The correlation between Mortgage-Backed Securities and capital market─Empirical evidence from the U.S.A. capital market

• Main Authors: Huei-Chuan Chu, 朱惠娟
• Other Authors: Ging-Chung Lin
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/13266776013908834955

碩士 === 淡江大學 === 財務金融學系碩士在職專班 === 91 === Abstract： Due to the depression and the rise of unemployment rate, the nonperforming loan ratio has been raised and the liquidity of asset and asset-liabilities structure of financial institutions has been affected subsequently. To strengthen the capital structure of financial institutions, the government has enacted and promulgated the Financial Asset Securitization Law in 2002. The financial institutions can issue the mortgage-backed securities (called MBS) to increase the channels of fundraisings and the capital adequacy ratio and to improve the asset-liabilities structure. The objective of this paper is to investigate whether investors should include MBS in their investment portfolio to diversify investment risks while conducting asset allocation after issuing of MBS in Taiwan. Since MBS in Taiwan has not been issued yet, this paper takes the following data as samples to investigate the correlation between MBS yield rate and other variances (treasury bonds yield rate, corporate bonds yield rate and Dow Jones stock index) of US capital market from May 1995 to April 2002. It applies the statistics and time series method to explore whether the cointegration exist among four variances. The empirical evidences are as follows: 1. Applying the Johansen Cointegration Test, no cointegration among four variances has been found, i.e. there is no long-term equilibrium relationship. Which indicates investors can invest in MBS, treasury bonds, corporate bonds and stock simultaneously to diversify investment risks. Thus, investors should include MBS in their investment portfolio while conducting asset allocation. 2. Applying VAR to the Granger Causality Test, a negative correlation has been found between MBS yield rate and stock index. Moreover, the capital market yield rate has a leading relationship to stock index. 3. Using VAR to the Impulse Response Function, we have found that investors can invest in MBS and corporate bonds or in stock, MBS and treasury bonds at the same time to diversify investment risks. 4. Analyzing the risk and return of investment portfolio, we have found that the return rate of stock is the highest in capital market, but its risk is also the greatest. However, the return rate of MBS is only second to that of stock, but its risk is much lower than that of stock and corporate bonds. There is a negative relationship between MBS and Stock portfolio, so that MBS can diversify the investment risks. In addition, this study also resorts US MBS issuing quantity, MBS issuing rate, GDP growth rate and M2 growth rate from May 1995 to December 2001 as data to analyze if any interactive, and leading, lagging or feedback relationships exist among macroeconomic variances. Applying VECM to the Granger Causality Test, it has been found the MBS issuing quantity, MBS issuing rate and GDP growth rate significantly lead the growth rate of M2. Using VECM to the Impulse Response Function, it has been found the impact of GDP growth rate, MBS issuing rate and M2 growth rate has influenced MBS issuing quantity, and the growth rate of M2 has been found to be the greatest variance of macroeconomic affecting MBS issuing quantity. We may thus conclude that the influence of government monetary policy on MBS issuing quantity is greater than MBS issuing rate and any other macroeconomic variances.