Printer Friendly

A critical review on analysts' earnings forecasts.


Analysts' earnings forecasts are routinely employed as a predictor of future earnings in many stock valuation models and as a proxy for the market's earnings expectations in capital market-based accounting and finance research. The common practice is justified by superior accuracy of analysts' forecasts to predictions derived from time-series models of earnings and the market's positive responses to analysts' forecast revisions. However, analysts' forecasts do differ significantly from actual earnings (Dreman and Berry 1995) and the market's earnings expectations do not consistently follow analysts' forecasts (Hopwood and McKeown 1990; Walther 1997). While various types of empirical research in accounting and finance have utilized analysts' forecasts, limitations of analysts' forecasts have not been well understood (Mikhail et al. 1999). Unlike prior literature reviews on analysts' earnings forecasts (e.g., Brown 1993; Givoly and Lakonishok 1984; Schipper 1991), this study draws together the results of some research on limitations of analysts' forecasts. It focuses on limited superiority of analysts' forecast accuracy, disconnection between analysts' and the market's expectations of earnings, and bias in analysts' forecasts. Understanding limitations of analysts' forecasts would improve the interpretation of empirical results that use analysts' forecasts.


It has been well established that analysts provide more accurate and precise earnings forecasts than time-series models of earnings. Table 1 shows that (unsigned) forecast errors of Earnings Forecaster, Value Line, and I/B/E/S are consistently smaller than any time-series models. Analysts' superior predictive ability to time-series models is explained by timeliness of forecasts and broad information set. Analysts have a timing advantage in three ways: First, analysts can use more recent information, which becomes available after forecast initiation date of time-series models, but before analysts' initiation (Brown et al. 1987a; Fried and Givoly 1982). Second, analysts are able to respond to situations such as strikes or sudden swings in earnings, whereas time-series models are either unable to respond or slow to do so (Collins and Hopwood 1980). Third, analysts are relatively better than time-series models at distinguishing among permanent, transitory and price-irrelevant earnings shocks (Ramakrshnan and Thomas, 1998). After eliminating the timing advantage, the superior accuracy of analysts' forecasts over time-series predictions still exists because analysts presumably utilize all available information that includes non-accounting information on firms, industry and general economy, while time-series models rely exclusively on past earnings (Brown et al. 1987a; Fried and Givoly 1982).

Paradoxically, these advantages limit individual analysts' forecast accuracy. Table 2 categorizes prior studies' results on limitations of analysts' forecast accuracy into forecast timing and available information set. These empirical results form the following model:


First, analysts' forecast accuracy (ACCURACY) declines with the length of forecast horizon (HORIZON), but increases with forecast recency (RECENCY) and frequency (FREQUENCY) because the latest earnings-relevant information can be incorporated in forecasts. Brown (1991), Mozes and Wiliams (1999), and O'Brien (1988) show that the most current clusters of analysts' forecasts are more accurate than contemporaneous consensus of analysts' forecasts. Second, firm size and analyst coverage (the number of analysts following a firm) have a positive effect on availability of information (Barth and Hutton 2001; Bhattacharya 2001; Hong et al. 2000) and thus analysts' forecasts are more accurate for firms with larger size (SIZE) and higher analyst coverage (COVERAGE). Brokerage firms' size (BROKERAGE) also have a positive impact on analysts' forecast accuracy because large brokerage firms provide superior resources (Clement 1999), better research networks (Stickel 1995), and closer relationships with management of companies (Jacob et al. 1999).

Less experience and task complexity are likely to limit analysts' forecast accuracy. Prior studies, however, report mixed empirical results. Clement (1999) and Mikhail et al. (1997) provide evidence that more experienced analysts provide more accurate forecasts than less experienced analysts, but Jacob et al. (1999) show that there is no significant influence of experience on forecast accuracy when specific analyst-company alignments are controlled. Jacob et al. (1999) attribute the positive association between analysts' experience and forecast accuracy to survival bias because more capable analysts survive longer, thus they are over-represented among the observations with high values of experience. Analysts' task complexity has been measured either by the number of firms followed by an analyst or by the extent to which an analyst concentrates his forecasting in an industry. Consistent with the general expectations, absolute forecast errors are positively related with the number of firms followed by an analyst. However, the relation of forecast accuracy with analysts' industry concentration is mixed. For instance, Mikhail et al. (1997) find little support for a relation between forecast accuracy and industry concentration that is defined as the proportion of firms in same industry among the total number of firms followed by an analyst, but Jacob et al. (1999) find a negative coefficient for the variable of industry concentration in the regression of forecast errors, using the definition of Mikhail et al. (1997).

