Printer Friendly

ASEAN 5 stock markets, currency risk and volatility spillover.

INTRODUCTION

Generally, volatility spillover occurs when changes in price volatility in one market create a lagged impact in other markets. When applied to currencies and stock markets, volatility spillover occurs when changes in foreign currency markets affect stock markets, over and above local effects. As several European and Asian countries consider the benefits of joining the Eurozone and ASEAN, respectively, the impact of volatility transmissions and spillovers raises key financial and policy questions that need to be further studied. From a business perspective, the prevalence of volatility spillovers can guide multinational corporations in managing their currency risk and exposure in these countries, a key element in their international diversification efforts. (Kanas, 2000).

This research investigates the interdependence of stock returns and exchange rate changes in the ASEAN5 countries. The countries included are the Philippines, Singapore, Malaysia, Thailand and Indonesia for the period January 4, 2000 to December 31, 2010. This study will also examine if there are volatility spillovers from stock returns to exchange rate changes present in each country and the ASEAN5.

THEORETICAL AND CONCEPTUAL FRAMEWORK

The Nature of Volatility Transmission and Volatility Spillover

Two approaches provide the possible link between exchange rates to the other economic and financial sectors. The first, so-called "flow model" looks at the impact of exchange rates on the balance of trade, such as those studied by Mundell in 1963 and by Dornbusch and Fisher in 1980. The flow model posits that changes in exchange rates affect international competitiveness and trade balances, thereby influencing real income and output. Stock prices, generally interpreted as the present values of future cash flows of firms, react to exchange rate changes and form the link among future income, interest rate innovations, and current investment and consumption decisions. (Yang and Doong, 2004)

The other model, "stock-oriented" models of exchange rates such as those studied by Branson (1983) and Frankel (1983) models view exchange rates as equating the supply and demand for assets such as stocks and bonds. This approach gives the capital account an important role in determining exchange rate dynamics. Since the values of financial assets are determined by the present values of their future cash flows, expectations of relative currency values play a considerable role in their price movements, especially for internationally held financial assets. Therefore, stock price innovations may affect, or be affected by, exchange rate dynamics. (Ibid, 1984)

An illustration of the second approach can be seen in Figure 1, where transmission and spillover is seen as an input-process-output model:

Because there has been no dominant approach to explain the impact of volatility spillover, numerous studies have populated the literature in recent years. The residual effect of the Global Financial Crisis still being felt in many countries as well as those "integrated" economies such as the Eurozone and ASEAN provide the motivation for sustained interest in this field of study.

LITERATURE REVIEW

Kanas (1998 and 2000) was one of the first to have examined volatility spillovers in the foreign exchange and stock markets. Using EGARCH, he studied the interdependence of stock returns and exchange rate changes among six industrialized countries, namely the United States (US), the United Kingdom (UK), Japan, Germany, France and Canada. The study concluded that there is evidence of volatility spillovers from stock returns to exchange rates changes for all countries except Germany. However, volatility spillovers from exchange rate changes to stock returns are insignificant for all countries.

[FIGURE 1 OMITTED]

Mishra and Rahman (2010) examined the dynamics of stock market returns volatility of India and Japan using the Threshold Generalized Autoregressive Conditional Heteroskedasticity (TGARCH-M) model. They concluded that return volatility persists in both countries. Savva, Osborn, Gill (2009) used the asymmetric Dynamic Conditional Correlations(DCC) version of the VAR-multivariate Exponential Generalized Autoregressive Conditional Heteroskedasticity (EGARCH) model for daily stock market returns across four major world markets, namely New York, London, Frankfurt and Paris. The results showed that the correlation in the post-euro period was highest between Frankfurt and Paris. Also, the presence of spillover effects from foreign markets for both returns and volatilities exists. The results are consistent with Kanas (1998) where he used the EGARCH model to capture potential asymmetric effects of volatility across the three largest European stock markets, namely London, Frankfurt and Paris. The results showed that there were reciprocal spillovers between London and Paris, and between Paris and Frankfurt, and unidirectional spillovers from London to Frankfurt. Yang and Doong (2004) explored the nature of the mean and volatility transmission mechanism between stock and foreign exchange markets for the G-7 countries. The results point to significant volatility spillovers and an asymmetric effect from the stock market to the foreign exchange market for France, Italy, Japan and the US, suggesting integration between stock and foreign exchange markets in these countries. (O' Donnell and Morales, n.d.)

Three other studies examined if there are price and volatility spillovers from the US market to other countries. In the case of Hong Kong, Singapore, Taiwan and Malaysia except Korea, there was a decrease in price and volatility spillovers from the US market since the 1997 Asian financial crisis. The study used the EGARCH model for the prior- and post-crisis periods. Data is the daily stock prices from January 3, 1995 to April 24 2001. (Nam, Yuhn, and Kim, 2008). In the case of Europe and the US, Anaraki, (n.d.) investigated the link by using Granger causality. The causality runs from the US to European stock market and that the US fundamentals including the Federal Fund Rate (FFR), the Euro-dollar exchange rate, and the US stock market indices affect European stock market volatility. Using a multivariate generalized autoregressive conditional heteroskedasticity (GARCH-M) model, Chancharoenchai and Dibooglu (2006) examined volatility spillovers in six Southeast Asian stock markets pre and after the 1997 Asian crisis and its interactions with the U.S. market (using the New York Stock Exchange as the global market), Japan (using the Tokyo Stock Exchange as a regional market). The study concluded that there were some interdependence in volatility between emerging markets and developed markets before and after crisis. The study showed evidence of the Asian contagion which started in Thailand and affected other financial markets.

Most of the literature on the international interactions of stock returns, foreign exchange rate changes and volatility spillovers employ Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models (Bollerslev, 1986). These models determine if there is volatility clustering, fat tails, volatility spillover. Volatility clustering in asset returns means that "large price changes follow large price changes of either sign and small price changes follow small prices changes of either sign." (Mandelbrot , 1963). The GARCH (1,1) is often used since the model is parsimonious. Another GARCH model was developed by Nelson (1991), the Exponential GARCH to study the asymmetrical effects of shocks on stock return volatility, known as leverage effect. The results showed that negative shocks have larger effects on volatility than positive shocks. (Kanas 1998).

