We develop a model in which the dependence of the brokerage commission rate on share price provides an incentive for brokers to produce research reports on firms with low share prices. Stock splits therefore affect the attention paid to a firm by investment analysts. Managers with favorable private information about their firms have an incentive to split their firm's shares in order to reveal the information to investors. We find empirical evidence that is consistent with the major new prediction of the model, that the number of analysts following a firm is inversely related to its share price. THE CLASSICAL THEORY OF finance assigns great importance to the aggregate market value of the equity of a firm, but has no role for the number of shares in a firm's capital stock, or for the price of a single share. Consequently, the attention paid to this seemingly irrelevant variable by firms, investors, and brokers, as well as by legal and regulatory authorities, has so far defied plausible explanation. It is clear that firms attempt to manage the unit price of their shares by stock splits and occasional reverse splits, and there is a strong relation between the price per share and the size of the firm.1 Moreover, investors pay attention to stock splits, the abnormal return consequent on a split announcement being strongly related to the projected postsplit share price. This reaction has been explained by Brennan and Copeland (1988a) as the rational response to a costly signal by the firm. The basis of their argument is that it is costly for a firm to reduce its share price by splitting because the structure of brokerage commissions makes it more costly to trade in low priced shares. The relation between splits and trading Irwin and Goldyne Hearsh Professor of Banking and Finance at University of California, Los Angeles, and University of Southern California, respectively. We are grateful to Craig Holden and Jim Brandon for research assistance. We also thank (1983) show that, based upon portfolios formed on firm size, average price per share is monotonically increasing in size, and both the mean portfolio return and beta are monotonically decreasing in size. 1666The Journal of Finance costs has been acknowledged in the business press,2 and the signaling argument finds support in the work of McNichols and Dravid (1990) and others who show that stock splits are followed by unexpected increases in earnings. Less easy to explain is the change in stock price behavior following the date the split becomes effective. Ohlson and Penman (1985) observe an increase in the variance of returns following the split ex-date, and Brennan and Copeland (1988b) find that the systematic risk of firms also increases following the split.The Brennan and Copeland signaling model is at best a partial explanation of the stock price reaction to split announcements, for it relies on the observed structure of brokerage commissions which is taken as exogenous. Yet why should brokerage commissions depend on such a seemingly irrelevant variable as the s...

No abstract

This paper examines whether financial disclosures on acquired entities allow investors to effectively predict goodwill impairment, a task that has become more important following the recent abolishment of goodwill amortization. In predicting goodwill impairment, we use variables relating to the postacquisition performance of the operating segment(s) to which the acquired company's assets are allocated as well as to the characteristics of the acquisition. We find that available disclosures do not provide financial statement users with information to adequately predict future write-offs of goodwill. In fact, the characteristics of the original acquisitions are more powerful predictors of eventual goodwill write-offs than those based on segment disclosures of the acquired entities' performance. We also find that goodwill write-offs lag behind the economic impairment of goodwill by an average of three to four years. For one-third of the companies examined, the delay can extend up to ten years. Although most of our analyses are conducted on goodwill generated before the introduction of Statement of Financial Accounting Standards No. 142 (SFAS 142), certain features of the sample and the analysis suggest that the results are generalizable to the current reporting regime. Sensitivity tests on a smaller sample of goodwill write-offs made upon the adoption of SFAS 142 confirm this expectation.