Another potential limitation of analysts' forecast accuracy is caused by analysts' herding behavior. DeBondt and Forbes (1999), Hong et al. (2000), and Trueman (1994) provide evidence of the reputation-based herding theory that career-concerning analysts herd toward consensus forecasts. Therefore, followers in time provide more accurate forecasts than leaders, but followers are more likely to herd on the consensus or the predecessor's forecasts (Shroff et al. 2003). However, Bernhardt et al. (2002) and Zitzewitz (2001) document that analysts do not herd as is often assumed, but rather they exaggerate their differences with the consensus forecasts. Analysts' herding behavior that has impact on forecast accuracy is an issue for future research on the limitation of analysts' forecast accuracy.


Since analysts provide more accurate forecasts than do time-series models, analysts' forecasts are expected to be a more precise surrogate for the market's expectations of earnings (Brown et al. 1987b; Fried and Givoly 1982). However, prior studies' empirical results indicate that the market's earnings expectations do not consistently follow analysts' forecasts. For instance, analysts' forecast errors have a higher association with abnormal stock returns than the random walk model's forecast errors (Wiedman 1996), but a lower association than the quarterly autoregressive model's forecast errors (O'Brien 1988). Hughes and Ricks (1987) show that analysts' forecast errors are significantly associated with abnormal stock returns, but do not outperform time-series models as a proxy for unexpected earnings. This anomalous relation of analysts' forecasts with stock returns suggests analysts' inefficiency in incorporating available information and/or investors' failure in impounding some portion of the value-relevant information in analysts' forecasts.

There is mounting evidence of the analysts' inefficiency. First, analysts do not fully incorporate past information available at the time of their forecasts. For example, analysts are not likely to fully use past forecast errors (Givoly 1985), past stock returns (Abarbanell 1991), past earnings (Ali et al. 1992, Mikhail et al. 2003), and past accruals (Bradshaw and Richardson 2001). Second, analysts have tendency to interpret information in a manner of bias. For instance, they systematically underreact (Abarbanell 1991; Abarbanell and Bernard 1992; Elliot et al. 1995; Lys and Sohn 1990; Teoh et al. 1998) or overreact (DeBont and Thaler 1990) to new information. More specifically, evidence in Easterwood and Nutt (1999) shows that analysts underreact to negative information, but overreact to positive information. Kim et al. (2001) analytically show that analysts' consensus forecasts inefficiently aggregate information by assigning too much weight to analysts' common information relative to their private information. Third, analysts are likely to mispredict earnings when a firm reports a loss. Benesh and Peterson (1986) and Clayman and Schwartz (1994) find that analysts are much more likely to get the sign of the earnings number wrong when a firm reports a loss. Degeorge et al. (1999) also suggest that the distribution of analysts' forecast errors for losses is much greater than for profits. Finally, the complexity of information contributes to the failure of analysts to properly interpret the information when forecasting earnings. Chen et al. (2002) provide evidence of analysts' failure to distinguish between permanent and transitory effects of deferred tax adjustment. To interpret this adjustment correctly, analysts need to understand the unique and complex nature of this transaction, such as accounting rules for deferred tax assets as well as certain portions of SFAS 109.

As a number of studies suggest that investors fail to fully assimilate publicly available value-relevant information in forming earnings (e.g., Ball and Bartov 1996; Bernard et al. 1997; Bernard and Thomas 1990; Freeman and Tse 1989; Rendleman et al. 1987), Elgers et al. (2001), Givoly and Lakonishok (1979), Mendenhall (1991), Shane and Brous (2001) and Stickel (1991) provide evidence of investors' underreliance on analysts' forecasts. For instance, Walther (1997) find that investors place more weight on analyst forecasts and less weight on time-series predictions as investors' sophistication increases.

Overall, prior studies indicate disconnection between analysts' and the market's earnings expectations, contrary to widespread use of analysts' forecasts as a proxy for the market's expectations of earnings. An alternative explanation for the contradictory finding is the analysts' inefficiency in processing value-relevant information. However, prior studies have not yet full modeled how analysts process information. This unresolved question that is essential to understand properties of analysts' forecasts makes this research area lively and rewarding for both theoreticians and empiricists interested in the operation of the analyst industry, the formation of investors' expectations and the interaction between accounting numbers and stock returns (Amir and Sougiannis 1999).


Absolute forecast errors may be larger for firms which are hard to predict, but analysts should err equally in both direction (under- and overestimates of future earnings) if they reveal their expectations in an unbiased manner (Das et al. 1998; Sougiannis and Yaekura 2001). However, numerous studies have confirmed analysts' optimistic bias (i.e., overestimates of future earnings) since it was reported more than 20 years ago (Barefield and Comiskey 1975; McDonald 1973). Table 3 shows research designs and findings of studies on analysts' forecast bias.