METHODOLOGY

We used daily closing stock prices denominated in local currency for the Philippines Stock Exchange Index (denoted in the runs as PSEi) in the case of the Philippines, the FTSE Straits Times Index (denoted as FSSTI Index) for Singapore, the FTSE Bursa Malaysia KLCI Index (denoted as FBMKLCI) for Malaysia, the Stock Exchange of Thailand (denoted as SET) for Thailand, and the Jakarta Composite Index (denoted as JCI Index) for Indonesia for the period from 4 January 2000 to 31 December 2010. The stock indices are not adjusted for dividends since they will not affect the results (Kanas, 2000). For exchange rate, the currencies used were: the Philippine Peso, Indonesian Rupiah, Singapore Dollar, Thai Baht and Malaysian Ringgit. The exchange rate series for each country and indices were all derived from Bloomberg.

Following Kanas (2000), we compute for the stock returns (denoted as ret in the results) and exchange rate changes (denoted by [S.sub.t] and [E.sub.t], respectively) are calculated as the difference between the natural logarithms of the closing values for two consecutive trading days, i.e. [S.sub.t] = ln([P.sup.S.sub.t]) - ln([sup.PS.sub.t-1]), and [E.sub.t] = ln([P.sup.E.sub.t]) - ln ([P.sup.E.sub.t-1]), where [P.sup.S.sub.t] and [P.sup.E.sub.t] are the stock price and the exchange rate at period t, respectively.

GARCH (1,1) specification was used since it is a parsimonious representation of conditional variance of time series data (Bollerslev, 1987), in this case, the stock returns and foreign exchange changes.

RESEARCH FINDINGS: MODEL RESULTS

Descriptive statistics for stock returns of the indices are reported in Table 1. The sample means of returns are positive and statistically different from zero. The variances range from 0.009 (Malaysia) to 0.1525 (Indonesia). The measures for skewness indicate that all countries except the Philippines (PSEi) are negatively skewed and excess kurtosis indicate that the distributions of returns for all markets are leptokurtic.

Descriptive statistics for foreign exchange rate changes are reported in Table 2. The sample means of returns are positive except for Malaysia and Singapore and statistically different from zero. The variances range from 0.0000208 (Thailand) to 0.007 (Indonesia) The measures for skewness indicate that all countries except Thailand are negatively skewed and excess kurtosis indicate that the distributions of the foreign exchange rate changes for all markets are leptokurtic.

Table 3 shows the correlation coefficient for each variable. The PSEi stock return is negatively correlated to the Indonesian and Singaporean stock returns, while positively correlated to Malaysian and Thai stock returns. The Malaysian, Indonesian, Thai, and Singaporean stock returns are positively correlated to each other. The foreign exchange rate change, retpeso (Philippines) is negatively correlated to all countries' stock returns including the retpsei (Philippines). The Indonesian rupiah and Thai baht foreign exchange rate changes are also negatively correlated with their respective stock returns on the indices. Singapore dollar and Malaysia ringgit foreign exchange rate changes are positively correlated with their respective stock returns on the indices. An increase in the foreign exchange in one country is viewed as favorable in the other. This is typical of these two countries as both are active trading and financial partners.

Table 4 shows the results of GARCH (1,1) for retpsei and retpeso. The p-value is significant. Also, volatility spillover exists between the Philippine stock market to the Philippine foreign exchange rate based on the alpha and beta results.

Table 5 shows the results of GARCH (1,1) for retjcindex and retidr. The p-value is not significant. Also, volatility spillover exists between the Indonesian stock market to the Indonesian foreign exchange rate based on the alpha and beta results.

Table 6 shows the results of GARCH (1,1) for retset and retthb. The p-value is not significant. Also, volatility spillover exists between the Thai stock market to the Thai foreign exchange rate based on the alpha and beta results.

Table 7 shows the results of GARCH (1,1) for retfblmklc and retmyr. The p-value is significant. Also, volatility spillover exists between the Malaysian stock market and the Malaysian foreign exchange rate based on the alpha and beta results.

Table 8 shows the results of GARCH (1,1) for retFSSTI Index and retsgd. The p-value is not significant. Also, volatility spillover exists between the Singaporean stock market and the Singaporean foreign exchange rate based on the alpha and beta results.

Tables 9, 10, and 11 show that volatility spillovers from exchange rate changes to stock returns are significant for Singapore, Indonesia and the Philippines. However, software runs for Malaysia and Thailand resulted into error despite several attempts. This implies that the convergence criterion was not present.

Tables 12, 13, and 14 show that the stock returns for the Philippine, Indonesian and Singaporean markets have persistent volatility and that there are spillovers to the other stock markets and foreign exchange rates. This means that the ASEAN5 stock market and foreign exchange markets are integrated and that any news on these markets will affect the other countries' markets.

The study included a GARCH (1,1) specification for both the stock returns on indices and foreign exchange changes to cover a subperiod starting June 1, 2007. On June 2007, Bear Sterns were forced to sell assets after their hedge funds with large holdings of subprime mortgages suffered large losses(Guillen, n.d). We will use this date as a subperiod to determine if volatility persists during the start of a global financial crisis and if there are evidences of volatility spillovers until the end of December 2010.

Tables 15, 16, 17, 18, 19 show that the p-value of the retmyr is significant in all stock returns on indices except for Malaysia. The retmyr has a positive relationship with both the Singaporean and Thai stock returns on indices. The retmyr has a negative relationship with both the Philippine and Indonesian stock returns on indices.

Table 16 shows that the p-value of retsgd is significant and has a positive relationship with retjci. Table 17 and 18 show that the p-value of retpeso is significant and has a negative relationship with retset and retfbmklci_inde, respectively. Table 19 shows that the p-value of retthb is significant and has a negative relationship with retfssti.

Tables 15, 16, 17,18, and 19 show that volatility is persistent and that there are spillovers to the other stock markets and foreign exchange rates. The alpha and beta values are higher during this subperiod of the financial crisis which indicates that the market takes time to absorb the impact of information and volatility persists.

Tables 20, 21, 22, 23, and 24 show that volatility is persistent in the foreign exchange changes in all countries. Table 20 shows that the p-values of retthb and retfbmklci_inde are significant and are both negatively related with retpeso. Table 21 shows that retmyr, retfssti have significant p-values and are both positively related with retidr. Table 22 shows that retsgd, retpeso, retset have significant p-values. Both retpeso and retset have negative relationship with retthb while retsgd is positively related. Based on Table 22,it is only the Thai stock returns (retset) and Thai foreign exchange changes that showed evidence of volatility spillover in the same country and is negatively related. Table 23 shows that retsgd, retidr have significant p-values and positively related with retmyr. Table 24 shows retjci, retthb, retmyr have significant p-values and are all positively related with retsgd. Based on Tables 20, 21, 22, 23,and 24, we find evidence of volatility spillover from foreign exchange markets to stock markets other than its own country.

