Indicates significance at the 1% level. ＊＊Indicates significance at the 5% level. 3.1. Spillover Effects As shown in Table 2, the part A shows the empirical results of the ARJI model parameters. We find that as far as the five countries stock markets of Indonesia, Thailand, Malaysia, the Philippines and Singapore are concerned, all of the parameters describing the dynamics are almost significant. In terms of model goodness-of-fit, the ARJI model of these five countries also passes the Ljung-Box Q(10) test and Q2(10) test, thereby indicating that the ARJI model appears to perform relatively well with regard to describing behavior within each of stock markets of Indonesia, Thailand, Malaysia, the Philippines and Singapore. 3.2. Information Efficiency First, the empirical results as shown in Table 2, within the mean equation of Indonesia, Thailand, Malaysia, the Philippines and Singapore, we find that the significance of ψ 1 indicates that the US stock market ups and downs does have significant negative impact on the stock markets of Malaysia and the Philippines, showing that when the US investors’ panic extent increases, the stock prices of Malaysia and the Philippines will fall; but the impact of changes in the US stock market on all of the three countries stock markets of Indonesia, Thailand and Singapore is not significant. In additions, the statistical significance of ς1 、ς 2 、ς 3 、ς 4 andς 5 indicates that there is indeed some degree of spillover effects among the five countries stock markets of Indonesia, Thailand, Malaysia, the Philippines and Singapore. For Indonesia, the stock index returns of Thailand, Malaysia, the Philippines and Singapore on its stock index return shows significant and positive impact, means any one of the four countries stock price ups (downs) will simultaneously cause this country’s stock price ups (downs), where the impact of Singapore stock market on Indonesian stock market is greatest; for Thailand, the stock index of Indonesia, Malaysia, the Philippines and Singapore on its stock index also shows significant and positive impact, means there is a comovement relationship between these four countries stock market and Thailand stock market, when any one of the countries stock price ups (downs), it will also cause this country’s stock price ups (downs), where the impact of Singapore stock market on Thailand stock market is greatest; For Malaysia, only the three countries stock indices of Indonesia, the Philippines and Singapore on its stock price shows significant and positive impact, but Thailand’s stock price has no effect on the Philippines’ s. For the Philippines, the stock index of Indonesia, Malaysia and Singapore on its market shows significant and positive impact, where Thailand’s stock price has no effect on Malaysia’ s. Whereas for Singapore, all of the stock indices of 210
11. Indonesia, Thailand, Malaysia, the Philippines on its stock index shows significant and positive impact. Overall, due to changes in the US VIX will only affect Malaysia and the Philippines stock markets; therefore, the international spillover effect to some extent exists in this region; there are two-way causality between Indonesia Thailand, Indonesia Malaysia, Indonesia the Philippines, Indonesia Singapore, Thailand Singapore, Malaysia the Philippines, Singapore Malaysia, Singapore the Philippines. While there is only one-way causality between Malaysia → Thailand, the Philippines → Thailand, but the regional spillovers do exist among the ASEAN FIVE countries stock markets. 