Prior studies provide three plausible explanations why analysts' forecasts are systematically optimistic. First, analysts have the incentive to maintain good relations with management because management is the most important source of non-public information. Moreover, management penalizes analysts based upon the content of their forecasts by limiting or cutting off analysts' future contact with management who disclose pessimistic earnings predictions. Therefore, analysts are likely to report optimistic predictions to cultivate relations with management. Empirical and analytical studies consistently support this management relation hypothesis. For example, Francis and Philbrick (1993) contend that a less favorable stock recommendation creates an incentive for the analyst to report optimistic forecasts because analysts wish to counter negative effects on their relations with management generated by an unfavorable stock recommendation. Das et al. (1998) posit that analysts more overstate earnings for firms that are hard to predict using publicly available information because optimism facilitates access to management's non-public information. Lim (2001) analytically proves that analysts trade off optimistic forecast bias to improve management access and forecast accuracy. However, the management relations hypothesis fails to consider the fact that managers may be displeased with optimistic forecasts because they have incentives to avoid negative earnings surprises (Burgstahler and Eames 2000). Matsumoto (1999) presents evidence consistent with management's preference for pessimistic forecasts to avoid negative earnings surprises. Numerous articles in the popular financial press (e.g., Ip 1997a,b; McGee 1997) reiterate that optimistic forecasts are not an effective way to curry management favor (Eames et al. 2002).

Second, compensation is an important economic incentive motivating sell-side analysts to issue optimistic forecasts. Brokerage or investment-banking firms may compensate analysts for their research service to stimulate brokerage commissions or securities-issue commissions. Thus, sell-side analysts have incentives to boost trade for their brokerage firms (trade-boosting incentive) and to favorably evaluate investment-banking client companies (investment-banking incentive). Kim and Lustgarten (1998) and Eames et al. (2002) test analysts' trade-boosting incentive. Kim and Lustgarten (1998) indicate that broker-analysts (Multex Investment Inc. analysts) are likely to report more (or less) optimistic earnings forecasts for buy (or sell) recommendations than non-broker analysts (Standard and Poor's and Value Line analysts). Eames et al. (2002) show that broker-analysts' earnings forecasts (34,612 annual forecasts for 1988-96 from Zacks) are significantly optimistic for buy recommendations and significantly pessimistic for sell recommendations. On the other hand, Lin and McNichols (1993) and Dechow et al. (1997) investigate analysts' investment-banking incentive. They consistently find that affiliated analysts who work for an investment-banking firm report more optimistic earnings forecasts for investment-banking clients than do unaffiliated analysts. For example, Lin and McNichols (1993) report that affiliated analysts' five-year earnings growth rate forecasts (21.29% of price) are significantly more favorable than those (20.73% of price) made by unaffiliated analysts. Dechow et al. (1997) document that analysts issue more optimistic forecasts for firms which are underwriting clients.

Third, several studies attempt to explain analysts' optimism by the cognitive bias, statistical and memory errors that are common to all human beings. For instance, Easterwood and Nutt (1999) document that analysts overreact to positive information, but underreact to negative information. Elton et al. (1984) also show that analysts have a marked tendency to overestimate the growth rates of securities they believed will perform well and to underestimate the growth rate of companies they believed will perform poorly. Moreover, analysts are likely to add firms they view favorably and drop firms they view unfavorably in their forecast announcements (i.e., selection bias) so that unfavorable forecasts are withheld (Affeck-Graves et al. 1990; McNichols and O'Brien 1997).

Existing empirical studies report the effect of earnings sign, firm size, analyst coverage, and forecast horizon on analysts' forecast bias. Ali et al. (1992), Brown (1998) Dowen (1996) and Hwang et al. (1996) document that analysts are more optimistic in forecasting earnings for firms whose actual earnings are losses. For example, Hwang et al. (1996) report that the optimistic bias in analysts' forecasts is ten times greater when firms report losses. Brown (1998) shows a dramatic difference of optimistic bias between firms reporting profits and losses on a sample of I/B/E/S for 1984-96; an optimistic bias of 11.7 cents exists when firms report losses, while a pessimistic bias of 2.3 cents exists when firms report profits. The difference in forecast bias between profit and loss firms is attributed to managers' incentives. When managers report profits, they try to meet or slightly beat analysts' forecasts. If managers are unable to report profits that meet or slightly beat analysts' forecasts, they may manipulate accruals in order to report small positive earnings surprises and to avoid small negative earnings surprises (Degeorge et al. 1999; Eames and Burgstahler 2003). Consequently, analysts' forecasts of firms that report profits are likely to be slightly underestimated than actual earnings. On the other hand, managers are unconcerned about meeting analysts' forecasts when they report losses. Instead, they may undertake actions to enhance future earnings, possibly taking a big bath that increases expenses and losses in the current period and reduces expenses and losses in subsequent periods. If managers fail to forewarn analysts that they are about to report a loss, analysts are likely to get the sign of the earnings number wrong. Moreover, if managers do not forewarn analysts that they are about to take a big bath, analysts' forecasts are extremely overestimated. Thus, loss firms' forecasts are shown as being large optimistic.