Our findings can be summarized as follows:

1). There is presence of volatility clustering in the stock markets and foreign exchange markets as evidenced by the (significant) high alpha (1) value This means volatility in the previous period will have an impact on the volatility of the current period returns.

2). The high beta (1) value in all models mean that volatility is quick to react to movements in the stock market and foreign exchange market and volatility tend to be spikier.

3). There is also evidence of spillovers from stock returns to exchange rate changes for all countries. This implies that there is some form of interaction between the stock and foreign exchange markets within the ASEAN5 countries.

4). Volatility spillovers from exchange rate changes to stock returns are also significant for all countries.

5). There is more volatility during the sub-period June 1, 2007-December 31, 2010 based on the higher alpha and beta results for all stock market and foreign exchange markets as compared for the whole period January 4, 2000 to December 31, 2010.

6). Since the ASEAN5 countries have moved towards increasing interdependence with each other, any news affecting either the stock market or foreign exchange market of one country would have a volatility spillover effect in that country and spread in the region.

Of particular interest in the findings is the significance of the results for the Philippine peso and the PSEi. This implies the presence of numerous foreign investors in the Philippine stock market and any instance where volatility exists, usually adverse, allows the opportunity for these foreign investors to pull out almost immediately, affecting foreign currency levels. This also affirms the impact of so-called "hot money" as one of the main drivers of prices in the Philippine stock market. This is not as strong or even present in the other ASEAN5 countries.

CONCLUSION AND AREAS FOR FURTHER STUDY

This study provides evidence of volatility spillover within and among the ASEAN5 countries. This study also affirms the applicability of GARCH to determine the levels of transmission and spillover among the countries. The study had two periods, from January 4, 2000-December 31, 2010 and a sub-period June 1, 2007 to December 31, 2010 to capture the volatility during the global financial crisis. However, it would be interesting to find out if there are periods where spillovers as not significant. It would also be interesting if the spillovers were present before and after the Global Financial Crisis of 2008, or even during the Asian Financial Crisis of 1997. Dividing the study into sub periods would reveal more details that are otherwise not captured in this study. The EGARCH model can also be used to capture leverage of asymmetric effects for future studies.

REFERENCES

Anaraki, N. (n. d.). The European Stock Market Impulse to the US Financial Crisis. Retrieved March 22, 2011 from http://www.pdfdownload.org/pdf2html/pdf2html.php?url= http%3A%2F%2Fwww.aabri.com%2FOC09man uscripts%2FOC09012.pdf&images=yes

Awayan, N., Guerra, F., Quipones, M., and Tang Woo, S. (2010). A Study on the Volatility Transmission in the Exchange Rates of the Currencies of the Philippines, South Korea, and Indonesia and the Impacts (sic) of the Global and Asian Financial Crises for the Period 1990-2009. Unpublished Manuscript, De La Salle University

Bollerslev, T. (1986). Generalized Autoregressive Conditional Heteroskedasticity. Journal of Econometrics 31, 307-327

Branson, W. H. (1983). Macroeconomic Determinants of Real Exchange Risk. Managing Foreign Exchange Risk. R. J. Herring ed., Cambridge: Cambridge University Press.

Dornbusch, R. and S. Fischer, (1980). Exchange Rates and the Current Account. American Economic Review, 70(5), 960-971.

Frankel, J. A., (1983). Monetary and Portfolio-Balance Models of Exchange Rate Determination. Economic Interdependence and Flexible Exchange Rates. J. S. Bhandari and B. H. Putnam eds., Cambridge: MIT Press.

Guillen, M. (n.d). The The Global Economic & Financial Crisis: A Timeline. Retrieved on April 27, 2011 from http://lauder.wharton.upenn.edu/pdf/Chronology%20Economic%20%20Financial% 20Crisis.pdf

Kanas, A. (1998). Volatility Spillovers Across Equity Markets: European Evidence. Applied Financial Economics. Taylor and Francis Journals, 8(3), 245-56.

Kanas, A. (2000). Volatility Spillovers between Stock Returns and Exchange Rate Changes: International Evidence. Journal of Business Finance and Accounting,. 27, 447-467.

Chancharoenchai, K.. and Dibooglu, S. (2006). Volatility Spillovers and Contagion During the Asian Crisis: Evidence from Six Southeast Asian Stock Markets. Emerging Markets Finance and Trade, March-April 42(2), 4-17.

Mandelbrot, B. (1963). The variation of certain speculative prices. Journal of Business, 36,.394-419.

Mishra, B. and Rahman, M. (2010). Dynamics of Stock Market Return Volatility: Evidence from the Daily Data of India and Japan. The International Business & Economics Research Journal, May, 9(5), 79-84.

Mundell, R., (1963). Capital Mobility and Stabilization Policy under Fixed and Flexible Exchange Rate. Canadian Journal of Economics and Political Science, 29, 475-467.

Nam, J. H., Yuhn, K., and Kim, S. B. (2008). What happened to Pacific-Basin emerging markets after the 1997 financial crisis? Applied Financial Economics. 18,639-658.

Nelson, D. B. (1991). Conditional Heteroskedasticity in Asset Returns: A New Approach. Econometrica, 59, 347370.

O' Donnell, M. and Morales. L. (n.d.) Volatility Spillovers Between Stock Returns and Foreign Exchange Rates: Evidence from Four Eastern European Countries. Retrieved on March 11, 2011 from http://www.fma.org/Prague/Papers/EECVOLATILITY.MoralesnOnnDonnell.pdf

Savva, C.S., Osborn, D.R., and Gill, L. (2009). Spillovers and Correlations between US and Major European Stock Markets: The Role of the Euro. Applied Financial Economics, 19, 1595-1604.

Yang, S. and Doong, S. (2004). Price and Volatility Spillovers between Stock Prices and Exchange Rates: Empirical Evidence from the G-7 Countries. International Journal of Business and Economics, 3(2), 139-153.

