Managerial choice of discretionary accounting methods: An empirical evaluation of security market response

Abstract

This study presents a comprehensive evaluation of managerial choice of discretionary accounting methods and the use of the reporting process in income smoothing. After analyzing concepts of smoothing behavior, including managerial motivations to engage in smoothing, security market reactions to smoothing efforts are empirically tested to corroborate managerial motivations for, and security market evaluations of, income smoothing. This study looks at income smoothing using reporting variables as involving the choice of a portfolio of discretionary accounting methods over a period of time with the effect of reducing the fluctuations of reported earnings relative to cash flows. After identifying smoothing behavior, the study tests hypotheses of differential information content of accounting earnings by looking at the security market reactions to the release of financial reports. In addition to changes in security prices, the study also looks at other capital market variables such as the systematic risk and changes in dividends. A pooled time-series cross sectional regression analysis using an estimated generalized least squares estimator finds significant differences in the risk-adjusted security returns of 104 smoothing and 29 non-smoothing firms around the quarterly earnings announcement dates from June 1984 to March 1987. Non-smoothing firms have a higher market reaction around the announcement time, implying a higher level of prior market uncertainty about their future earnings and cash flows. This result is consistent across different estimation methods and sub-samples. In addition, a cumulative average abnormal return analysis shows similar differences in the indirect information content of accounting earnings over a period of 12 months for the three years 1984 to 1986. In additional empirical tests, the study finds that the average systematic risk of the smoothing firms is significantly lower than the average systematic risk of non-smoothing firms. This is in line with the conclusions about the higher prior level of uncertainty of the market about the future earnings and cash flows of non-smoothing firms. The non-smoothing firms had fewer changes in dividends declared, probably due to the dividend payments being used as a signalling device to reduce some of the prior uncertainties about the future financial performance of the non-smoothing firms.