Small firms have relatively more optimistic bias because they are much more (nearly 10 times as) likely to report losses and/or their managers are less likely to manage profits (Brown 1998; Lim 1998). For example, Lim (1998) find that optimistic bias is considerably higher at 2.5% of price for small firms versus 0.53% of price for large firms. Analyst coverage and forecast horizon have negative impact on the magnitude of analysts' optimistic bias. Brown (1997, 1998) and Das et al. (1998) show that firms followed by few analysts have relatively more optimistic bias. Richardson et al. (1999) report that the optimistic bias declines dramatically, from 0.91% of price to 0.09% of price, as the forecast horizon is shortened from one year to one month,

The optimism of analysts' forecasts appears to be waning in recent years. Brown (1997, 1998) show that the optimistic bias in analysts' forecasts has been mitigated over time. His annual analysis on I/B/E/S reveals a range of optimistic bias from 3.35 cents per share in 1985 to 0.30 cents per share in 1997. Richardson et al. (2001) provide evidence of a switch from upward-biased to downward-biased annual forecasts as the earnings announcement date approaches. Consistent with Brown (1997, 1998), this effect is strongest in the post-1992 years and especially in 1997. Gu and Wu (2000) suggest that such forecast pessimism is a natural result of positive earnings skewness because analysts strive to minimize mean absolute forecast error.


Analysts' earnings forecasts have been used extensively as a predictor of future earnings and a proxy for the market's expectations of earnings. The superior accuracy of analysts' forecasts to predictions derived from time-series models of earnings justifies this common practice in empirical studies in accounting and finance. However, analysts' forecast accuracy is limited by forecast timing (i.e., forecast horizon, recency, and frequency) and available information set (i.e., firm size, analyst coverage, and brokerage size). Also, analysts' forecasts do not consistently generate measures of earnings surprise that are more highly associated with contemporaneous stock returns, suggesting that analysts' forecasts are not necessarily better representations of the market's earnings expectations than are time-series predictions. The disconnection between analysts' and the market's expectations of earnings is reconciled by analysts' inefficiency in incorporating available information and investors' failure in impounding some portion of the value-relevant information in analysts' forecasts. Another limitation of analysts' earnings forecasts is analysts' optimism in forecasting future earnings although the optimistic bias is waning in recent years. The findings in this study indicate the need for further research in several areas: the relation of analysts' herding behavior on forecast accuracy, how analysts process information to produce forecasts, and a structural shift toward forecast pessimism in recent years.


Abarbanell, J. 1991. Do analysts' earnings forecasts incorporate information in prior stock price changes? Journal of Accounting and Economics 14: 147-165.

Abarbanell, J. and V. Bernard. 1992. Tests of analysts' overreaction/underreaction to earnings information as an explanation for anomalous stock price behavior. The Journal of Finance 47: 1181-1207.

Affeck-Graves, J., R. Davis, and R. Mendenhall. 1990. Forecasts of earnings per share: possible sources of analyst superiority and bias. Contemporary Accounting Research 6: 501-517.

Ali, A., A. Klein, and J. Rosenfeld. 1992. Analysts' use of information about permanent and transitory earnings components in forecasting annual EPS. The Accounting Review 87: 183-198.

Amir, E. and T. Sougiannis. 1999. Analysts' Interpretation and Investors' Valuation of Tax Carryforwards. Contemporary Accounting Research 16: 1-21.

Ball, R. and E. Bartov. 1996. How na e is the stock market's use of earnings information? Journal of Accounting and Economics 21: 319-337.

Barefield, R. and E. Comiskey. 1975. The accuracy of analysts' forecasts of earnings per share. Journal of Business Research 3: 241-252.

Barth, M. and A. Hutton. 2001. Financial analysts and the pricing of accruals. Working Paper, Standford University.

Benesh, G. and P. Peterson. 1986. On the relation between earnings changes, analysts' forecasts and stock price fluctuations. Financial Analysts Journal 42: 29-40.

Bernard, V. and J. Thomas. 1990. Evidence that stock prices do not fully reflect the implications of current earnings for future earnings. Journal of Accounting and Economics 13: 305-340.

