PROBLEMS OF COMPARING or choosing among models of a stochastic process are frequently encountered in empirical research. In many such situations, conventional statistical procedures offer little guidance since they assume that the model is given. If the alternative models can be nested in a more general model, standard estimation and testing procedures can be employed. Often, however, such general models are not readily available and other considerations may dictate against their use. Recently, there has been considerable progress in the development of methods for comparing alternative non-rested models. A review of this work both Bayesian and non-Bayesian, is given in Gaver and Geisel [1]. Discussions of the Bayesian approach to the comparison of linear regression models are given in Zellner [3, Ch. 10] and Lempers [2], among others. In this paper we consider Bayesian comparison of linear models in which the disturbances have non-scalar covariance matrices. General posterior odds expressions are given and specialized to the case of first order autoregressive disturbances. We also consider a specification error problem in this context; that is, we examine the effect of ignoring the non-scalar covariance structure on the posterior odds ratio. For the first order autoregressive disturbance case we give an approximate expression indicating the magnitude of the error involved in computing the posterior odds ignoring the serial correlation. The accuracy of this approximation is investigated via a small sampling experiment. We use the following notatiori: Let the ith model, Mi (i = 1, 2,. .., N), be y = Xi/3i + yi where y is a T x 1 vector of observations on the random (dependent) variable of interest, Xi is a T x ki matrix of observations on the explanatory variables of Mi (Xi is assumed to be non-stochastic with rank ki), p3i is a ki x 1 vector of unknown parameters of Mi, and ui is a T x 1 vector of disturbances of Mi (yi is assumed to have a normal distribution with E(yi) = 0, and E(yiyii) = U22i where 2Ji is an unknown T x T positive definite symmetric matrix with trace (i) = T).2 Probability functions for the models are denoted by P( ), densities for parameters by 7r( ), and densities for observations by p( ).

[This study investigates the role of alternative earnings components in the CEO cash compensation function. We find that cash compensation is significantly positively related to above the line earnings, as long as results are positive. Compensation is shielded from the effects of above the line losses. Similarly, nonrecurring transactions that increase income flow through to compensation, but nonrecurring losses do not. This effect is noted for gains and losses that arise both from extraordinary transactions, discontinued operations and nonrecurring items that do not qualify for below the line presentation. Thus, the data tell a remarkably consistent story: gains flow through to compensation, but losses do not. The classification of the gain or loss on the income statement is of relatively little importance.]

This paper presents additional evidence on the relation between the investment opportunity set and financing, dividend, and compensation policies. Our results are based on a sample of 237 growth firms and 237 nongrowth firms. We find that growth firms have significantly lower debt/equity ratios and exhibit significantly lower dividend yields than nongrowth firms. We also find that growth firms pay significantly higher levels of cash compensation to their executives and have a significantly higher incidence of stock option plans than nongrowth firms. However, controlling for firm size, the incidence of bonus plans, performance plans, and restricted stock plans does not differ between growth and nongrowth samples.

[Often overlooked in empirical analyses of the relation between audit quality and audit fees is the recognition that they are mutually determined by the interaction of the client's demand for, and the audit firm's supply of, audit quality. Failure to account for this endogeneity can lead to biased inferences concerning the audit quality/audit fee relation. We adopt a simultaneous equations estimation procedure (offered by Amemiya 1978) applicable to jointly determined endogenous variables when one of the variables (in our case the quality review outcome) is qualitative in nature.1 To illustrate the procedure, we use a dataset developed by the United States General Accounting Office (GAO 1987) in its study of the audit procurement practices of entities receiving federal financial assistance. Inferences using this procedure differ from those of single-stage analyses. The results suggest that within the context of the application examined, audit fees appear to be positively related to the supply of audit quality and inversely related to the demand for audit quality. These findings have implication for studies involving audit fees and studies examining the demand characteristics of auditing.]

[This study examines the stock market reaction to the adoption of long-term compensation agreements for top management that are based on accounting goals. Prior researchers (Baril 1988; Larcker 1983; Smith and Watts 1982) argue that these agreements, known as performance plans, motivate executives to improve firm performance by working harder, lengthening their decision horizons, and becoming less risk-averse in their investment decisions. However, because the agreements are based on accounting goals, performance plans can distract managerial attention from the maximization of share price and can also encourage the manipulation of the accounting system (Gibbons and Murphy 1989). Managers may also reject projects with large, positive net present values in favor of less valuable projects with higher accounting profits. In an earlier study, Larcker (1983) reports a significantly positive stock market reaction for a sample of 21 firms that adopted performance plans between 1971 and 1978. The reaction occurs on the trading day after the SEC receives the proxy statement announcing the plan adoption. However, we find no significant reaction during the two-day announcement period beginning with the SEC stamp date for a sample of 209 adoptions that occur between 1971 and 1980. Further, no evidence of abnormal stock performance is detected for the two-day periods beginning with either the date on which the board of directors voted to approve the plan or the proxy statement date. Our study highlights the difficulties inherent in analyzing the stock market reaction to announcements made in proxy statements. First, pinpointing the timing of information dissemination is problematic. Second, firm-specific information unrelated to the event of interest is often released in the proxy statement and at the annual meeting. This complicates the interpretation of any unusual stock price behavior. In our study, significantly positive abnormal performance is observed for both adopting and nonadopting firms during the period beginning two days after the SEC stamp date and ending the day after the shareholders' meeting date. This suggests that the reaction observed for the adopting firms is due to the impending shareholders' meeting rather than to the performance plan adoption per se. We suggest that the results of any event study involving announcements made around the time of the annual shareholders' meeting should be interpreted with caution.]

We extend Healy (1985) by examining the relation between discretionary accruals and bonus plan bounds for a sample of 102 firms for the 1980–1990 period. Contrary to Healy, we find that when earnings before discretionary accruals fall below the lower bound, managers select income-increasing discretionary accruals (and vice versa). We believe that our results are more consistent with the income smoothing hypothesis than with Healy's bonus hypothesis. However, mechanical selection bias in portfolio formation cannot be entirely ruled out as an alternative explanation for our results.

This study investigates the rote of alternative earnings components in the CEO cash compensation function. We find that cash compensation is significantly positively related to above the line earnings, as long as results are positive. Compensation is shielded from the effects of above the line losses. Similarly, nonrecurring transactions that increase income flow through to compensation, but nonrecurring losses do not. This effect is noted for gains and losses that arise both from extraordinary transactions, discontinued operations and nonrecurring items that do not qualify for below the line presentation. Thus, the data tell a remarkably consistent story: gains flow through to compensation, but losses do not. The classification of the gain or loss on the income statement is of relatively little importance.

Examines the stock market reaction to the adoption of long-term compensation agreements for top management that are based on accounting goals. Method of study; Reaction observed for the adopting firms due to the impending shareholders' meeting rather than to the performance plan adoption per se; Caution in the interpretation of announcements made around the time of the annual shareholders' meeting.