Earnings management and loss reversal

Abstract

This research aims to detect and measure earnings management using a newly modified version of the standard Jones model (Jones, 1991). The standard model is extended to include a measure of discretionary accruals as an additional regressor instead of using the residuals. The variable used to measure discretionary accruals is a composite variable that consists of two components, one that represents the incentive and the other represents the tool of manipulation. The model is applied to detect earnings management in loss reversal companies for listed companies in Jordan and examine the market reaction to the loss reversal. The model is also applied on loss reversal companies for listed companies in the UK and the US. In chapter three, the new model is applied on listed companies in Amman Stock Exchange (ASE). The ASE is structured into two markets: the first market and the second market. Companies are motivated to be listed or remain listed in the first market since it only lists profitable companies. Companies reporting losses more frequently are listed in the second market. Results provide evidence of earnings management for companies listed in the first market. Companies that report a loss in a previous period manipulate in the following period to report profits. As a result of loss reversal, they preserve their place in the first market and avoid dropping back to the second market. This research conducts statistical simulation tests to compare the extended Jones model with the standard model. Results show that the extended model detects earnings management better than the standard one. This new model also separates discretionary accruals from measurement error (i.e. residuals) and makes it possible to accurately measure the whole amount of manipulation. Chapter four examines the investor reaction to the manipulation taking place in the first market. Results show that the market is pricing the discretionary accruals (the manipulation) as a component of net income, although they result only from earnings management. In chapter five, the model is applied on loss reversal firms listed in the UK and in the US. Results show that the companies manipulate to reverse losses and the manipulation depends on to the presence of R&D activities and the changing level in these activities.