Bernard, V., J. Thomas and J. Wahlen. 1997. Accounting-based stock price anomalies: Separating market inefficiencies from risk. Contemporary Accounting Research 14: 89-135.

Bernhardt D., M. Campello and E. Kutsoati. 2002. Who Herds? Working Papers, University of Illinois.

Bhattacharya, N. 2001. Investors' trade size and trading responses around earnings announcements: an empirical investigation. The Accounting Review 76: 221-244.

Bradshaw, M. and S. Richardson. 2001. Do analysts and auditors use information in accruals. Journal of Accounting Research 39: 45-74.

Brown, L. 1991. Forecast selection when all forecasts are not equally recent. International Journal of Forecasting 7: 349-356.

Brown, L. 1993. Earnings forecasting research: its implications for capital markets research. International Journal of Forecasting 9: 295-320.

Brown, L. 1998. Managerial behavior and the bias in analysts' earnings forecasts. Working paper, Georgia State University.

Brown, L., R. Hagerman, P. Griffin, and M. Zmijewski. 1987a. Security analyst superiority relative to unwanted time-series models in forecasting. Journal of counting and Economics 9: 61-88.

Brown, L., R. Hagerman, P. Griffin, and M. Zmijewski. 1987b. An evaluation of alternative proxies for the market's assessment of unexpected earnings. Journal of Accounting and Economics 9: 159-194.

Burgstahler, D. and M. Eames. 2000. Management of earnings and analyst forecasts. Working paper. University of Washington.

Chen, X. and Q. Cheng. 2002. Institutional holding and analysts' stock recommendations. working paper. University of Chicago.

Clayman, M. and R. Schwartz. 1994. Falling in love again-analysts' earnings and reality. Financial analysts Journal 50: 66-68.

Clement, M. 1999. Analyst forecast accuracy: do ability, resources, and portfolio complexity matter? Journal of Accounting and Economics 27: 285-303.

Collins, W. and W. Hopwood. 1980. A multivariate analysis of annual earnings forecasts generated from quarterly forecasts of financial analysts and univariate time-series models. Journal of Accounting Research 18: 390-406.

Das, S., C. Levine, and K. Sivaramakrishnan. 1998. Earnings predictability and bias in analysts' earnings forecasts. The Accounting Review 73: 277-294.

DeBondt, W. and R. Thaler. 1990. Do security analysts overreact? The American Economic Review 80: 52-57.

Dechow, P., A. Hutton, and R. Sloan. 1999. An empirical assessment of the residual income valuation model. Journal of Accounting and Economics 26: 1-34.

Degeorge, F., J. Patel, and R. Zeckhauser. 1999. Earnings management to exceed thresholds. Journal of Business 72: 1-33.

Dowen, R. 1996. Analyst Reaction to Negative Earnings for Large Well-Known Firms. Journal of Portfolio Management 23: 49-55.

Dreman, D. and M. Berry. 1995. Analyst forecasting errors and their implications for security analysis. Financial Analysts Journal 51: 30-41.

Eames, M. and D. Burgstahler. 2003. Earnings management to avoid losses and earnings decreases: Are analysts fooled? Contemporary Accounting Research 20: 253.

Eames, M., S. Glover, and J. Kennedy. 2002. The association between trading recommendations and broker-analysts' earnings forecasts. Journal of Accounting Research 40: 85-103.

Easterwood, J. and S. Nutt. 1999. Inefficiency in analysts' earnings forecasts: systematic misreaction or systematic optimism? The Journal of Finance 54: 1777-1797.

Elgers, P., M. Lo, and R. Pfeiffer. 2001. Delayed security price adjustments to financial analysts' forecasts of annual earnings. The Accounting Review 76: 613-632.

Elliott, J., D. Philbrick, and C. Wiedman. 1995. Evidence from archival data on the relation between security analysts' forecast errors and prior forecast revisions. Contemporary Accounting Research 11: 919-938.

Elton, E., M. Gruber, and M. Gultekin. 1984. Professional expectations: accuracy and diagnosis of errors. The Journal of Financial Quantitative Analysis 19: 351-363.

Francis, J. and D. Philbrick. 1993. Analysts' decision as products of a multi-task environment. Journal of Accounting Research 31: 216-230.

Freeman, R. and S. Tse. 1989. The multi-period information content of accounting earnings: confirmations and contradictions of previous earnings reports. Journal of Accounting Research 27: 49-84.

Fried, D. and D. Givoly. 1982. Financial analysts' forecasts of earnings: a better surrogate for market expectations. Journal of Accounting and Economics 4: 85-107.

