Many economists believe that the stock market plays an important role in efficiently allocating capital to its most productive uses. This standard story of the stock market was called into question by events in the late 1990s, when some observers believed that stock market overvaluation - or a bubble - led to overinvestment. Both the standard and overinvestment stories involve discount rates and, to differentiate between the two stories, this paper examines the discount rates used by firms in making their investment decisions. We use a revealed preference approach that relies on the pattern of investment spending - combined with investment theory - to estimate the discount rates used by managers. The standard story predicts that firms with high stock prices and good investment opportunities should have discount rates that do not differ systematically from the risk-adjusted market rate. The overinvestment story predicts that firms with high stock prices and poor investment opportunities should have discount rates consistently below the market rate. Based on a panel dataset of over 50,000 firm-year observations, we find support for both stories. The behavior of high stock price firms with good measured investment opportunities is best described by the standard story, while the overinvestment story provides the most appropriate interpretation of the behavior of high stock price firms with poor investment opportunities. Firms in this latter category accumulate between 15.1% and 45.2% too much capital. These estimates suggest that, even before they burst, bubbles adversely affect economic activity by misallocating capital.