盈余质量与盈余的定价模式的影响外文翻译.docx

1、盈余质量与盈余的定价模式的影响外文翻译原文：Earnings Quality and the Pricing Effects of Earnings Patterns Earnings Patterns and Earnings Quality Previous research on pricing effects of earnings patterns has considered increasing earnings, earnings that meet or exceed analyst forecasts, and smooth earnings. The research i

2、s motivated by the observation that managers appear to focus on maintaining such patterns. We build on this observation by exploring the notion that managers accruals choices either reinforce or undercut the pricing effects of earnings patterns. In this section, we describe previous research (sectio

3、n 2.1), discuss why the three earnings patterns we consider might be economically distinct, as opposed to manifestations of a single underlying construct (section 2.2), and discuss the links between earnings patterns and earnings quality (section 2.3).1、Previous research on pricing effects of earnin

4、gs patterns Beginning with the observation that many managers appear to strive to report steadily increasing earnings, Barth, Elliott and Finn 1999 document rewards to such behavior: firms with increasing earnings have higher price-earnings multiples than other firms, after controlling for growth an

5、d risk. The control for growth is based on both increases in book value of equity and analyst earnings growth forecasts, and the control for risk is the variance of the most recent six years percentage earnings changes. They attempt to rule out the possibility that firms with long patterns of increa

6、sing earnings also share a valuation relevant factor that is known before the earnings pattern develops; results are mixed although the authors conclude that the weight of the evidence does not support the existence of such a factor. Barth et al. do not consider whether the quality of earnings might

7、 affect their results, although they note that if earnings increases are obtained via earnings management which investors discern and discount, their tests will be biased against finding pricing effects associated with patterns of increasing earnings (see their note 6).Kasznik and McNichols 2002 and

8、 Bartov, Givoly and Hayn 2002 focus on market effects of meeting or exceeding analyst forecasts. Kasznik and McNichols document that consistently meeting or exceeding analyst forecasts is associated with a higher multiple relating earnings to price, and this pricing effect is due to two distinct sub

9、-effects: higher future earnings and a premium associated with the act of meeting or exceeding forecasts for at least three years. They interpret their results as indicating that consistently meeting or exceeding analyst expectations connotes lower risk and, therefore, a higher earnings multiple. Ho

10、wever, they also note two open issues. First, they do not investigate why investors would associate lower risk with consistently meeting analyst forecasts. Second, they do not investigate the role of manipulating either expectations of earnings or earnings themselves. We believe our analyses shed li

11、ght on these issues; specifically, we posit that investors price firms that meet or exceed analyst earnings expectations and have high earnings quality higher than firms that meet the first but not the second condition.Using a different research design than Kasznik and McNichols, Bartov et al. 2002

12、document higher quarterly abnormal returns in quarters where firms meet or exceed forecasts. Several of their tests distinguish between shifts in analyst forecasts over the quarter and the earnings surprise, based on IBES reported earnings compared to the most recent IBES forecast. They interpret th

13、e finding that the return to an earnings surprise greatly exceeds the return to shifts in analyst forecasts as rationalizing expectations managementthe market penalty for dampening analyst expectations is swamped by the reward to a favorable earnings surprise. Subsequent tests support this interpret

14、ation.Most directly relevant to our analyses, Bartov et al. test whether the premium for meeting or exceeding analyst forecasts is smaller when there is evidence of either or both expectations management and earnings management. They define the former as the combination of negative analyst forecast

15、revisions and zero or positive earnings surprises, and the latter as the presence of unexpected accruals based either on the Jones 1991 model as modified by Dechow, Sloan and Sweeney 1995 or on defining normal accruals as working capital accruals plus depreciation and amortization. They conclude tha

16、t both expectations management and earnings management are associated with statistically reliable but small decreases in the market premium to meeting or exceeding analyst expectations, and that the “economically minor” (p. 198) effect of earnings management might be due to low power. We extend thes

17、e analyses by providing direct tests of whether the pricing effects of consistently meeting or exceeding analyst forecasts are a function of earnings quality, an inverse measure of earnings management.Thomas and Zhang 2002 report higher price-earnings multiples for firms with smooth earnings. Their

