Abstract
In asset pricing models, the uncertainty surrounding firm fundamentals plays a central role, driving expected returns, volatility, and valuation ratios. In this paper, we extract estimates of the uncertainty embedded in earnings announcements using option prices. To do this, we take seriously the fact that the timing of earnings announcements, although not the response of equity prices, is known in advance. We develop a no-arbitrage option pricing model incorporating jumps on earnings announcement dates. We estimate the uncertainty in a simple extension of the Black-Scholes model and in a more complicated stochastic volatility model. The uncertainty is large, has important pricing implications and interesting time-variation. Adding jumps on earnings announcement dates drastically reduces the model pricing errors. Not surprisingly, the estimates of the earnings uncertainty increase in periods of market stress, such as 2000 and 2001.