The Good, the Bad, and the Ambiguous: The Aggregate Stock Market Dynamics around Macroeconomic News

Using a representative agent model in which the investor is averse to ambiguity (Knightian uncertainty) and sees an ambiguous piece of news about the fundamental value of a risky asset, I show a number of predictions for the dynamics of stocks around news: Stocks respond more strongly to bad news than to good news, respond positively to neutral news, and increase on average through news. In times of high ambiguity, the magnitudes of each effect is larger, and the volatility of stocks around news changes in a predictable manner as well. I provide empirical evidence consistent with the model by analyzing the high-frequency behavior of the aggregate stock market around macroeconomic news announcements from November, 1997 to March, 2014. The model helps to understand features of the data that challenge existing frameworks; e.g., the findings that the stock market reacts especially strongly to bad news versus good news during crisis periods and that about 1/3 of equity returns in the 17 year sample accrues in the 10 minutes around the release of macroeconomic data. In addition to providing evidence for the role of ambiguity in financial markets generally and in how financial assets reflect macroeconomic shocks specifically, the empirical results also have implications for the behavior of investors. Investors treat bad news as more relevant in bad times than in good times but treat good news the same in good and bad times. ∗Harvard University. [email protected]. I thank Geert Bekaert, John Campbell, Itamar Drechsler, Sam Hanson, Marie Hoerova, Erik Stafford, Jeremy Stein, Luis Viceira, and seminar participants at Harvard for helpful discussions and suggestions.