This paper uses two different groups of firms, a group of innovative firms and a group of non innovative firms, in computer and software industry and compares them to estimate the stock market's valuation of innovation activities and examines how the stock price response to the innovation events. Two groups show different pattern of stock price during the high-tech boom period, while they have an identical pattern before and after the boom period. The observed different patterns are tested based on Tobin's q theory, and the results show that stock market valued significantly highly on innovation activity during 1996-1999. I employ simulation method to overcome the limitation of event study. This simulation method leads me to conclude that the usual event study method can misidentify the abnormal returns. I find the statistically significant cumulative abnormal returns before patent application events and patent grant announcement events. The information concentration, characteristics of innovation and previous cumulative abnormal returns affect the stock market's response to innovation events.