DescriptionFinance and accounting research has recently focused on extracting the tone or sentiment of a document by using positive or negative words/phrases in the document. The first essay of this dissertation exploits the information content of qualitative data in addition to quantitative signals in selecting optimal portfolios. Using optimization techniques developed by Brandt, Santa-Clara, and Valkonov[2009], this essay shows that significantly higher returns can be obtained combining quantitative and qualitative data obtained from firms’ Management Discussion and Analysis sections of their Form 10-Q (10-K) SEC filings than just using quantitative signals. The second essay uses option market characteristics to examine the two leading explanations for the Post-Earnings-Announcement Drift (PEAD) anomaly. PEAD points towards an inexplicable inefficiency in the equity markets where traders seem to ignore the autocorrelations in extreme earnings surprises across adjacent quarters. By contrast, there is mounting evidence that option markets are very efficient. If so, there should be no PEAD like anomaly in the pricing of equity options. This essay tests this using a straddle strategy around earnings announcements and its empirical results indicate that option traders already incorporate the autocorrelation in extreme earnings surprises in option prices. It also uses the change in implied volatilities obtained from options prices immediately before and after the earnings announcements as risk metric to examine the risk premium hypothesis of PEAD. However, its findings favor the competing underreaction hypothesis, which assumes equity traders do not completely utilize the autocorrelations of earnings surprises. The third essay examines the potential explanations for the observed decline in the value relevance of earnings over the years by exploring the time-series change of information transfer. Specifically, it examines how the importance of information transfer itself changes over time and the time-series change in value relevance of earnings by industry. It shows that the decline in usefulness of earnings is not significant in all industries, although the decline is significant on average. It also indicates that the time-series change in the magnitude of information transfer is insignificant on average and for most industries, after controlling for the decline in the value relevance of earnings over time.