Leila C. Kabigting, University of Guam

Rene B. Hapitan, De La Salle University
Table 1: Summary Statistics, Stock Returns
using the observations 2000/01/04-2010/12/30
(missing values were skipped)

Variable              Mean         Median       Minimum      Maximum

retpsei           0.000246919    0.000107464   -0.130887    0.161776
retjciindex       0.000624071    0.00127921    -0.109540    0.0762312
retset            0.000270504    0.000404790   -0.160633    0.105770
retfbmklci        0.000221836    0.000462797   -0.0997851   0.0450273
retfSSTI Index    8.12490e-005   0.000449251   -0.0869598   0.0753053

Variable           Std. Dev.        C.V.        Skewness       Ex.
                                                            kurtosis

retpsei            0.0143023       57.9230      0.500516     15.6014
retjciindex        0.0152522       24.4399     -0.627927     5.67023
retset             0.0151458       55.9912     -0.756638     8.99236
retfbmklci         0.00940011      42.3741     -0.863193     9.14827
retfSSTI Index     0.0132466       163.037     -0.241634     4.36202

Table 2: Summary Statistics, FOREIGN EXCHANGE,
using the observations 2000/01/04-2010/12/30
(missing values were skipped)

Variable       Mean           Median        Minimum        Maximum

retpeso    3.60732e-005      0.000000      -0.142778      0.0371944
retidr     8.16137e-005      0.000000      -0.0897804     0.0590335
retthb     3.63517e-005    2.62916e-005   2.46655e-005   8.21164e-005
retmyr     -7.52771e-005     0.000000      -0.0231778     0.0175549
retsgd     -8.88746e-005   -0.000115380    -0.0203807     0.0157647

Variable     Std. Dev.         C.V.         Skewness     Ex. kurtosis

retpeso     0.00488392       135.389        -9.52797       280.322
retidr      0.00735742       90.1493       -0.395801       17.1222
retthb     2.08127e-005      0.572537       1.50691        0.293686
retmyr      0.00264459       35.1314       -0.345536       11.0646
retsgd      0.00311705       35.0724       -0.179246       3.74293

Table 3: Correlation coefficients, using the observations
2000/01/04-2010/12/30 (missing values were skipped)
5% critical value (two-tailed) = 0.0366 for n = 2868

retpsei   retjciindex  retset   retfbmklci  retfSSTI
                                             Index

1.0000      -0.0119    0.0099     0.0092    -0.0059      retpsei
            1.0000     0.0270     0.0323     0.0036    retjciindex
                       1.0000     0.0184     0.0002       retset
                                  1.0000     0.0538     retfbmklci
                                             1.0000   retfSSTI Index

retpeso     retidr     retthb     Retmyr     retsgd

-0.0015     0.0127     -0.0105    0.0278     0.0134      retpsei
-0.0468     -0.0093    0.0054    -0.0209    -0.0371    retjciindex
-0.0031     0.0010     -0.0068    0.0012    -0.0121       retset
-0.0254     -0.0012    -0.0062    0.0274     0.0227     retfbmklci
-0.0261     -0.0168    0.0360    -0.0163     0.0441   retfSSTI Index
1.0000      0.0613     0.0002     0.0169     0.0055      retpeso
            1.0000     0.0230     0.0192     0.1268       retidr
                       1.0000     0.0304     0.0315       retthb
                                  1.0000     0.0452       retmyr
                                             1.0000       retsgd

TABLE 4: Model 1 ARCH, using observations 2000/01/05-2010/05/11
(T = 2700)
Dependent variable: retpsei
Standard errors based on Hessian

                         Coefficient           Std. Error

const                    0.000409585           0.00023327
retpeso                    -0.42622            0.0382203
alpha(0)                 1.46418e-05          2.48929e-06
alpha(1)                   0.199983            0.0211614
beta(1)                    0.746274             0.023964
Mean dependent var         0.000235
Log-likelihood             7870.915
Schwarz criterion         -15694.42

                              z                 p-value

const                       1.7558             0.07912 *
retpeso                    -11.1517           <0.00001 ***
alpha(0)                    5.8819            <0.00001 ***
alpha(1)                    9.4504            <0.00001 ***
beta(1)                    31.1415            <0.00001 ***
Mean dependent var    S.D. dependent var        0.014317
Log-likelihood         Akaike criterion        -15729.83
Schwarz criterion        Hannan-Quinn          -15717.03

Unconditional error variance = 0.00027244

Table 5: Model 2: GARCH, using observations
2000/01/04-2010/03/26 (T = 2669)
Dependent variable: retjciindex
Standard errors based on Hessian

                           Coefficient             Std. Error

const                       0.00132724            0.000251939
retidr                      -0.049531              0.0331003
alpha(0)                    1.4793e-05            2.71361e-06
alpha(1)                     0.141121               0.017445
beta(1)                      0.795718              0.0239791
Mean dependent var           0.000624
Log-likelihood               7634.732
Schwarz criterion           -15222.13

                                z                   p-value

const                         5.2681              <0.00001 ***
retidr                       -1.4964                0.13455
alpha(0)                      5.4514              <0.00001 ***
alpha(1)                      8.0895              <0.00001 ***
beta(1)                      33.1839              <0.00001 ***
Mean dependent var      S.D. dependent var          0.015252
Log-likelihood           Akaike criterion          -15257.46
Schwarz criterion          Hannan-Quinn            -15244.68

Unconditional error variance = 0.000234213

Table 6: Model 3: GARCH, using observations
2000/01/04-2010/04/29 (T = 2693)
Dependent variable: retset
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      0.000761275            0.000506557
retthb                       4.06065                11.9368
alpha(0)                   2.03909e-05            3.41554e-06
alpha(1)                     0.120092              0.0165926
beta(1)                      0.786669              0.0274207
Mean dependent var           0.000271
Log-likelihood               7685.133
Schwarz criterion           -15322.87

                                z                   p-value

const                         1.5028                0.13288
retthb                        0.3402                0.73372
alpha(0)                      5.9700              <0.00001 ***
alpha(1)                      7.2377              <0.00001 ***
beta(1)                      28.6888              <0.00001 ***
Mean dependent var      S.D. dependent var          0.015146
Log-likelihood           Akaike criterion          -15358.27
Schwarz criterion          Hannan-Quinn            -15345.47

Unconditional error variance = 0.000218695

Table 7: Model 4: GARCH, using observations
2000/01/05-2010/05/24 (T = 2709)
Dependent variable: retfbmklci
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      0.000577577            0.000134222
retmyr                      0.0824977              0.0474972
alpha(0)                   1.07521e-06            3.17802e-07
alpha(1)                     0.123606              0.0158681
beta(1)                      0.874124              0.0153776
Mean dependent var           0.000222
Log-likelihood               9185.873
Schwarz criterion           -18324.32