3.3. Market risk In terms of variance equation, the empirical results show that: owing to λ0 , ρ1 andγ 1 are almost significantly different from zero, not only showing that the jump processes are certainly time-varying (ex. Fig. 3.1 to Fig. 3.6 presented here), but also for the importance of jumping provides the proof, rejecting the assumption of jump being fixed proposed by Chen and Shen (2004), consistent with the result of Chiang, Chung and Huang (2012). We have further found that the unconditional jump intensities of Indonesia、Thailand、Malaysia、the Philippines and Singapore are separate as 0.0838、0.0227、0.5378、0.3056 and 0.3696, representing on average, the returns of Indonesia、Thailand、Malaysia、the Philippines and Singapore occur jumping once every 1.09、1.02、2.16、1.44 and 1.59 business days93. In this way, though the jump occurring frequencies of these five countries stock markets are high, they all will completely disappear every one to two days. Part B in Table 2 shows that the total variances of Indonesia、Thailand、Malaysia、the Philippines and Singapore stock indices are affected by the GARCH variance, whereas the proportion of jump variation is not small; indicates variability caused by the jump process cannot be ignored, therefore, when studying the characteristics of the five countries stock markets, the jump variability is a very important factor. As Fig. 3.1 to Fig. 3.6 represent, we have found that: when major events occur in each country, the total volatility, GARCH volatility and jump volatility of stock markets will enhance. 3.4. Event analysis This paper collected seven stake events during the whole sample of 2005-2014, and the jump intensities and total volatilities results reported in Table 3, for the following seven related events examined in this study: (1) the Southern Philippines 6.5 Earthquake Occurred in 2007; (2) the 12th General Election of Malaysia in 2008; (3) the Thailand Bloodshed Demonstrations in 2008; 93 λ0 Formula of jump intensity = 1− ρ1 211
12. (4) the Indonesian Equities Indefinite Cease Trading when Asian Stock Crash in 2008; (5) the Subprime Mortgage Crisis in 2008; (6) the Easing Monetary Policy Announced by the Monetary Authority of Singapore in April to May 2009; (7) the US QE Exit in 2013. As shown in Fig. 3.6 - Fig. 3.10, depending on the events occurring global and in every country, we can find: the jump intensity and total volatility generated during the events occurring are almost all higher than the average of the whole sample period, as compared to periods of no important news. The unlisted data (-) owing to its impact being small are almost negligible. It can be seen: the volatility of the stock index is indeed affected by the incident, the only difference is affecting the length of time and the impact of the volatility. Therefore, when an event occurs, the stock index would indeed appear abnormally. These results indicate that when investigating the impact of events, consistent with the results of Chiu et al. (2006) and Chiang et al. (2009), it is necessary to consider jump intensity in order to avoid incorrect decision making. 4. Conclusions Because the US is the focus of global financial markets, and the implementation of various policies and the occurrence of an event (such as: the subprime mortgage crisis, etc.) are violent waves of global stock market turmoil, not only the US stock market downturn, leading to the US economic downturn, rising unemployment, but also the global economy having a negative knock-on effect. Therefore, the US as a global financial center, the impact of its economic dynamics on every country’s economy is self-evident, the spillover effect generated has become an inevitable phenomenon. When the panic psychology originating from the US market investors, is it contagious? Furthermore, whether will the panic psychology generate the spillover effects arising from the geographic close or not? We adopt the ARJI model to examine the spillover effects between the US market and the ASEAN FIVE countries stock indices, and further to explore when a major event occurs, each volatility and jump of individual stock index, so as to provide the investors to invest and hedge for reference. The empirical results show: (1) American investors’ panic psychology will affect the stock markets of Malaysia and the Philippines, but it won’t affect the ones of Indonesia, Thailand and Singapore, indicating that there exists a partial international spillover effect between the US and the ASEAN FIVE countries stock markets; (2) There also exists one-way impact among the stock indices of Indonesia, Thailand, Malaysia, the Philippines and Singapore (Malaysia and the Philippines will affect Thailand) or two-way impact, showing that the spillover effects among the ASEAN FIVE countries stock indices are not only strong, but also are highly regional; (3) Although the higher returns of Indonesia and the Philippines stock markets, the risks are relatively high; while the risks of Thailand and Singapore stock markets are the highest, their returns are belonging to the lowest regions, the relatively less suitable market for investing among the five stock markets; (4) The total variance of Indonesia, Thailand, Malaysia, the Philippines and Singapore stock indices are mainly influenced by the GARCH variance; (5) When major 212
