New Methods in Empirical Finance
Abstract
This dissertation studies two new methods in empirical finance. Section 2 applies a rolling estimation window approach to adjust for time-varying risk parameters in asset pricing models when estimating long-run abnormal returns after major corporate events. Abnormal returns are defined as realized returns minus predicted returns on each day in a five-year, post-event period. A variety of asset pricing models are employed to compute out-of-sample predicted returns in different estimation windows for seasoned equity offerings (SEOs) and mergers and acquisitions (M&As). We find that, after an initial significant return response in the month or two after corporate announcements, abnormal returns thereafter disappear. Robustness checks corroborate our results: (1) with or without matched and random control samples, (2) for different asset pricing models including the CAPM market model, and (3) in robustness tests of share repurchases (SRs), stock splits (SPLTs) as well as different subperiods. Also, simulation tests confirm the robustness of the RPE method to potential risk shifts. In summary, after dynamic risk adjustment, long-run abnormal returns are not evident after these major corporate actions.
Section 3 proposes a new empirical method to estimate the global minimum variance portfolio without the covariance matrix to avoid associated estimation errors. Unlike previous studies, we employ extant asset pricing models to test efficiency, sort portfolios, and assign weights to individual assets. Based on out-of-sample analyses of U.S. stock returns in the sample period from 1968 to 2019, empirical results show that the global minimum variance portfolio has relatively high expected returns, low variance, and high Sharpe ratios. The results are robust with respect to different asset pricing models, extreme weights for individual stocks, and different subperiods.
Subject
abnormal returnlong-run event study
global-minimum variance portfolio
optimal portfolio
asset pricing models
Citation
Han, Yao (2021). New Methods in Empirical Finance. Doctoral dissertation, Texas A&M University. Available electronically from https : / /hdl .handle .net /1969 .1 /193120.