                                z                   p-value

const                         4.3032              0.00002 ***
retmyr                        1.7369               0.08241 *
alpha(0)                      3.3833              0.00072 ***
alpha(1)                      7.7896              <0.00001 ***
beta(1)                      56.8439              <0.00001 ***
Mean dependent var      S.D. dependent var          0.009400
Log-likelihood           Akaike criterion          -18359.75
Schwarz criterion          Hannan-Quinn            -18346.94

Unconditional error variance = 0.000473662

Table 8: Model 5: GARCH, using observations
2000/01/05-2010/08/04 (T = 2761)
Dependent variable: retFSSTI Index
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      0.000547155            0.000181741
retsgd                      0.0263514              0.0624312
alpha(0)                    1.4354e-06            4.16955e-07
alpha(1)                    0.0998542              0.0107854
beta(1)                      0.895657              0.0102046
Mean dependent var           0.000081
Log-likelihood               8451.841
Schwarz criterion           -16856.14

                                z                   p-value

const                         3.0106              0.00261 ***
retsgd                        0.4221                0.67296
alpha(0)                      3.4426              0.00058 ***
alpha(1)                      9.2582              <0.00001 ***
beta(1)                      87.7703              <0.00001 ***
Mean dependent var      S.D. dependent var          0.013247
Log-likelihood           Akaike criterion          -16891.68
Schwarz criterion          Hannan-Quinn            -16878.84

Unconditional error variance = 0.000319773

Table 9: Model 6: GARCH, using observations
2000/01/05-2010/05/11 (T = 2700)
Dependent variable: retpeso
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      8.07153e-05            5.00607e-05
retpsei                    -0.00129963             0.00390811
alpha(0)                   3.47397e-07            7.30592e-08
alpha(1)                     0.222078              0.0180842
beta(1)                      0.777922              0.0174963
Mean dependent var           0.000032
Log-likelihood               11572.62
Schwarz criterion           -23097.83

                                z                   p-value

const                         1.6123                0.10689
retpsei                      -0.3325                0.73948
alpha(0)                      4.7550              <0.00001 ***
alpha(1)                     12.2802              <0.00001 ***
beta(1)                      44.4620              <0.00001 ***
Mean dependent var      S.D. dependent var          0.004881
Log-likelihood           Akaike criterion          -23133.24
Schwarz criterion          Hannan-Quinn            -23120.43

Unconditional error variance = 6.16567e+007

Table 10: Model 7: GARCH, using observations
2000/01/04-2010/03/26 (T = 2669)
Dependent variable: retidr
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      3.43288e-05            9.40919e-05
retjciindex                -0.00515879             0.00555879
alpha(0)                   3.44955e-06            4.37192e-07
alpha(1)                     0.335201              0.0305775
beta(1)                      0.664799              0.0261348
Mean dependent var           0.000089
Log-likelihood               9832.963
Schwarz criterion           -19618.59

                                z                   p-value

const                         0.3648                0.71523
retjciindex                  -0.9280                0.35339
alpha(0)                      7.8902              <0.00001 ***
alpha(1)                     10.9623              <0.00001 ***
beta(1)                      25.4373              <0.00001 ***
Mean dependent var      S.D. dependent var          0.007566
Log-likelihood           Akaike criterion          -19653.93
Schwarz criterion          Hannan-Quinn            -19641.14

Unconditional error variance = 960324

Table 11: Model 8: GARCH, using observations
2000/01/05-2010/08/04 (T = 2761)
Dependent variable: retsgd
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      -0.000105167           5.12557e-05
retfSSTI Index              0.00627846             0.00401432
alpha(0)                   1.06892e-07            3.26818e-08
alpha(1)                    0.0437789              0.0069778
beta(1)                      0.945068              0.00892546
Mean dependent var          -0.000073
Log-likelihood               12246.09
Schwarz criterion           -24444.63

                                z                   p-value

const                        -2.0518               0.04019 **
retfSSTI Index                1.5640                0.11781
alpha(0)                      3.2707              0.00107 ***
alpha(1)                      6.2740              <0.00001 ***
beta(1)                      105.8844             <0.00001 ***
Mean dependent var      S.D. dependent var          0.003088
Log-likelihood           Akaike criterion          -24480.17
Schwarz criterion          Hannan-Quinn            -24467.34

Unconditional error variance = 9.58388e-006

Table 12: Model 9: GARCH, using observations
2000/01/05-2010/03/26 (T = 2668)
Dependent variable: retPSEi
Standard errors based on Hessian

                           Coefficient             Std. Error

Const                        0.051502              0.0315025
retfSSTI Index               0.151885                1.3426
retjciindex                 -0.759426                1.0524
Retset                      -0.0530194              1.16144
retfbmklci                   -1.07831                1.8491
Retpeso                      -30.5301               2.70618
Retidr                       1.94724                 2.1933
Retthb                       -549.092               735.818
Retmyr                       9.85199                 7.1398
Retsgd                       -2.01639               5.45151
alpha(0)                     0.068975               0.011935
alpha(1)                     0.202876              0.0214085
beta(1)                      0.746174              0.0239392
Mean dependent var           0.017888
Log-likelihood              -3541.332
Schwarz criterion            7193.112

                                z                   p-value

Const                         1.6349                0.10208
retfSSTI Index                0.1131                0.90993
retjciindex                  -0.7216                0.47053
Retset                       -0.0456                0.96359
retfbmklci                   -0.5832                0.55979
Retpeso                      -11.2816             <0.00001 ***
Retidr                        0.8878                0.37464
Retthb                       -0.7462                0.45553
Retmyr                        1.3799                0.16763
Retsgd                       -0.3699                0.71147
alpha(0)                      5.7792              <0.00001 ***
alpha(1)                      9.4764              <0.00001 ***
beta(1)                      31.1695              <0.00001 ***
Mean dependent var      S.D. dependent var          0.999071
Log-likelihood           Akaike criterion           7110.664
Schwarz criterion          Hannan-Quinn             7140.498

Unconditional error variance = 1.35378

Table 13: Model 10: GARCH, using observations
2000/01/05-2010/03/26 (T = 2668)
Dependent variable: retjciindex
Standard errors based on Hessian

                           Coefficient             Std. Error

const                       0.00110156            0.000494091
retfSSTI Index              -0.0136058             0.0192688
retset                      0.0138238              0.0183159
retfbmklci                  0.0432274              0.0286173
retpeso                     -0.0389996             0.0556795
retidr                      -0.0404195             0.0337233
retthb                        4.9839                11.6034
retmyr                      -0.158373               0.109879
retsgd                      -0.130468              0.0840113
retPSEi                     -0.0222121              0.018222
alpha(0)                   1.50416e-05            2.75839e-06
alpha(1)                     0.141404              0.0176693
beta(1)                      0.793987              0.0243974
Mean dependent var           0.000636
Log-likelihood               7638.708
Schwarz criterion           -15166.97

