แบบจำลองอัตราผลตอบแทนและความผันผวนของตลาดหลักทรัพย์ ตลาดพันธบัตรและตลาด อัตราแลกเปลี่ยนสำหรับประเทศไทยและประเทศญี่ปุ่น

Abstract

Portfolio manager invests in many markets or international market to diversify risk efficiently than investment in only one market or domestic market. Therefore, the paper would like to construct model include 3 markets, namely, stock market, bond market and exchange rate market. The study examine 2 countries, Thailand and Japan, by using multivariate volatility models, namely, Constant Conditional Correlation (CCC) model; Vector Autoregressive Moving Average Generalized Autoregressive Conditional Heteroskedasticity (VARMA-GARCH) model, which allow volatility spillover and VARMA-Asymmetric GARCH (VARMA-AGARCH) model, which allow both volatility spillover and asymmetric effect. The results showed that VARMA-AGARCH model is better than VARMA- GARCH model in variance equations of all markets except variance equation between Japan bond market and Thailand exchange rate market. Asymmetric effect in the variance equation between Japan bond market and Thailand exchange rate market is not significant, that mean the magnitude of effect of good and bad news to volatility are equal.