Givoly, D. 1985. The formation of earnings expectations. The Accounting Review 60: 372-386.

Givoly, D. and J. Lakonishok. 1979. The information content of financial analysts' forecasts of earnings. Journal of Accounting and Economics 2: 165-185.

Givoly D. and J. Lakonishok. 1984. Properties of analysts' forecasts of earnings: A Review and Analysis of the Research. Journal of Accounting Literature 3: 119-152.

Gu, Z. and J., Wu. 2000. Earnings skewness and analyst forecast bias. Working paper. University of Rochester.

Hong, H., T. Lim, and J. Stein. 2000. Bad news travels slowly: size, analyst coverage, and the profitability of momentum strategies. The Journal of Finance 55: 265-295.

Hopwood, W. and J. Mckeown. 1990. Evidence on surrogates for earnings expectations within a capital market context. Journal of Accounting, Auditing, and Finance 5: 339-368.

Hughes, J. and W. Ricks. 1987. Associations between forecast errors and excess returns near to earnings announcements. Accounting Review 62: 158-175.

Ip, G. 1997a. Traders laugh off the official estimate on earnings, act on whispered numbers. Wall Street Journal (January): C.1.

Ip, G. 1997b. Rise in profit guidance dilutes positive surprises. Wall Street Journal (June): C.1.

Jacob, J., T. Lys, and M. Neale. 1999. Expertise in forecasting performance of security analysts. Journal of Accounting and Economics 28: 51-82.

Kim, O., S. Lim, and K. Shaw. 2001. The inefficiency of the mean analyst forecast as a summary forecast of earnings. Journal of Accounting Research 39: 329-335.

Lim, T., 1998. Are analysts' forecasts optimistically biased? Working paper, Dartmouth. University.

Lim, T. 2001. Rationality and analysts' forecast bias. The Journal of Finance 56: 369-385.

Lin, H. and M. McNichols. 1998. Underwriting relationships, analysts' earnings forecasts and investment recommendations. Journal of Accounting and Economics 25: 101-127.

Lys, T. and S. Shon. 1990. The association between revisions of financial analysts' earnings forecasts and security price change. Journal of Accounting and Economics 13: 341-364.

Matsumoto, D. 2001. Management's incentives to avoid negative earnings surprises. Working Paper. University of Washington.

McDonald, C. 1973. An empirical examination of the reliability of published predictions of future earnings. The Accounting Review 48: 502-510.

McGee, S. 1997. As stock market surges ahead, 'predictable' profits are driving it. Wall Street Journal (May): C.1.

McNichols, L. and P. O'Brien. 1997. Self-selection and analysts coverage. Journal of Accounting Research 35: 167-199.

Mikhail, M., B. Walther, and R. Willis. 1997. Do security analysts improve their performance with experience? Journal of Accounting Research 35: 131-157.

Mikhail, M., B. Walther and R. Wills. 1999. Does forecast accuracy matter to security analysts? The Accounting Review 74: 185-200.

Mikhail, M., B. Walther and R. Wills. 2003. Security analyst experience and post-earnings-announcement drift. Journal of Accounting, Auditing, and Finance 18: 529-550.

Mendenahll, R. Evidence on the possible underweighting of earnings-related. Journal of Accounting Research 29: 170-179.

Mozes H. and P. Williams. 1999. Modeling earnings expectations based on clusters of analyst forecasts. Journal of Investing 8: 25-38.

O'Brien, P. 1988. Analysts' forecasts as earnings expectations. Journal of Accounting and Economics 10: 53-83.

Ramakrishnan, R. and R. Thomas. 1998. Valuation of permanent, transitory, and price-irrelevant components of reported earnings. Journal of Accounting, Auditing, and Finance 13: 301-336.

Rendleman, R., C. Jones, and H. Latane. 1987. Further insight into the standardized unexpected earnings anomaly: size and serial correlation effects. Financial Review 22: 131-144.

Richardson, S. S. Teoh, and P. Wysocki. 1999. Tracking analysts' forecasts over the annual earnings horizon: Are analysts' forecasts optimistic or pessimistic? Working paper, MIT.

Richardson, S., S. Teoh, and P. Wysocki. 2001. The walkdown to beatable analyst forecasts: the roles of equity issuance and insider trading incentives. Working paper, University of Michigan.

Schipper, K. 1991. Commentary on analysts' forecasts. Accounting Horizons 5: 105-121.

Shane, P. and P. Brous. Investor and (Value Line) Analyst Underreaction to Information about Future Earnings: The Corrective Role of Non-Earnings-Surprise Information. Journal of Accounting Research 39: 387-404.

Shroff, P., R. Venkataraman and B. Xin. 2003. Leaders and followers among security analysts: Analysis of impact and accuracy. Working paper, University of Minnesota.

