This dissertation is made of three distinct chapters. In the first chapter, we introduce a new measure of herding that allows for tracking dynamics of individual herding. Using a database of nearly 8 million trades executed between 1999 and 2006 by 87,373 individual investors, we show that individual herding is persistent over time and that past performance and the level of sophistication influence this behavior. We are also able to answer a question that was previously unaddressed in the literature: is herding profitable for investors? Our unique dataset reveals that the investors trading against the crowd tend to exhibit more extreme returns and poorer risk-adjusted performance than the herders. In the second chapter, we show that measuring the accuracy of a target price is not sufficient to assess its quality, because the forecast predictability (which depends on the stock return volatility and on the forecast horizon) is likely to vary across stocks and over time. We argue that the evidence of time persistent differences in analysts' target price accuracy, obtained in previous studies, cannot be interpreted as a proof of persistent differential abilities. Our analysis indicates that the persistence in accuracy is driven by persistence in stock return volatility. We introduce a measure of target price quality that considers both the forecast inaccuracy and the forecast predictability. Using elements from option-pricing theory, we provide a simple solution to the issue of estimating target price predictability. Our empirical analysis reveals that, when forecast predictability is taken into account, financial analysts do not exhibit significant persistent differential abilities to forecast future stock prices. In the third chapter, we show that experienced financial analysts tend to cover different firms than inexperienced analysts. Experienced analysts tend to follow blue chips (i.e., large, international, mature firms) while inexperienced analysts focus on small, young, growth-oriented firms. These differences in coverage decisions imply that inexperienced analysts issue target prices on firms for which stock returns are more volatile, and thus less predictable. As a consequence, the accuracy measure of target prices fails to evaluate differences in ability between experienced and inexperienced analysts. When taking into account these differences in coverage decisions, we still find that experienced analysts do a better job at forecasting stock prices. Our results on the influence of analysts' characteristics on target price quality are statistically significant but economically weak.
