過去的實證研究指出，在討論當期的會計盈餘穩健性時，研究者必須要控制期初會計盈餘穩健性的水準。具體而言，期初會計盈餘穩健性與本期會計盈餘穩健性的負向關係，會稀釋我們對當期會計盈餘穩健性的解釋效果。基於這個理由，本研究以期初市價對淨值比做為期初會計盈餘穩健性的代理變數，去檢視我國1999年至2004年會計盈餘的穩健性以及探討盈餘穩健性與外資法人持股比例之關聯性。穩健原則之定義，係以 Basu(1997)模型做為衡量穩健性的指標。研究結果發現，我國企業近年來之會計盈餘存在穩健原則之特性，而期初市價對淨值比與盈餘穩健程度之間呈現顯著之負向相關。至於針對外資持股與盈餘穩健性之關聯性的分析，在控制期初市價對淨值比的影響之後，本研究未能找到外資持股比例之高低對盈餘穩健程度有顯著影響之證據。然而，在期初市價對淨值比最高的組別中，盈餘穩健程度與次期外資持股比例之間存在正向之關係。這個現象意味著對於期初盈餘穩健性很高的公司而言，其在當期之盈餘若相對較穩健，則次期之外資持股比例將會較高。但是，在本研究改以工具變數衡量外資持股比例之後，不論是當期或次期的外資持股比例，其與盈餘穩健程度之間則均未存在顯著之關聯性。彙總而言，利用1999年至2004年的資料，本研究發現，期初會計盈餘穩健性與本期會計盈餘穩健性有顯著的負向關係；無論有無控制期初會計盈餘穩健性，我國的會計盈餘均具有穩健性的特性；未能得到非常充份的證據去宣稱外資對於我國會計盈餘穩健性有系統性的影響。 / Prior empirical studies indicate that researchers have to control for the level of beginning-of-period earnings conservatism when discussing earnings conservatism in the current period. The negative association between beginning-of-period earnings conservatism and earnings conservatism in the current period will dilute the effect of our explanation of earnings conservatism in the current period. Using beginning-of-period price-to-book ratios to proxy for beginning-of-period earnings conservatism, this study examines earnings conservatism in Taiwan during the period from 1999 to 2004, and discusses the relationship between earnings conservatism and shareholding ratios of foreign institutional investors. The definition of earnings conservatism is based on Basu(1997). The empirical results show that earnings in Taiwan demonstrate the characteristic of earnings conservatism in recent years, and that beginning-of-period price-to-book ratios are negatively associated with the level of earnings conservatism. As for the analysis of the link between shareholding ratios of foreign institutional investors and earnings conservatism, this study fails to find the evidence that shareholding ratios of foreign institutional investors have significant impacts on earnings conservatism after controlling for the level of beginning-of-period price-to-book ratios. However, in the portfolio of the highest beginning-of-period price-to-book ratios, there is a positive relationship between earnings conservatism and shareholding ratios of foreign institutional investors in the next period. This result means that firms with the highest level of beginning-of-period earnings conservatism will have higher shareholding ratios of foreign institutional investors in the next period if their earnings in the current period are relatively more conservative. However, when this study uses the instrumental variable to measure shareholding ratios of foreign institutional investors, shareholding ratios of foreign institutional investors, whether in the current or next period, do not have a significant relationship with the level of earnings conservatism. In summary, using data from 1999 to 2004, this study finds that there is significantly negative association between beginning-of-period earnings conservatism and earnings conservatism in the current period, and that earnings in Taiwan demonstrate the characteristic of earnings conservatism whether controlling for the level of beginning-of-period earnings conservatism or not, but this study fails to obtain sufficient evidence to assert that foreign investment has systematic effects on earnings conservatism in Taiwan.