[This paper reviews and critiques the positive accounting literature following publication of Watts and Zimmerman (1978, 1979). The 1978 paper helped generate the positive accounting literature which offers an explanation of accounting practice, suggests the importance of contracting costs, and has led to the discovery of some previously unknown empirical regularities. The 1979 paper produced a methodological debate that has not been very productive. This paper attempts to remove some common misconceptions about methodology that surfaced in the debate. It also suggests ways to improve positive research in accounting choice. The most important of these improvements is tighter links between the theory and the empirical tests. A second suggested improvement is the development of models that recognize the endogeneity among the variables in the regressions. A third improvement is reduction in measurement errors in both the dependent and independent variables in the regressions.]

We estimate a firm-year measure of accounting conservatism, examine its empirical properties as a metric, and illustrate applications by testing new hypotheses that shed further light on the nature and effects of conservatism. The results are consistent with the measure, C_Score, capturing variation in conservatism and also predicting asymmetric earnings timeliness at horizons of up to 3 years ahead. Cross-sectional hypothesis tests suggest firms with longer investment cycles, higher idiosyncratic uncertainty and higher information asymmetry have higher accounting conservatism. Event studies suggest increased conservatism is a response to increases in information asymmetry and idiosyncratic uncertainty.

Using an accounting conservatism theory that reflects accounting's role in practice, we investigate the relation between two extensively used measures of conservatism: asymmetric timeliness of earnings and the market-to-book ratio (MTB). We predict and observe that when asymmetric timeliness is measured cumulatively over long periods, its relation with end-of-period MTB is positive. When asymmetric timeliness is measured over short periods, its dependence on beginning-of-period composition of equity value (EV) is responsible for its negative association with MTB. Further, asymmetric timeliness appears to measure conservatism more efficiently when it is estimated cumulatively over multiple periods.

In this study we specify a simple earnings model, present managerial discretion hypotheses from existing literature, and assume efficient markets in order to evaluate five discretionary-accrual models. The five discretionary accrual models are the same as those evaluated in Dechow, Sloan, and Sweeney [1995]. The models are Healy [1985]; DeAngelo [1986]; Jones [1991]; Jones as modified in Dechow, Sloan, and Sweeney [1995]; and the industry model proposed by Dechow and Sloan [1991]. We specify three managerial discretion hypotheses. First, under the performance measure hypothesis, discretionary accruals help managers produce a reliable and more timely measure of firm performance (i.e., earnings) than using nondiscretionary accruals alone. Second, the opportunistic accrual management hypothesis is that discretionary accruals are employed to hide poor performance or postpone a portion of unusually good current earnings to future years. Finally, discretionary accruals are noise in earnings. This is the noise hypothesis. Our contribution is to make the joint hypotheses explicit and generate explicit predictions about the relative variability of earnings components,

Value maximization requires either that knowledge is transferred to those with the right to make decisions, or that decision rights are transferred to those who have the knowledge. A tradeoff of knowledge transfer costs and control costs is required. Characteristics of firms' investment opportunity sets (IOSs) that affect knowledge transfer costs and control costs are identified. Testable predictions about the relations between these characteristics and firms' decentralization decisions are developed and tested. The evidence presented is consistent with our predictions and is robust to different ways of measuring variables.

[In competitive product markets, product prices and thus firms' revenues incorporate the cost of equity capital. In a competitive capital market, the cost of equity capital (the expected return on equity) increases with the risk of firms' investments. Because accounting earnings are calculated without deducting the cost of equity capital, they are expected to be an increasing function of firms' investment risks. This simple competitive equilibrium analysis predicts a positive relation between changes in investment risk and expected earnings. The presence of corporate debt complicates the analysis because leverage effects seem likely to affect the relation between changes in investment risk and expected earnings. Using annual earnings and return data from 1950 to 1988, we document a statistically significant positive association between changes in equities' relative risks and in earnings. However, on average, only a small proportion of changes in earnings can be attributed to changes in risk. A much larger proportion is attributable to changes in economic rents (windfall gains and losses). The observed positive association between changes in earnings and changes in equities' risks suggests that leverage effects do not fully offset the effect of changes in investment risks. This association is robust with respect to subperiod analysis, alternative specifications of the earnings change variable, alternative data-availability requirements, and the number of portfolios formed.]

A model of earnings, cash flows and accruals is developed assuming a random walk sales process, variable and fixed costs, and that the only accruals are accounts receivable and payable, and inventory. The model implies earnings better predict future operating cash flows than current operating cash flows and the difference varies with the operating cash cycle. Also, the model is used to predict serial and cross-correlations of each firm's series. The implications and predictions are tested on a 1337 firm sample over 1963–1992. Both earnings and cash flow forecast implications and correlation predictions are generally consistent with the data.