No abstract

No abstract

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. This content downloaded from 131.172.. We acknowledge the helpful comments of Masako Darrough, Rob Heinkel, and Brett Trueman and workshop participants at UBC, UCLA, Berkeley, and the Journal of Accounting Research Conference. l "In 1979 the IRS estimated that if all firms which could have switched that year had done so, approximately $18 billion less in corporate income taxes would have been paid. Current estimates indicate potential savings of at least $6 billion annually" (Biddle and Martin [1985, p. 34]). 2 Morse and Richardson [1983] find that for a sample of firms which switched to LIFO in 1939-78, positive mean tax benefits from LIFO existed for each of the five years prior to the switch. Many papers have alluded to the potential for an accounting method choice to serve as a signal of inside information about investment and production activities (e.g., Gonedes and Dopuch [1974] and Dye [1985]).In this paper we attempt to explain the LIFO/FIFO puzzle through an information-signaling approach. First, we define the relevant costs and benefits of a FIFO choice and then derive the conditions under which a FIFO choice may serve as a credible signal of favorable private information by managers. Second, we analyze the conditions under which all firms select LIFO (FIFO) and no firm has an incentive to deviate to FIFO (LIFO). Finally, we study the market reaction to switches from FIFO to LIFO, and, in particular, we establish that there may be conditions under which a negative price reaction will be observed for the switching firms.The conceptual basis of our model is that managers possess private information regarding the firm's future prospects. The manager must select an inventory cost-flow assumption while recognizing that his compensation, either explicitly or implicitly, depends on both current and future market prices. Investors attempt to learn the manager's private information by observing the LIFO/FIFO choice and price the firm's securities accordingly. Managers anticipate investor reactions to the LIFO/FIFO choice and include these in their decision problem.Rational expectations equilibria in which managers maximize their compensation and investors correctly price securities are obtained using the Nash concept that no agent has incentives to deviate from the equilibrium, taking the actions of other agents as given.We show that there is an equilibrium in which managers truthfully signal their private information by not switching from LIFO to FIFO if and only if their information is favorable. Firms with favorable information remain at FIFO, thereby foregoing reduced taxes and the opportunity of reducing expected bankruptcy costs. The benefit of not switching is an immediate...

We present a model of managerial discretionary disclosure under litigation risk. In our model, the manager bears a personal cost of litigation in addition to the costs prescribed by the legal system. Investors share litigation damages with their attorneys and therefore are less than fully insured against market losses. We find that the interaction between the manager's level of aversion to litigation and investors' degree of insurance determines the quality of the manager's disclosure. This is a “cheap-talk” model similar to that of Crawford and Sobel [1982]. Like C&S, we find that, in general, the manager's and investors' preferences do not coincide and that there is a partition equilibrium that is consistent with managers providing a range forecast of earnings. In our setting, we are able to extend C&S and determine what condition is necessary for congruence of preferences resulting in full disclosure and informationally efficient security prices, consistent with managers providing point estimates of earnings. We also address a suggestion of C&S to test the robustness of the partition equilibrium by relaxing the assumption that investors have complete knowledge of the manager's preferences, and we find that, in general, the partition equilibrium is robust. We also show that under restrictive conditions, when the preferences of the manager are not known, a biasing equilibrium exists, consistent with managers providing biased forecasts of earnings.

This paper is an empirical test of the Arbitrage Pricing Theory using monthly security returns for 220 Canadian firms for the ten‐year period 1971‐80. Specific hypotheses tested are the existence of multiple factors generating Canadian security returns, the APT pricing relationship, the existence of a constant intercept when a security's expected return is expressed as a linear function of the risk premia on the factors, and the equality of the constant intercept to the riskfree rate of interest. A partial test of the congruence of the factor structure across groups of securities is also performed. Empirical procedures include factor analysis and the formation of minimum‐variance portfolios in order to derive the time series of returns on general economic factors, sensitivities of each firm's return to that of each factor, and the expected returns on the factors. It is found that at least twelve common factors generate Canadian security returns and that three to four of these factors have statistically significant risk premia. There is strong support for the existence of a constant intercept and marginal support for its being the riskfree rate. It is also found that, while there is a consistency in the underlying factor structures of the two groups, an identification problem prohibits matching the factors between groups. Résumé Ce texte présente des tests empiriques de la théorie de l'établissement des prix des titres par arbitrage (APT), fondés sur les taux de rendement mensuels de 220 titres canadiens pour la période 1971–1980. Les hypothèses testées sont celles de l'existence de facteurs multiples qui génèrent les taux de rendement, la structure des prix proposée par l'APT, l'existence d'une constante lorsqu'un taux de rendement est relié linéairement aux primes de risque attachées aux facteurs, et l'égalité de la constante et du taux d'intérět sans risque. Les techniques utilisées comprennent l'analyse factorielle et la formation de portefeuilles à variance minimale, qui permettent d'établir la série chronologique des rendements inhérents à chacun des facteurs, la sensibilité de chacun des titres aux variations de chacun des facteurs et la partie du rendement qui lui est attribuable. On découvre qu'au moins douze facteurs différents génèrent les rendements des titres canadiens et que les primes de risque de trois ou quatre d'entre eux sont statistiquement significatives. Les résultats appuient fortement l'hypothèse d'une constante, et faiblement celle de son égalité au taux d'intérět sans risque. Bien que les structures des deux groupes soient similaires, on ne peut les accoupler à cause d'un problème d'identification.