13. events occur in each country, the total volatility, GARCH volatility and jump volatility of stock markets will enhance. Our results confirm the existence of spillover effects originating in the US, along with varying jumps with time. Of the ASEAN FIVE economies examined in this study (Indonesia, Thailand, Malaysia, the Philippines and Singapore), the returns and volatility of the US stock market are found to have the greatest contagious effects on the markets of Malaysia and the Philippines. As for market efficiency, the Indonesia and the Philippines stock markets are found to be most efficient, with the Thailand and Singapore stock markets are exhibiting inverse situations. In terms of market risk, Indonesia and the Philippines are found to present the higher risks with higher returns, while the Thailand and Singapore markets presents the highest risks with lowest returns. That is, portfolio investors can receive sound returns from taking diversified positions in the index of these countries, a result similar to Bhar and Nikolova (2009). As noted in both Chung and Chiang (2006) and Chiu et al. (2006), the different responses in the ASEAN FIVE markets to the arrival of news, whether originating regionally or globally, may stem mainly from the distinct market microstructure characteristics. From a perspective of asset allocation, investors should consider investing more funds in the Indian stock market, while being conservative their investment weighting in Brazil and China in order to avoid higher risks and lower efficiency. In conclusion, based upon our finding of the existence of spillover effects from the US stock market, we suggest that the governments of the ASEAN FIVE should aim to gain a complete understanding of the impact of the US stock market, along with the changes in policies and trading systems, which will ultimately lead to improvements in efficiency and enable these governments to better manage and control market risks. Failure to consider these issues is likely to lead to very costly mistakes. 5. REFERENCES Bhar, R., & Nikolova, B. (2009). Return, volatility spillovers and dynamic correlation in the BRIC equity markets: An analysis using a bivariate EGARCH framework. Global Finance Journal, 19, 203-218. Bowman, R. G., Chan, K. F., & Comer, M. R. (2010). Diversification, rationality and the Asian economic crisis. Pacific-Basin Finance Journal, 18, 1-23. Chan, W.H and Maheu, J.M. (2002), Conditional jump dynamics in stock market returns,” Journal of Business & Economic Statistics, 20, 377-389. Cheung, W., Fung, S., & Tsai, S. C. (2010). Global capital market interdependence and spillover effect of credit risk: Evidence from the 2007-2009 global financial crisis. Applied Financial Economics, 20(1 & 2), 85-103. Chiang, S. M., Yeh, C. P., & Chiu, C. L. (2009). Permanent and transitory components in the Chinese stock market: The ARJI-Trend model. Emerging Markets Finance and Trade, 45(3), 35-55. Chiu, C. L., Chiang, S. M., & Kao, F. (2006). The relationship between the S&P 500 spot and futures indices: Brothers or cousins? Applied Financial Economics, 16, 405-412. 213