                                z                   p-value

const                         2.2295               0.02578 **
retfSSTI Index               -0.7061                0.48012
retset                        0.7547                0.45040
retfbmklci                    1.5105                0.13091
retpeso                      -0.7004                0.48366
retidr                       -1.1986                0.23070
retthb                        0.4295                0.66754
retmyr                       -1.4413                0.14949
retsgd                       -1.5530                0.12043
retPSEi                      -1.2190                0.22285
alpha(0)                      5.4530              <0.00001 ***
alpha(1)                      8.0028              <0.00001 ***
beta(1)                      32.5440              <0.00001 ***
Mean dependent var      S.D. dependent var          0.015242
Log-likelihood           Akaike criterion          -15249.42
Schwarz criterion          Hannan-Quinn            -15219.58

Unconditional error variance = 0.000232811

Table 14: Model 11: GARCH, using observations
2000/01/05-2010/03/26 (T = 2668)
Dependent variable: retset
Standard errors based on Hessian

                           Coefficient             Std. Error

const                       0.00112611            0.000507964
retfSSTI Index              0.0126462               0.020667
retfbmklci                  0.0462272               0.029768
retpeso                     0.0686826              0.0636375
retidr                      -0.020453              0.0368582
retthb                       -5.33834               11.9532
retmyr                       0.193123               0.118983
retsgd                     1.95525e-05             0.0877279
retPSEi                    -0.00648415             0.0189839
retjciindex                 0.0229849              0.0197718
alpha(0)                   2.04548e-05            3.49307e-06
alpha(1)                     0.128036              0.0191421
beta(1)                      0.780403              0.0291884
Mean dependent var           0.000289
Log-likelihood               7615.112
Schwarz criterion           -15119.78

                                z                   p-value

const                         2.2169               0.02663 **
retfSSTI Index                0.6119                0.54060
retfbmklci                    1.5529                0.12044
retpeso                       1.0793                0.28046
retidr                       -0.5549                0.57896
retthb                       -0.4466                0.65516
retmyr                        1.6231                0.10456
retsgd                        0.0002                0.99982
retPSEi                      -0.3416                0.73268
retjciindex                   1.1625                0.24503
alpha(0)                      5.8558              <0.00001 ***
alpha(1)                      6.6887              <0.00001 ***
beta(1)                      26.7367              <0.00001 ***
Mean dependent var      S.D. dependent var          0.015139
Log-likelihood           Akaike criterion          -15202.22
Schwarz criterion          Hannan-Quinn            -15172.39

Unconditional error variance = 0.0002234

Table 15: Model 12: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retpsei
Standard errors based on Hessian

                           Coefficient             Std. Error

Const                      0.000981244            0.000447434
Retjci                      -0.0271703             0.0246361
Retset                     -0.00408886             0.0313633
retfbmklc_inde              0.0731688              0.0638906
Retfssti                    -0.0104083             0.0293733
Retidr                      0.0611328              0.0614346
Retmyr                      -0.215454               0.111388
Retsgd                      0.0682212               0.123933
Retpeso                      0.158323               0.107717
Retthb                      -0.0623793              0.143085
alpha(0)                   9.10178e-06            3.96753e-06
alpha(1)                     0.150747              0.0319415
beta(1)                      0.819348              0.0388831
Mean dependent var           0.000247
Log-likelihood               2360.239
Schwarz criterion           -4626.310

                                z                   p-value

Const                         2.1931               0.02830 **
Retjci                       -1.1029                0.27009
Retset                       -0.1304                0.89627
retfbmklc_inde                1.1452                0.25212
Retfssti                     -0.3543                0.72308
Retidr                        0.9951                0.31969
Retmyr                       -1.9343               0.05308 *
Retsgd                        0.5505                0.58200
Retpeso                       1.4698                0.14161
Retthb                       -0.4360                0.66287
alpha(0)                      2.2941               0.02179 **
alpha(1)                      4.7195              <0.00001 ***
beta(1)                      21.0721              <0.00001 ***
Mean dependent var      S.D. dependent var          0.016347
Log-likelihood           Akaike criterion          -4692.477
Schwarz criterion          Hannan-Quinn            -4667.109

Unconditional error variance = 0.000304357

Table 16: Model 13: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retjci
Standard errors based on Hessian

                           Coefficient             Std. Error

Const                       0.00119088            0.000506397
Retset                      0.0259006              0.0355295
retfbmklci_inde             0.0556889              0.0560895
Retfssti                    0.00152385             0.0296245
Retidr                      0.00719751             0.0712532
Retmyr                      -0.247524               0.122619
Retsgd                       0.243843               0.14483
Retpeso                     0.0816309               0.106845
Retthb                      -0.0707309              0.17237
Retpsei                     -0.0254544             0.0334898
alpha(0)                   1.29768e-05            3.87661e-06
alpha(1)                     0.14556               0.0263961
beta(1)                      0.817742              0.0290423
Mean dependent var           0.000554
Log-likelihood               2262.394
Schwarz criterion           -4430.622

                                z                   p-value

Const                         2.3517               0.01869 **
Retset                        0.7290                0.46601
retfbmklci_inde               0.9929                0.32078
Retfssti                      0.0514                0.95898
Retidr                        0.1010                0.91954
Retmyr                       -2.0186               0.04352 **
Retsgd                        1.6837               0.09225 *
Retpeso                       0.7640                0.44486
Retthb                       -0.4103                0.68155
Retpsei                      -0.7601                0.44722
alpha(0)                      3.3475              0.00082 ***
alpha(1)                      5.5144              <0.00001 ***
beta(1)                      28.1569              <0.00001 ***
Mean dependent var      S.D. dependent var          0.018595
Log-likelihood           Akaike criterion          -4496.789
Schwarz criterion          Hannan-Quinn            -4471.420

Unconditional error variance = 0.000353605

Table 17: Model 14: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retset
Standard errors based on Hessian

                           Coefficient             Std. Error

const                       0.00125671            0.000441034
retfbmklc_inde              -0.0220331             0.0482013
retfssti                   -0.00241563              0.028872
retidr                      0.0289232              0.0725929
retmyr                       0.200491               0.114925
retsgd                      -0.0535125              0.136243
retpeso                     -0.213458               0.102941
retthb                      -0.230843               0.144696
retpsei                     0.0100106              0.0304021
retjci                       0.018723              0.0282761
alpha(0)                   5.13683e-06            2.02241e-06
alpha(1)                     0.119977              0.0208121
beta(1)                      0.863977              0.0205942
Mean dependent var           0.000319
Log-likelihood               2357.312
Schwarz criterion           -4620.458

