Essay I: Management Earnings Guidance, Earnings Momentum and Price Momentum

This paper studies the impact of management earnings guidance on momentums. I find that management earnings guidance can reduce EC earnings momentum (where EC is seasonal earnings changes) and price momentum, but cannot reduce ES earnings momentum (where ES is actual earnings minus analyst forecast consensus). Further evidence suggests that information content rather than the presence of management earnings guidance affects its impact on momentums. Specifically, I find that management earnings guidance can increase rather than decrease earrings momentum when the bias in the guidance is high. Finally, I show that investors are responsive to the prior information contained in management earnings guidanc...[ Read more ]

Essay I: Management Earnings Guidance, Earnings Momentum and Price Momentum

This paper studies the impact of management earnings guidance on momentums. I find that management earnings guidance can reduce EC earnings momentum (where EC is seasonal earnings changes) and price momentum, but cannot reduce ES earnings momentum (where ES is actual earnings minus analyst forecast consensus). Further evidence suggests that information content rather than the presence of management earnings guidance affects its impact on momentums. Specifically, I find that management earnings guidance can increase rather than decrease earrings momentum when the bias in the guidance is high. Finally, I show that investors are responsive to the prior information contained in management earnings guidance. The evidence in this paper supports information diffusion theory in explaining earnings and price momentum and suggests that informative disclosure can accelerate information diffusion process.

Essay II: How Do Heterogeneous Priors Generate Disagreement

This study uses an innovative setting to study how heterogeneous beliefs affect information incorporation. I pair analysts who have identical earnings forecasts for quarter t+1 before earnings announcements of quarter t, to control their prior beliefs about earnings of quarter t+1. I define the order of paired analysts’ latest forecasts for quarter t as prior order and the order of their earliest forecasts for quarter t+1 after earnings announcements of quarter t as posterior order. I find that posterior order is more likely to be consistent with rather than flip over prior order. This finding cannot be explained by existing literature in finance, such as Kandel and Pearson (1995) and Hong and Stein (2007). One possible explanation for my finding is that analysts’ information processing is biased by their prior beliefs.