The Time Series Behavior of Earnings

摘要: The time series behavior of earnings is an important area for empirical research because its implications related in several areas accounting and finance. Although many other examples could be provided, three "accounting" issues immediately come to mind: (1) income smoothing, (2) the relative forecast ability alternative measurements, (3) interim reporting. hypothesis that management uses discretionary practices smooth was first posited by Gordon [21] later tested Gordon, Horwitz, Meyers [22] Copeland [14] among others. As stated smoothing involves minimizing deviations reported from some standard, where standard defined terms normal income. Normal has never been precisely at conceptual level, but cases it appears have used sense expected value process a given point time. A variety models used, fact assess or period. Each model makes specific assumptions about generating numbers. Any inferences drawn evidence regarding existence (or lack it) are dependent upon validity made underlying process. Moreover, as shown paper, certain processes attempts can exactly opposite effect. Yet literature represent only narrow range possible alternatives, little justification (either priori empirical) offered their behalf, nor there any direct, rigorous investigation nature itself.