14. Christiansen, C. (2007). Volatility-spillover effects in European bond markets. European Financial Management, 13(5), 923-948. Chung, H. M., & Chiang, S. M. (2006). Price clustering in E-mini and floor-traded index futures. Journal of Futures Markets, 3, 269-295. Dickey, D. and Fuller, W.A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association, 74(366a), 427- 431. Engle, R.F. and Granger, C.W.J. (1987). Co-integration and error correction: representation, estimation, and testing. Econometrica, 55(2), 251-276. Engle, R.F. and Yoo, B. S. (1987). Forecasting and testing in co-integrated systems. Journal of Econometrics, 35(1), 143-159. Granger, C. and Newbold, P. (1974). Spurious regressions in econometrics. Journal of Econometrics, 2(2), 111-120. Kim, H. Y., & Mei, J. P. (2001). What makes the stock market jump? An analysis of political risk on Hong Kong stock returns. Journal of International Money and Finance, 20(7), 1003-1016. Liu, Y. A., & Pan, M. S. (1997). Mean and volatility spillover effects in the U.S. and Pacific-Basin stock markets. Multinational Finance Journal, 1, 47-62. Maheu, J., & McCurdy, T. (2004). News arrival, jump dynamics and volatility components for individual stock returns. Journal of Finance, 59, 755-793. Ng, A. (2000). Volatility spillover effects from Japan and the US to the Pacific-Basin. Journal of International Money and Finance, 19(2), 207-233. Phillips, P.C.B. and Perron, P. (1988). Testing for a unit root in time series regression. Biometrika, 75(2), 335-346. Said, S.E. and Dickey, D.A. (1984). Testing for unit roots in autoregressive-moving average models of unknown order. Biometrika,71(3), 599-607. Vrugt, E. B. (2009). U.S. and Japanese macroeconomic news and stock market volatility in Asia-Pacific. Pacific-Basin Finance Journal, 17, 611-627. Wilson, D., & Purushothaman, R. (2003). Dreaming with BRICs: The path to 2050. Global Economics Paper No: 99 by Goldman Sachs. 214
15. Table 2 The Estimation and Testing Result of the VIX ARJI Model Part A：The Estimation Result of the ARJI Model Parameters Country Indonesia Thailand Malaysia the Philippines Singapore Variables 0.1473 0.0705 0.0344 0.0733 0.0555 µ (0.0248) (0.0221) (0.0113) (0.0225) (0.0159) -0.0113 0.0125 0.0341* 0.0224 -0.1313 φ 1 (0.0224) (0.0222) (0.0192) (0.0189) (0.0222) ψ -0.0051 -0.0001 -0.0129 -0.0264 -0.0017 1 (0.0034) (0.0032) (0.0016) (0.0032) (0.0026) 0.0885 0.0427 0.0843 0.0642 ς 1 (0.0150) (0.0070) (0.0174) (0.0127) 0.0849 0.0017 0.0270 0.0879 ς 2 (0.0178) (0.0084) (0.0172) (0.0128) 0.1630 0.0798 0.5888 0.2548 ς 3 (0.0366) (0.0260) (0.0320) (0.0234) 0.1200 0.0625 0.1406 0.0921 ς 4 (0.0226) (0.0150) (0.0080) (0.0115) 0.1785 0.1589 0.0747 0.0393* ς 5 (0.0272) (0.0200) (0.0109) (0.0202) 0.0554 0.0620 0.0351 0.2373 0.0182 ω (0.0140) (0.0049) (0.0027) (0.0711) (0.0021) 0.0726 0.1125 0.0623 0.0794 0.0754 α 1 (0.0227) (0.0054) (0.0121) (0.0225) (0.6000) 0.8513 0.8263 0.6722 0.5684 0.8656 β 1 (0.0296) (0.0050) (0.0191) (0.0989) (0.0064) -1.3329 -1.9173 -0.0478 -0.1427 -0.2029 θ (0.4595) (0.8878) (0.0294) (0.1011) (0.0661) 2.7430 4.5166 0.8102 1.5642 1.1112 δ (0.3571) (0.4582) (0.0379) (0.2127) (0.0890) 0.0437 0.0001 0.0064 0.0088 0.0017 λ 0 (0.0417) (0.0000) (0.0014) (0.0071) (0.0006) 0.4786 0.9956 0.9881 0.9712 0.9954 ρ 1 (0.6276) (0.0017) (0.0037) (0.0187) (0.0025) 0.2496* 0.0050 0.1950 0.2064 0.1554 γ 1 (0.1321) (0.0066) (0.0358) (0.0992) (0.0395) Q(20) 29.5770 21.5179 18.057 19.220 21.4530 2 Q (20) 6.9190 0.9058 16.697 17.483 23.7730 Likelihood function value -3383.1369 -3236.1680 -1907.8268 -3199.5768 -2774.1773 Part B: Fundamental Statistics of the Conditional Jump Joint Process 215
16. Value % Value % Value % Value % Value % GARCH Variance 26.447 91.82 35.992 99.60 5.768 77.35 9.936 81.15 15.605 90.81 Jump Variance 2.358 8.19 0.275 0.76 1.689 22.65 2.308 18.85 1.662 9.67 Total Conditional Variance 28.804 100 36.138 100 7.457 100 12.244 100 17.184 100 Note: 1. *, , separately indicates significance at the 10%、5%、1% level. 2.Q(20) and Q2(20) are separately 2 lags Ljung-Box Q-statistics. ς ς ς ς 3. For Indonesia, 2 、 3 、 4 、 5 are coefficients of the other 4 countries stock index returns. ς ς ς ς 4. For Thailand, 1 、 3 、 4 、 5 are coefficients of the other 4 countries stock index returns. ς ς ς ς 5. For Malaysia, 1 、 2 、 4 、 5 are coefficients of the other 4 countries stock index returns. ς ς ς ς 6. For the Philippines , 1 、 2 、 3 、 5 are coefficients of the other 4 countries stock index returns. ς ς ς ς 7. For Singapore, 1 、 2 、 3 、 4 are coefficients of the other 4 countries stock index returns. Figure 3.1 The Trend of Indonesia GARCH Volatility、Jump Volatility and Total Volatility Figure 3.2 The Trend of Thailand GARCH Volatility、Jump Volatility and Total Volatility 216