Sougiannis, T. and T. Yaekura. 2000. The accuracy and bias of equity values inferred from analysts' earnings forecasts. Working paper, University of Illinois.

Stickel, S. 1991. Common stock returns surrounding earnings forecast revisions: More puzzling evidence. Accounting Review 66: 402-416.

Stickel, S. 1995. The anatomy of the performance of buy and sell recommendation. Financial analysts Journal 51: 25-39.

Teoh, S., I. Welch, and T. Wong. 1998. Earnings management and the underperformance of seasoned equity offerings. Journal of Financial Economics 50: 63-99.

Trueman, B. 1994. Analyst forecasts and herding behavior. The Review of Financial Studies 7: 97-124.

Walther, B. 1997. Investor sophistication and market earnings expectations. Journal of Accounting Research 35: 157-179.

Wiedman, C. 1996. The relevance of characteristics of the information environment in the selection of a proxy for the market's expectations for earnings: an extension of Brown, Richardson, and Schwager [1987]. Journal of Accounting Research 34: 313-324.

Zitzewitz, E. 2001. Measuring herding and exaggeration by equity analysts and other opinion sellers. Working paper, Stanford University

Do-Jin Jung, West Texas A&M University
Table 1: Analysts' Superior Forecast Accuracy

 Databases Time-series Models


et al.
(1978) X X X

Brown &
(1978) X X X

Brown &
(1980) X X X X

Collins &
(1980) X X X X

Fried &
(1982) X X X

et al.
(1987a) X X X X

et al.
(1987c) X X X

(1988) X X

 Forecast Errors
Studies Period [absolute [(F-A).sup.2] [absolute [(VT/VA).
 value of value of sup.2]
 (F-A)] (F-A)/A]

et al.
(1978) 1967-76 X

Brown &
(1978) 1972-75 X

Brown &
(1980) 1973-76 X

Collins &
(1980) 1951-70 X

Fried &
(1982) 1969-79 X

et al.
(1987a) 1975-80 X

et al.
(1987c) 1977-82 X

(1988) 1975-82 X X

These studies consistently provide evidence that analysts' earnings
forecast errors are smaller than predictions derived from time-series
models of earnings.

Forecast Databases: (EF) Earnings Forecaster published by Standard
and Poor's, (VL) Value Line, (IB) I/B/E/S.

Time-series Models: (NA) Naive model, (MA) Moving Average model, (MT)
Martingale model, (BJ) Box-Jenkins (1970) model, (BR) Brown-Rozeff
(1979) model, (GW) Griffin (1977)-Watts (1975) model, (FO) Foster
(1977) model, (IN) Index model, (RW) Random Walk model.

Forecast Errors: (F) Forecasts of Earnings, (A) Actual Earnings, (VT)
Variance of Time-Series Models' Forecast Errors, (VA)

Variance of Analysts' Forecast Errors.

Table 2: Limitations of Analyst Forecast Accuracy

Studies Forecast Sample Forecast Error
 Database Period

Crichfield et al. (1978) Earnings 1967-76 [(F-A).sup.2]
Imhoff & Pare (1982) Earnings 1971-74 [absolute value
 of A-F]
 Forecaster [absolute value
 of (A-F)/A]
 [absolute value
 of (A-F)/F]

Brown et al. (1987c) Value Line 1977-82 [(VT/VA).sup.2]

O'Brien (1990) I/B/E/S 1975-81 [absolute value
 of A-F]

Butler & Lang (1991) I/B/E/S 1983-86 (|F-A|-M|)/P

Stickel (1992) Zacks 1981-85 [absolute
 value of A-F]

Sinha et al. (1997) I/B/E/S 1984-93 [absolute
 value of A-F]

Mikhail et al. (1997) Zacks 1980-95 [absolute
 value of

Clement (1999) IBES 1983-94 (|A-F|-M)/M

Jacob et al. (1999) Zacks 1981-92 |A-F|/M

 Association with Forecast Accuracy

Studies Forecast Forecast Forecast
 Horizon Frequency Recency

Crichfield et al. (1978) --
Imhoff & Pare (1982) --
Brown et al. (1987c) --
O'Brien (1990) --
Butler & Lang (1991)
Stickel (1992) -- --
Sinha et al. (1997) -- --
Mikhail et al. (1997) --
Clement (1999)
Jacob et al. (1999) --

 Association with Forecast Accuracy

Studies FirmSize Analyst Brokerage
 Coverage Size

Crichfield et al. (1978)
Imhoff & Pare (1982)
Brown et al. (1987c) --
O'Brien (1990)
Butler & Lang (1991)
Stickel (1992)
Sinha et al. (1997)
Mikhail et al. (1997) --
Clement (1999) --
Jacob et al. (1999) --

