In the third essay, I study the information contents and impacts of stock split. To provide an alternative perspective to distinguish among several hypotheses proposed to explain the abnormal return associated with splits, I examine bond, EPS forecast, and realized earnings of split companies. The results show that split companies experience significant decrease in bond yield and significant increase in the EPS forecast, and realized earning after splits, which evidence is consistent with the signaling hypothesis.This dissertation contains three individual and internally related essays.In the second essay, the information contents and impacts of the Canadian S&P/TSX Composite Index revision on firm performance are studied. The empirical results show that added companies experience significant increases in their expected earnings following the index change, whose pattern is similar to that of addition to S&P 500. Whereas, the expected earnings for companies removed from S&P/TSX Composite, still show positive change following deletion, but after adjusting for matched benchmark firms' EPS forecast change, they become significantly negative. For realized earnings, removed companies don't have significant decrease, and added companies experience significant increase after May 2002, whose patterns are different from those of S&P 500.In the first essay, I study the information content contained in the S&P 500 and the impact of index revision on the firm performance and bond yields of concerned companies. The regression analysis is applied to EPS forecasts, realized earnings and bond yield. The empirical results show that analysts and investors draw positive (negative) information from the index addition (deletion) and adjust their expectations about firm's performance, and required rate of return on bond after the index revision.