                                Z                   p-value

const                         2.8495              0.00438 ***
retfbmklc_inde               -0.4571                0.64760
retfssti                     -0.0837                0.93332
retidr                        0.3984                0.69031
retmyr                        1.7445               0.08106 *
retsgd                       -0.3928                0.69449
retpeso                      -2.0736               0.03812 **
retthb                       -1.5954                0.11063
retpsei                       0.3293                0.74195
retjci                        0.6621                0.50788
alpha(0)                      2.5400               0.01109 **
alpha(1)                      5.7648              <0.00001 ***
beta(1)                      41.9524              <0.00001 ***
Mean dependent var      S.D. dependent var          0.016576
Log-likelihood           Akaike criterion          -4686.625
Schwarz criterion          Hannan-Quinn            -4661.257

Unconditional error variance = 0.000320141

Table 18: Model 15: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retfbmklci_inde
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      0.000758725            0.000252621
retfssti                    0.0113419              0.0168251
retidr                      0.0254291              0.0365779
retmyr                      -0.0187246             0.0607504
retsgd                      0.0460012              0.0720567
retpeso                     -0.187645              0.0627862
retthb                      -0.0611178             0.0836623
retpsei                     0.0289376              0.0199808
retjci                      0.0126959              0.0158835
retset                      -0.0129801             0.0181719
alpha(0)                   1.70056e-06            7.68163e-07
alpha(1)                     0.174617               0.028838
beta(1)                      0.825383              0.0272578
Mean dependent var           0.000107
Log-likelihood               2785.127
Schwarz criterion           -5476.086

                                Z                   p-value

const                         3.0034              0.00267 ***
retfssti                      0.6741                0.50024
retidr                        0.6952                0.48693
retmyr                       -0.3082                0.75791
retsgd                        0.6384                0.52321
retpeso                      -2.9886              0.00280 ***
retthb                       -0.7305                0.46507
retpsei                       1.4483                0.14754
retjci                        0.7993                0.42411
retset                       -0.7143                0.47504
alpha(0)                      2.2138               0.02684 **
alpha(1)                      6.0551              <0.00001 ***
beta(1)                      30.2806              <0.00001 ***
Mean dependent var      S.D. dependent var          0.010085
Log-likelihood           Akaike criterion          -5542.253
Schwarz criterion          Hannan-Quinn            -5516.885

Unconditional error variance = 1.84212e+007

Table 19: Model 16: : GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retfssti
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      0.000500276            0.000402032
retpsei                     -0.033138              0.0289111
retjci                      0.0289614              0.0210753
retidr                      0.00101502             0.0738605
retpeso                     -0.0630895             0.0864475
retthb                       -0.22987               0.117993
retmyr                       0.176334               0.105469
retsgd                      0.0108712               0.131793
retset                      0.00896206             0.0252185
retfbmklci_inde             0.0866703              0.0445245
alpha(0)                   2.35098e-06            1.30289e-06
alpha(1)                     0.13431               0.0230304
beta(1)                      0.863584              0.0209251
Mean dependent var          -0.000149
Log-likelihood               2373.436
Schwarz criterion           -4652.704

                                Z                   p-value

const                         1.2444                0.21336
retpsei                      -1.1462                0.25171
retjci                        1.3742                0.16938
retidr                        0.0137                0.98904
retpeso                      -0.7298                0.46551
retthb                       -1.9482               0.05140 *
retmyr                        1.6719               0.09454 *
retsgd                        0.0825                0.93426
retset                        0.3554                0.72231
retfbmklci_inde               1.9466               0.05159 *
alpha(0)                      1.8044               0.07116 *
alpha(1)                      5.8319              <0.00001 ***
beta(1)                      41.2703              <0.00001 ***
Mean dependent var      S.D. dependent var          0.016893
Log-likelihood           Akaike criterion          -4718.871
Schwarz criterion          Hannan-Quinn            -4693.503

Unconditional error variance = 0.00111672

Table 20: Model 17: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retpeso
Standard errors based on Hessian

                           Coefficient             Std. Error

Const                      -4.53241e-05           0.000144234
Retidr                      0.0108114              0.0201272
Retmyr                      0.0209637              0.0364225
Retsgd                      0.0149479              0.0386298
retthb                      -0.111454              0.0469043
retpsei                     0.00495771             0.00973482
retjci                      0.00926667             0.00805189
retset                      -0.0135335             0.00936735
retfbmklci_inde             -0.0322533             0.0151236
retfssti                   -0.00772692             0.00863735
alpha(0)                   6.28294e-07            3.02312e-07
alpha(1)                     0.103505              0.0253881
beta(1)                      0.867795              0.0306009
Mean dependent var          -0.000076
Log-likelihood               3342.952
Schwarz criterion           -6591.736

                                z                   p-value

Const                        -0.3142                0.75334
Retidr                        0.5372                0.59116
Retmyr                        0.5756                0.56491
Retsgd                        0.3870                0.69879
retthb                       -2.3762               0.01749 **
retpsei                       0.5093                0.61056
retjci                        1.1509                0.24979
retset                       -1.4447                0.14853
retfbmklci_inde              -2.1326               0.03295 **
retfssti                     -0.8946                0.37100
alpha(0)                      2.0783               0.03768 **
alpha(1)                      4.0769              0.00005 ***
beta(1)                      28.3585              <0.00001 ***
Mean dependent var      S.D. dependent var          0.004605
Log-likelihood           Akaike criterion          -6657.903
Schwarz criterion          Hannan-Quinn            -6632.535

Unconditional error variance = 2.1892e-005

Table 21: Model 18: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retidr
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      -5.24677e-05           0.000139118
retmyr                       0.184379              0.0450983
retsgd                      -0.0168322             0.0449055
retthb                      0.00859123             0.0315043
retpsei                    -0.00255088             0.00691328
retjci                     -0.00785515             0.00694385
retset                      0.00661844             0.00652647
retfbmklc_inde             -0.00176059             0.0120432
retfssti                    0.0155717              0.00661603
Retpeso                     0.0328715              0.0265139
alpha(0)                   5.78678e-07            1.69907e-07
alpha(1)                     0.149061              0.0290481
beta(1)                      0.846949               0.025351
Mean dependent var          -0.000013
Log-likelihood               3243.926
Schwarz criterion           -6393.684

