本文利用臺灣中央銀行公布的存、放款利率資料，分析代表貨幣政策改變之隔夜拆款利率變化，對存、放款利率之零售利率的利率轉嫁、立即衝擊以及零售利率完成反應政策利率影響效果所需時間等議題。實證分析顯示，存款利率之利率轉嫁為不完整的，而放款利率的利率轉嫁則為完整或過度調整。而不論存款利率或放款利率，零售利率反應政策利率變化具有延遲、落後的現象。且存款利率可以在低於兩個月的時間內，完成反應政策利率變化的影響。這些實證結論對線性共整模型以及非線性門檻共整模型而言，皆可以成立，顯示文中的統計分析具有穩健性。此外，存款利率以及放款利率的利率轉嫁效果，也會因政策利率漲跌不同而出現易跌難漲的變化，此分別符合勾結假說以及負面顧客反應假說或資訊不對稱的觀點。

This research employs data published by the Taiwan Central Bank to address the degree of pass-through, immediate impact and adjustment speed of retail bank interest rates, including deposit rates and lending rates, in response to changes in the offical rate, represented by the overnight interest rate of the interbank call-loan market. The empirical evidence from linear and threshold cointegration models suggests that the pass-through from the offical rate to de-posit rates is incomplete, and that to lending rates is complete or overshooting in the long run. For both deposit rates and lending rates, the immediate impact from changes in the offical rate is less than the magnitude of long run pass-through, indicating sluggish adjustment. And the mean adjustment lag of a complete pass-through for deposit rates is less than two months. Besides, this research also examines whether interest rate rigidity is different when the offical rate is increasing or decreasing. It is concluded that both deposit rates and lending rates are more rigid when they are rising. Evidence of rigidity of deposit rates increases supports the collusive hypothesis and the rigidity of lending rates increases coincides with the hypothesis of adverse customer reaction and asymmetric information.