The Effects of Changes in the Level and Volatility of Interest Rate on Bank Stock Excess Returns----A GARCH-M Model

[[abstract]]受到金融技術的創新、國際市場整合及景氣波動的影響，利率的水準及利率波動之變化有愈加劇烈的跡象。而在討論金融相關議題方面，使用隨時間變動之自我相關異質變異數(ARCH)模型已凌駕傳統固定變異數迴歸模型，愈來愈受重視。 本論文以Elyasiani and Mansur(1998)的研究為基礎，向三個方向作延伸。第一，同時探討利率改變及利率波動對於銀行股票報酬波動及超額報酬的影響。再者，採用GARCH-M模型，摒棄線性，獨立，及固定條件變異數等現存研究中的不合理假設，允許「波動」隨著時間而變動，將超額報酬與報酬波動結合討論，並考慮景氣循環對波動可能造成的影響。第三，探討不同資產負債結構的銀行，其股票報酬受利率及利率波動的影響程度是否產生差異。本研究的樣本期間為1990年1月到1998年9月，資料型態為日資料。資料來源以台灣經濟新報資料庫(TEJ)為主。 實證結果指出：股票報酬波動確實隨著時間而變動，但其並不是銀行股票超額報酬的顯著決定因素。拒絕景氣循環與銀行股票報酬波動無關的虛無假設，顯示兩者有相關性存在，而景氣循環與股票超額報酬間的相關性呈現不顯著。 利率改變對於銀行股票超額報酬具有負向但不顯著的影響。對於兩類銀行而言，利率波動是其股票報酬波動及股票風險溢酬的一個重要決定因素。當利率的波動加大時，兩類銀行投資組合的股票報酬波動皆呈現顯著性的增加，但「借短貸長」投資組合風險溢酬的增加呈現不顯著，「借長貸短」投資組合的風險溢酬則沒有顯著的減少。 利率效果確實存在，其對於銀行股票超額報酬之產生過程是有影響的。但不同景氣循環下之利率波動不會造成銀行股票超額報酬不同的反應。而銀行資產負債結構這項因素在我們的討論中的影響並沒有顯著的結果出現。[[abstract]]The major issue addressed in this thesis is the effects of changes in the level and volatility of interest rate on bank stock excess returns. In recently years, interest rate has volatiled substantially due to several factors such as financial innovations, technological progress, international market integration, and business cycles in Taiwan as well as the other countries. We extend a recent study by Elyasiani and Mansur(1998), which employed the GARCH-M methodology to investigate this issue, in the following ways. First, we discuss the effects of levels and volatility of interest rate on bank stock returns at the same time. Second, we employ a GARCH-M methodology, as Elyasiani and Mansur(1998) did, to develop our empirical model which allows for volatility of stock returns to vary over time, relates the risk premia to the volatility, and considers the effects of business cycles. Third, we discuss whether the effects of interest rate and its volatility on stock returns differ under alternative asset-liabilit structures of selected banks. We use daily data from TEJ database. The sample period spans from January 1, 1990 to September 30, 1998. The empirical findings are as follows: Return volatility varies over time and is not a significant factor in determining portfolios'' excess returns. We find that business cycles are relevant to return volatility, however, the relationship between business cycles and stock excess returns is insignificant. The change in interest rate has an insignificant negative effect on excess returns. Interest rate volatility is one of the important factors in determining return volatility and risk premia. When interest rate becomes volatile, both portfolios'' return volatility increases significantly. Neither the increase of "borrow short and lend long" portfolio''s risk premium is significant, nor the decrease of "borrow long and lend short" portfolio''s risk premium. As a result, the asset-liability structures of banks play no role in discussing this issue. Our empirical finding suggests that the interest rate effects exist, which means that interest rate affects bank stock return generating process significantly