UC San Diego

There are probably only few other questions as central to economics as the question "How do market prices react to news?". The reaction of prices to new information has interested and puzzled economists since the early years of the field. This thesis addresses several dimensions of this basic question for the specific case of the stock market. This thesis develops new theoretical models about the reaction of stock prices to macroeconomic news using new mathematical tools and techniques and tests the implications of these and other models using new data sets on macroeconomic news. In the first chapter of my thesis, A Rational Model of Underreaction: The Effect of Macroeconomic News, I analyze the long-term effects of macroeconomic news on the return dynamics. I develop a dynamic general equilibrium asset pricing model where macroeconomic news is an additional state variable. In this framework, I show that the underreaction of stock prices to news is consistent with a rational expectations model rather than a behavioral specification as suggested by recent literature. Furthermore, I show that the reaction of the stock market to news depends on the state of the economy. The empirical results suggest that the stock market underreacts to news about the nominal U.S. Gross Domestic Product. In the second chapter of my thesis, Risk and Return Reaction of the Stock Market to Public Announcements about Fundamentals: Theory and Evidence, I analyze the short-term effects of public macroeconomic announcements about fundamentals on daily returns. This chapter presents new theoretical and empirical results on the effect of public announcements on the stock market. I develop a dynamic general equilibrium asset pricing model where investors learn about the unobserved state of the economy through dividend realizations and periodic public announcements. The main implications of my model can be summarized as follows: 1. If investors are more risk averse than log utility, returns react negatively to a positive unanticipated news in the announcement. 2. Returns react asymmetrically to the unanticipated news on announcement days. 3. The effect of the unanticipated news depends on the state of the economy which is revealed by the announcement. 4. On announcement days, the conditional volatility of returns is a decreasing function of the investors' uncertainty about the announcement. In other words, the higher the degree of uncertainty resolved on announcement days, the smaller the conditional volatility will be. Using real-time data and survey expectations, I develop measures of unanticipated news and uncertainty to test the implications of my theoretical model. I find that the implications of my model hold for the aggregate stock market returns on the U.S. Gross Domestic Product announcement days. In the last chapter of my thesis, I analyze the asymmetries in the reaction of returns on portfolios with different characteristics to the same macroeconomic news. The first empirical question addressed in this chapter is "Do the effects of macroeconomic news on stock returns differ across assets?". More specifically, I analyze whether stock returns on a portfolio of firms with high market capitalization and/or high book equity-to -market equity ratio react differently than stock returns on a portfolio of firms with low market capitalization and /or low book equity-to-market equity. I find that returns on a portfolio of firms with high market capitalization (large firms) and book-to-market ratio (value firms) react stronger (in magnitude) to macroeconomic news than returns on a portfolio of firms with low market capitalization (small firms) and book-to-market ratio (growth firms). I also find that firms with high market capitalization and low book-to-market ratio are sensitive to fewer macroeconomic variables than firms with low market capitalization and high book-to-market ratio. Having documented these asymmetries in the reaction of firms with different characteristics, I analyze the possible sources of these asymmetries by decomposing the effect of news into three parts, its effect through the market's discount rate component, its effect through the market's cash flow component and its direct effect. First of all, I find that the news does not have any direct effect on stock returns when one controls for the market's discount rate and cash flow components suggesting that the reaction is generally captured by the two market components. Furthermore, I find that the differential reaction across firms with different characteristics is generally due to the differential sensitivity to the market's cash flow component