AbstractWhile Eugene Fama has repeatedly expressed his discontent with the notion of an “irrational bubble,” he has never publicly expressed his opinion on “rational bubbles.” On empirical grounds Fama rejects bubbles by referring to the lack of reliable evidence that price declines are predictable. However, this argument cannot be used to rule out rational bubbles because such bubbles do not necessarily imply return predictability, and return predictability of the kind documented by Fama does not rule out rational bubbles. On data samples that include the 1990s, there is evidence of an explosive component in stock market valuation ratios, consistent with a rational bubble.