Modeling, Estimation and Application of STVEC-GARCH model:The Relationships among Stock Price ,House Price and Interest Rate in Taiwan

• Main Authors: Lin,Tsung-Shian, 林聰賢
• Other Authors: 劉祥熹
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/66275858242283750972

碩士 === 國立臺北大學 === 統計學系 === 95 === The purpose of this study is to detect the relationships among stock prices, house prices and interest rate of Taiwan. The STVEC-GARCH model is applied to examine these interactive relationships by using the sample period begins from Jan.1986 to Mar.2007.The main empirical results are as follows: 1. After differencing stock prices, house prices and interest rate, the three time series become stationary time series. There is a cointegrating relationship among the three time series based on the Johansen cointegration test. 2. Based on the result of linearity test, it can be found that these three time series of returns possess the nonlinearity characteristics. It furtherly indicates that the three time series of returns can be fitted by non-linear model. The trend and tendency of stock returns can be depicted by ESTVECM process while the trend and tendency of house returns and interest rate returns can be explained by LSTVECM process. 3. According to the estimation result of the mean equation of the three time series of returns, the causal relation is as follows: Before the state transited, there are not significant interactions among these three time series. After the state transited, house prices and stock prices lead interest rate to change and house prices lead stock prices to change. 4. According to the estimation results of the variance equation of the three time series of returns, the volatility of interest rate is influenced by the volatility of stock prices. This may imply that the investors need to consider the volatility of the stock market in order to strengthen risk management during the time they borrow the money. 5. With respect to the goodness-of-fit of the model we estimated, the result of the linearity test indicates that the three time series of returns present non-linear tendency. Furtherly, according to the result of the correlation test of error term series, it can be found that the returns of stock prices and interest rate have ARCH effects. Therefore, this study confirms that the STVEC-GARCH model can explain the behavior of these three time series and their interactions.