18、measures of earnings smoothness are derived from the standard errors of regressions of earnings on time and from the standard deviations of seasonally differenced quarterly earnings per share. They condition on intrinsic earnings volatility (proxied by the over-time variability of analyst earnings f

19、orecasts), to capture the valuation effects of managerial interventions to smooth earnings. Results indicate that both lower intrinsic earnings volatility and earnings smoothness are associated with higher price-earnings ratios and that this relation is due at least in part to an association between

20、 lower risk and lower intrinsic earnings volatility. Like Thomas and Zhang, Hunt, Moyer and Shevlin 2000 report that earnings multiples increase with earnings smoothness. Their measures of smoothness are derived from a separation of earnings into cash and total accruals, and total accruals into its

21、discretionary and nondiscretionary components, using the Jones 1991 model as modified by Dechow, Sloan and Sweeney 1995. Controlling for the standard deviation of cash flows (presumably as a proxy for unsmoothed earnings) Hunt et al. assess the incremental effects on pricing multiples when (1) the s

22、tandard deviation of cash plus nondiscretionary accruals is small relative to the standard deviation of cash flows and (2) the standard deviation of reported net income is small relative to the standard deviation of cash plus nondiscretionary accruals. The former measure is interpreted as smoothness

23、 induced by nondiscretionary accruals and the latter, as smoothness induced by discretionary accruals. Hunt et al. conclude that discretionary smoothness is associated with a larger positive pricing effect than nondiscretionary smoothness, and that discretionary smoothness is also associated with gr

24、eater earnings persistence.Taken together, prior research suggests two empirical regularities which we explore. First, there is a consistent positive pricing effect associated with earnings that are smooth, earnings that increase, and earnings that meet or exceed analyst forecasts. Previous research

25、 has considered these patterns one at a time and unconditionally. We consider the effects of each earnings pattern conditional on the others, to analyze whether the effects are in fact distinct, as opposed to varying manifestations of a single underlying factor. Second, there is (mixed) evidence sug

26、gesting that the pricing effects of earnings patterns might be conditional on whether earnings are managed. The effects of earnings management appear to be small, perhaps because of low power tests. We provide direct assessments of the effects of earnings quality (an inverse measure of earnings mana

27、gement).2、Relations among earnings patterns A pattern of increasing earnings simply implies that, over time, each successive earnings number exceeds that of the previous period. There is no necessary implication of, or conflict with, smoothness, measured either in terms of over time earnings variabi

28、lity (i.e., earnings might trend upward either smoothly or by varying amounts each period) or in terms of low reported income variability relative to intrinsic income variability. There is also no necessary implication of meeting, or not meeting, analyst expectations. In terms of managements behavio

29、r, earnings increases can be achieved by some combination of intrinsic growth and earnings management via either or both accruals manipulations and real actions. Depending on managements planning horizon and the underlying economics of the firm, earnings increases can be planned and manipulated over

30、 long periods by systematically building and then using “reserves” of accrualsthat is, some portion of unmanaged current period earnings can be delayed, via accruals management, and incorporated into later earnings, as needed. However, once an earnings number is announced, it becomes a fixed target

31、that must be exceeded if the pattern of increasing earnings is to be maintained.In contrast, patterns of exceeding analyst forecasts can be achieved by intrinsic performance, earnings management and expectations management. Evidence of expectations management is provided by, for example, Bartov et a

32、l. 2002 and Matsumoto 2002.2 To the extent managers can and do guide analysts toward a forecast that will ensure a zero or positive earnings surprise, the overtime pattern of earnings that meet or exceed analyst forecasts can take any form of increases and decreases, can exhibit any level of smoothn

33、ess, and can be entirely unmanaged. Thus, there is no necessary relation between meeting or exceeding analyst forecasts and any other earnings pattern, or between this earnings pattern and earnings quality. Finally, earnings smoothness is typically assessed by calibrating earnings variability over time, relative to the variability of so