                                Z                   p-value

const                        -0.3771                0.70607
retmyr                        4.0884              0.00004 ***
retsgd                       -0.3748                0.70778
retthb                        0.2727                0.78508
retpsei                      -0.3690                0.71214
retjci                       -1.1312                0.25796
retset                        1.0141                0.31054
retfbmklc_inde               -0.1462                0.88377
retfssti                      2.3536               0.01859 **
Retpeso                       1.2398                0.21506
alpha(0)                      3.4058              0.00066 ***
alpha(1)                      5.1315              <0.00001 ***
beta(1)                      33.4089              <0.00001 ***
Mean dependent var      S.D. dependent var          0.006830
Log-likelihood           Akaike criterion          -6459.851
Schwarz criterion          Hannan-Quinn            -6434.483

Unconditional error variance = 0.000145035

Table 22: Model 19: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retthb
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      -8.62565e-05           6.80432e-05
retmyr                      0.0016322              0.0185389
retsgd                       0.238865              0.0258216
retpsei                    -0.00757537             0.00522802
retjci                     -0.00165233             0.00440304
retset                      -0.0121791             0.00494023
retfbmklc_inde             -0.00918824             0.00676398
retfssti                   -0.00685871             0.00439909
retpeso                     -0.0548585             0.0172497
retidr                     -0.00316257              0.012035
alpha(0)                   3.28729e-07            1.08083e-07
alpha(1)                     0.262849              0.0452133
beta(1)                      0.737151              0.0403357
Mean dependent var          -0.000124
Log-likelihood               3794.682
Schwarz criterion           -7495.197

                                z                   p-value

const                        -1.2677                0.20491
retmyr                        0.0880                0.92984
retsgd                        9.2506              <0.00001 ***
retpsei                      -1.4490                0.14734
retjci                       -0.3753                0.70746
retset                       -2.4653               0.01369 **
retfbmklc_inde               -1.3584                0.17433
retfssti                     -1.5591                0.11897
retpeso                      -3.1803              0.00147 ***
retidr                       -0.2628                0.79272
alpha(0)                      3.0415              0.00235 ***
alpha(1)                      5.8135              <0.00001 ***
beta(1)                      18.2754              <0.00001 ***
Mean dependent var      S.D. dependent var          0.003580
Log-likelihood           Akaike criterion          -7561.364
Schwarz criterion          Hannan-Quinn            -7535.996

Unconditional error variance = 1.92831e+006

Table 23: Model 20: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retmyr
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      -0.000193129           0.000112758
retsgd                      0.0677273              0.0384396
retpsei                     0.00548671             0.0067486
retjci                     -5.19301e-05            0.00634895
retset                     -0.000655657            0.00756114
retfbmklci_inde            -0.00225435             0.0109943
retfssti                   -0.00874604             0.00662814
retpeso                     0.0171017              0.0234216
retidr                      0.0770925              0.0235447
retthb                      0.0108206              0.0254415
alpha(0)                   3.07789e-07             1.4771e-07
alpha(1)                     0.124471              0.0322547
beta(1)                      0.868968               0.02823
Mean dependent var          -0.000097
Log-likelihood               3490.296
Schwarz criterion           -6886.425

                                z                   p-value

const                        -1.7128               0.08675 *
retsgd                        1.7619               0.07808 *
retpsei                       0.8130                0.41621
retjci                       -0.0082                0.99347
retset                       -0.0867                0.93090
retfbmklci_inde              -0.2050                0.83753
retfssti                     -1.3195                0.18699
retpeso                       0.7302                0.46529
retidr                        3.2743              0.00106 ***
retthb                        0.4253                0.67061
alpha(0)                      2.0837               0.03718 **
alpha(1)                      3.8590              0.00011 ***
beta(1)                      30.7817              <0.00001 ***
Mean dependent var      S.D. dependent var          0.003919
Log-likelihood           Akaike criterion          -6952.592
Schwarz criterion          Hannan-Quinn            -6927.224

Unconditional error variance = 4.69133e-005

Table 24: Model 21: GARCH, using observations
2007/06/01-2010/08/11 (T = 834)
Dependent variable: retsgd
Standard errors based on Hessian

                           Coefficient             Std. Error

const                      -0.000118557           9.95123e-05
retpsei                    -0.000798756            0.00604538
retjci                      0.00982239             0.00544697
retset                      0.00354235             0.00669941
retfbmklc_inde             -0.00737134             0.00984218
retfssti                   -0.00146166             0.00618743
retpeso                    -0.000992164            0.0210715
retidr                      -0.013964              0.0195587
retthb                       0.129528              0.0310012
retmyr                      0.0528361              0.0293694
alpha(0)                   7.40557e-08            4.56461e-08
alpha(1)                    0.0495539              0.0110603
beta(1)                      0.944847              0.0121377
Mean dependent var          -0.000112
Log-likelihood               3611.545
Schwarz criterion           -7128.923

                                z                   p-value

const                        -1.1914                0.23351
retpsei                      -0.1321                0.89488
retjci                        1.8033               0.07134 *
retset                        0.5288                0.59698
retfbmklc_inde               -0.7490                0.45389
retfssti                     -0.2362                0.81325
retpeso                      -0.0471                0.96244
retidr                       -0.7140                0.47526
retthb                        4.1782              0.00003 ***
retmyr                        1.7990               0.07202 *
alpha(0)                      1.6224                0.10472
alpha(1)                      4.4803              <0.00001 ***
beta(1)                      77.8441              <0.00001 ***
Mean dependent var      S.D. dependent var          0.003683
Log-likelihood           Akaike criterion          -7195.090
Schwarz criterion          Hannan-Quinn            -7169.721

Unconditional error variance = 1.32275e-005
COPYRIGHT 2011 The DreamCatchers Group, LLC
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

 Reader Opinion

Title:

Comment:



 

Article Details
Printer friendly Cite/link Email Feedback
Author:Kabigting, Leila C.; Hapitan, Rene B.
Publication:Journal of International Business Research
Geographic Code:4E
Date:Dec 1, 2011
Words:7351
Previous Article:A comparison of the International Financial Reporting Standards (IFRS) and Generally Accepted Accounting Principles (GAAP) for Small and Medium-sized...
Next Article:Remittances as avenue for encouraging household entrepreneurial activities.
Topics:

Terms of use | Copyright © 2014 Farlex, Inc. | Feedback | For webmasters