Essays on Empirical Asset Pricing

Prasad Hegde, Purdue University

Abstract

This dissertation involves two topics in investments. In Chapter 1 we establish that a quantitative measure capturing the tonality (TONE) of a publicly traded company’s annual 10-K documents is indeed a priced risk factor employing a large panel and multiple econometric approaches. Using both the conventional portfolio level, as well as the more recent individual stock level, regressions, we show that TONE captures risk associated with a company that is unique from those captured by the other established risk factors like: HML, Rm − Rf , SMB, MOM, RMW and CMA. We also establish that TONE is distinct from the recently identified plethora of factors (the “factor zoo” problem). We find that the coefficient associated with TONE is significantly negatively (positively) correlated with excess stock returns for the most (least) pessimistic stock portfolios. Further, TONE is negatively correlated with returns for portfolios with the smallest stocks and those with the smallest bookto-market ratios. n Chapter 2, we investigate the contribution of bond markets to price discovery in equity markets, using Hasbrouck (1995) information share approach. Based on corporate bond and equity transaction for S&P 500 stocks over a sample period between 2002 and 2014 we find that the monthly average information share (IS) of corporate bond markets is about 19%. We also find that the information share of high-yield bonds is significantly greater than investment grade bonds. We also examine the relative informational efficiency in corporate bonds through the leadlag relationship between their daily returns. Our results demonstrate empirically the theoretical arguments that informed investors trade in both stock and corporate bond markets, suggesting an important informational role for corporate bonds.

Degree

Ph.D.

Advisors

Chakravarty, Purdue University.

Subject Area

Accounting|Finance|Web Studies

COinS