 Association with Forecast Accuracy

Studies Analyst Task Industry
 Experience Complexity Concentration

Crichfield et al. (1978)
Imhoff & Pare (1982)
Brown et al. (1987c)
O'Brien (1990)
Butler & Lang (1991)
Stickel (1992)
Sinha et al. (1997)
Mikhail et al. (1997) -- no
Clement (1999) -- --
Jacob et al. (1999) no -- --

Table 3: Bias in Analysts' Earnings Forecasts

Studies Forecast Sample (Time Period) Bias Test
 Databases Models

McDonald Wall Street 201 forecasts for 152 (A-F)/F
(1973) Journal firms (1966-70)

Barefield & Earnings 100 firms on NYSE (F-A)/F
Comiskey Forecaster (1967-72)

Crichfield Earnings 130 forecasts for 46 A = [alpha]
et al. (1978) Forecaster firms (1967-76) + [beta]
 F + e

Fried & Givoly Earnings 1,247 firm-years with (A-F)/
(1982) Forecaster 6,020 forecasts for 424 [absolute
 firms on NYSE (1969-79) value of A]

Givoly (1985) Earnings 1,247 firm-years with A = [alpha]
 Forecaster 6,020 forecasts for 424 + [beta]
 firms on NYSE (1969-79) F + e

O'Brien (1988) I/B/E/S 184 firms with 1,260 (A-F)
 firm-years (1975-82)

Abarbanell Value Line 1370 forecasts for 100 (F-A)
(1991) firms (1981-84)

Ali et al. I/B/E/S 5,365 Firm-Years A-F/P
(1992) (1978-89)

Francis & Value Line 918 annual and first (A-F)/P(orF)
Philbrick quarter forecasts
(1993) (1987-89)

Kang et al. Value Line 132 firms with 6,538 A = [alpha]
(1994) forecasts (1980-85) + [beta]
 F + e

Das et al. Value Line 1,195 firm-year with (A-F)/P
(1998) 239 firms (1989-93)

Easterwood & I/B/E/S 10,694 firm-year with (A-F)/P
Nutt (1999) 1,608 firms (1982-95)

Richardson et I/B/E/S 25,623 firm-year with (A-F)/P
al. (2001) 681,413 forecasts

Studies Findings

McDonald The mean of the relative forecast errors
(1973) is -13.6%. The number of overpredictions is
 65.3% of the total forecasts.

Barefield & Earnings forecasts exceeding actual earnings
Comiskey are approximately 64.9% of the total
(1975) forecasts.

Crichfield [alpha] and [beta] are not significantly
et al. (1978) different from zero and one, respectively, at
 the .10 level.

Fried & Givoly Analysts' forecast bias is present only in 6 of
(1982) the 11 years and appears to be quite small,
 except for the first three years. The mean of
 the relative forecast errors is -5.3%,
 indicating some tendency for analysts to
 overestimate next year's earnings.

Givoly (1985) [alpha] (-.023) and [beta] (1.009) are very close
 to 0 and 1. The mean of the relative forecast
 errors is 1.2% and insignificant.

O'Brien (1988) The mean of the forecast errors exhibit
 significantly negative bias, but the median is
 indistinguishable from zero.

Abarbanell The mean of the relative forecast errors is
(1991) $.04 and significant at

Ali et al. Both long-term (eight-month) and short-term
(1992) (one-month) forecasts are, on average,
 upwardly biased.

Francis & The mean of the relative forecast errors
Philbrick is--$.18 per share and -9.0% of the EPS
(1993) forecasts, significantly less than zero at the
 .01 level.

Kang et al. F-statistics under OLS and
(1994) [chi square]-statistics under GMM reject the
 null hypothesis of no bias at the .01 level.

Das et al. The mean of the relative forecast errors is
(1998) significantly negative (from -1.5% to -3.5%)
 over all four horizons.

Easterwood & The mean (median) of the relative forecast
Nutt (1999) errors is -1.93% (-.32%) and significant.

Richardson et The mean (median) of the relative forecast
al. (2001) errors is -.90% (.28%) and significant.

(A) Actual Earnings, (F) Forecasts of Earnings, (P): Stock Prices.
COPYRIGHT 2005 The DreamCatchers Group, LLC
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2005 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Jung, Do-Jin
Publication:Academy of Banking Studies Journal
Date:Jan 1, 2005
Previous Article:Bank loans and corporate debt ownership: some evidence from an economy in transition.
Next Article:The evolution of derivative use by community banks from 1995-2003.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters