Earnings quality and accrual mispricing: a country and firm-level investigation in the period surrounding SOX
thesisposted on 31.01.2017 by Strydom, Maria Elizabeth
In order to distinguish essays and pre-prints from academic theses, we have a separate category. These are often much longer text based documents than a paper.
This thesis investigates accrual mispricing through two related studies. The first examines the impact of earnings quality on accrual mispricing in the US and determines whether mispricing persists at the country level in the high earnings quality environment following the Sarbanes–Oxley Act (SOX) of 2002. The second study determines whether accruals are mispriced at the firm level and if such firm level mispricing persists so that profits can be generated by exploiting this trading strategy. The motivation to investigate the impact of earnings quality on the mispricing of accruals stems from the substantive literature that documents the persistent accrual anomaly (Sloan, 1996; Xie, 2001; Mashruwala et al., 2006) but fails to find its cause. This thesis examines whether this persistent country-level mispricing stems from investors being misled by low earnings quality. When earnings quality is low, investors will not be able to accurately price accruals. Earnings quality could therefore explain the existence of the accrual anomaly. Since SOX improved earnings quality, accrual mispricing should be less in the post-SOX environment. This thesis therefore also examines the mispricing of accruals post-SOX to determine whether it persists. This thesis’ second study is motivated by the cross-country (Pincus et al., 2007), country-level (Sloan, 1996; Xie, 2001), and industry-level (Zhang, 2007; Trejo-Pech et al., 2009) evidence that shows differences in mispricing. While cross-country, country-level, and industry-level mispricing have been investigated, there is no evidence on whether a firm-level anomaly exists, and this study therefore attempts to fill that void. This study is also motivated by the findings of Fama and French (2008) and Avramov et al. (2010). Fama and French (2008) investigate the pervasiveness of asset pricing anomalies and conclude that the accrual anomaly is one of few that persist in all size groups, cross sections, and sorts. Avramov et al. (2010) similarly investigate commonalities across asset pricing anomalies and conclude that whilst the majority of asset pricing anomalies are associated with downgrades in firm credit ratings, the accrual anomaly is an exception and remains unaccounted for and robust. Given the pervasiveness of this anomaly over time, this study investigates whether the firm-level accrual anomaly is similarly persistent. Two accrual mispricing models (Mishkin, 1983; Kraft et al., 2007) are employed in the first study to investigate the impact of earnings quality on the accrual anomaly. These models are augmented by including earnings quality proxies to determine their impact on mispricing. Given that both yield identical results, the second study employs only the Mishkin (1983) model to estimate firm-level mispricing. The mispricing model is employed for each firm in each year to estimate firm-level accrual mispricing. Significantly over- and underpriced accrual firms are identified, and a trading strategy of buying underpriced accrual firms and selling overpriced ones is examined for abnormal returns. The results from the first study indicate that earnings quality mitigates accrual mispricing. When investigated in the post-SOX environment, however, there is no evidence of mispricing. This is true even without considering earnings quality. These findings show that SOX have achieved its stated aim of improving disclosure quality so that investors are better able to estimate accrual persistence, mitigating the anomaly. The second study shows, however, that firm-level mispricing still exists. Specifically, it shows that both significantly over- and underpriced accrual firms exist in the same post-SOX sample, whereas at the country level no anomaly was documented. As with the differences in accrual mispricing documented at the aggregate market (Hirshleifer et al., 2009) and industry levels (Trejo-Pech et al., 2009), firm-level mispricing also differs from the country-level anomaly. Further analyses of firm-level mispricing show abnormal returns are available from a strategy of selling overpriced accrual firms and buying underpriced accrual firms. The first study contributes to the literature documenting the impact of earnings quality on accrual mispricing and thus provides evidence of the importance of good disclosure quality in ensuring efficient pricing. It also contributes by showing that SOX has achieved its stated aims of improving disclosure quality and has thus mitigated mispricing at the country level. A further contribution is the direct comparison of the accrual mispricing models of Mishkin (1983) and Kraft et al. (2007) and the evidence that they yield similar results. The second study makes two main contributions: First, it documents that firm-level accrual mispricing exists, even in the absence of a country-level anomaly. Second, the study shows that at the firm level both significantly over- and underpriced accrual firms exist, and it establishes the persistence of this firm-level mispricing. It also documents that investors can profit from a firm-level accrual mispricing strategy.