17. Figure 3.3 The Trend of Malaysia GARCH Volatility、Jump Volatility and Total Volatility Figure 3.4 The Trend of the Philippines GARCH Volatility、Jump Volatility and Total Volatility Figure 3.5 The Trend of Singapore GARCH Volatility、Jump Volatility and Total Volatility 217
18. Figure 3.6 The Time-Varying Jump Intensity of Indonesia Figure 3.7 The Time-Varying Jump Intensity of Thailand Figure 3.8 The Time-Varying Jump Intensity of Malaysia 218
19. Figure 3.9 The Time-Varying Jump Intensity of the Philippines Figure 3.10 The Time-Varying Jump Intensity of Singapore 219
20. Table3 The Jump Intensity and Total Volatility of ASEAN FIVE Stock Indices during the event occurred country Indonesia Thailand Malaysia the Philippines Singapore Jump Total Jump Jump Jump Jump date Intensity Volatility Intensity Total Intensity Total Intensity Total Intensity Total the Southern PhilippinesVolatility 6.5 EarthquakeVolatility Occurred in 2007Volatility Volatility 20070821 - - - - - - 0.5633 3.0614 - - 20070824 - - - - - - 0.7805 6.6516 - - 20070828 - - - - - - 0.7631 6.3084 - - 20070829 - - - - - - 0.7180 5.3295 - - 20070830 - - - - - - 0.6662 4.4543 - - 20070905 - - - - - - 0.7029 4.4073 - - the 12th General Election of Malaysia in 2008 20080311 - - - - 2.0792 7.4570 - - - - 20080312 - - - - 2.0425 6.1296 - - - - 20080313 - - - - 2.0005 4.9523 - - - - 20080314 - - - - 1.9851 4.2924 - - - - 20080317 - - - - 1.9268 3.4020 - - - - the Thailand Bloodshed Demonstrations in 2008 20081008 - - 0.0259 31.4685 - - - - - - 20081010 - - 0.0259 30.5292 - - - - - - 20081013 - - 0.0258 25.7941 - - - - - - 20081014 - - 0.0258 21.6739 - - - - - - 20081015 - - 0.0257 18.4632 - - - - - - Table3 The Jump Intensity and Total Volatility of ASEAN FIVE Stock Indices during the event occurred (continued) country Indonesia Thailand Malaysia the Philippines Singapore Jump Total Jump Total Jump Total Jump Total Jump Total date Intensity Volatility Intensity Volatility Intensity Volatility Intensity Volatility Intensity Volatility the Indonesian Equities Indefinite Cease Trading when Asian Stock Crash in 2008 20081010 0.4420 14.7028 - - - - - - - - 20081013 0.2247 12.0881 - - - - - - - - 20081014 0.1371 10.1029 - - - - - - - - 20081015 0.1070 10.8671 - - - - - - - - 20081016 0.0897 9.3824 - - - - - - - - 20081017 0.0848 8.3968 - - - - - - - - the Subprime Mortgage Crisis in 2008 20081028 0.0789 6.2807 0.0264 19.6779 1.2551 1.2724 0.9361 3.7165 1.3800 13.3302 20081029 0.5216 28.8043 0.0265 28.8780 1.6371 2.6255 1.4564 16.3049 1.3764 15.2574 20081030 0.2877 24.3660 0.0264 27.9035 1.6175 2.3387 1.4192 15.6600 1.3700 17.1837 20081031 0.1654 21.6258 0.0263 23.2235 1.6459 2.3250 1.3638 13.4012 1.3583 15.1909 20081103 0.1135 20.2171 0.0263 19.6963 1.5847 1.9960 1.3178 11.6616 1.3416 14.3325 20081104 0.0926 20.3870 0.0262 16.3909 1.7736 2.6275 1.2411 9.6100 1.3467 14.0279 220
21. Table3 The Jump Intensity and Total Volatility of ASEAN FIVE Stock Indices during the event occurred (continued) country Indonesia Thailand Malaysia the Philippines Singapore Jump Total Jump Total Jump Total Jump Total Jump Total date Intensity Volatility Intensity Volatility Intensity Volatility Intensity Volatility Intensity Volatility the Easing Monetary Policy Announced by the Monetary Authority of Singapore in April to May 2009 20090416 - - - - - 1.3340 6.0925 20090417 - - - - - 1.3081 5.6561 20090420 - - - - - 1.2879 5.1550 20090421 - - - - - 1.2588 4.6489 20090422 - - - - - 1.2611 4.5735 20090423 - - - - - 1.2296 4.1481 the US QE Exit in 2013 20130617 0.3249 3.7899 0.0204 7.3610 - - 1.1473 10.9693 - - 20130618 0.2308 3.5362 0.0204 7.0477 - - 1.1147 9.9189 - - 20130619 0.1682 3.6847 0.0204 5.9288 - - 1.0531 8.2087 - - 20130620 0.1041 3.1375 0.0204 5.4843 - - 1.0606 7.3983 - - 20130621 0.0814 2.8549 0.0203 4.8940 - - 1.0588 6.6635 - - Sample 0.0703 1.9137 0.0187 4.0881 0.4765 0.5996 0.2997 1.7039 0.3577 1.4143 mean 221