本研究之目的在探討臺灣短期票券持有報酬之機率分配模型。報酬之機率分配是極為重要的基礎研究，對報酬率之預測與市場投資行為之瞭解均有密切關係，但國内研究大多以股票報酬為主，本研究乃針對商業本票與銀行承兌匯票，利用對數常態機率分配的特性，探討持有報酬之機率分配模型。結果發現（1）設限之對數常態分配模型較非設限之對數常態分配模型更適用。（2）設限分配模型所形成的報酬分配具有持續穩定性。（3）經適度修訂後設限分配模型的上限可用模擬法做近似點估計，解決模型實際應用的問題，有助於對票券投資報酬分配之各級動差做進一步的研究，以了解其對投資行為可能之影響。

The purpose of this stud is to explore the holding return distribution model for short-term discount bonds. Return distributions are very important in forcasting expected returns and in building valuation models. Utilizing log-normal distribution characteristics, we study the single period holding return distributions for commercial papers and bankers' acceptances. It is found that (1) the bounded distribution model (BDM) fits better than the unbounded distribution model (UDM), (2) BDM is persistent, and (3) the upper bound can be estimated by an approximation method under a modified BDM. This is useful for further analysis of return distributions with different types of bonds of with different holding periods.