A Re-Examination of Financial Analysts' Differential Earnings Forecast Accuracy*

Abstract. This research re-examines whether there are differences in the forecast accuracy of financial analysts through a comparison of their annual earnings per share forecasts. The comparison of analyst forecast accuracy is made on both an ex post (within sample) and an ex ante (out of sample) basis. Early examinations of this issue by Richards (1976), Brown and Rozeff (1980), O'Brien (1987), Coggin and Hunter (1989), O'Brien (1990), and Butler and Lang (1991) were ex post and suggest the absence of analysts who can provide relatively more accurate forecasts over multiple years. Contrary to the results of prior research and consistent with the belief in the popular press, we document that differences do exist in financial analysts' ex post forecast accuracy. We show that the previous studies failed to find differences in forecast accuracy due to inadequate (or no) control for differences in the recency of forecasts issued by the analysts. It has been well documented in the literature that forecast recency is positively related to forecast accuracy (Crichfield, Dyckman, and Lakonishok 1978; O'Brien 1988; Brown 1991). Thus, failure to control for forecast recency may reduce the power of tests, making it difficult to reject the null hypothesis of no differences in forecast accuracy even if they do exist. In our analysis, we control for the differences in recency of analysts' forecasts using two different approaches. First, we use an estimated generalized least squares estimation procedure that captures the recency-induced effects in the residuals of the model. Second, we use a matched-pair design whereby we measure the relative forecast accuracy of an analyst by comparing his/her forecast error to the forecast error of another randomly selected analyst making forecasts for the same firm in the same year on or around the same date. Using both approaches, we find that differential forecast accuracy does exist amongst analysts, especially in samples with minimum forecast horizons of five and 60 trading days. We show that these differences are not attributable to differences in the forecast issuance frequency of the financial analysts. In sum, after controlling for firm, year, forecast recency, and forecast issuance frequency of individual analysts, the analyst effect persists. To validate our findings, we examine whether the differences in the forecast accuracy of financial analysts persist in holdout periods. Analysts were assigned a “superior” (“inferior”) status for a firm-year in the estimation sample using percentile rankings on the distribution of absolute forecast errors for that firm-year. We use estimation samples of one- to four-year duration, and consider two different definitions of analyst forecast superiority. Analysts were classified as firm-specific “superior” if they maintained a “superior” status in every year of the estimation sample. Furthermore, they were classified as industry-specific “superior” if they were deemed firm-specific “superior” with respect to at least two firms and firm-specific “inferior” with respect to no firm in that industry. Using either definition, we find that analysts classified as “superior” in estimation samples generally remain superior in holdout periods. In contrast, we find that analysts identified as “inferior” in estimation samples do not remain inferior in holdout periods. Our results suggest that some analysts' earnings forecasts should be weighted higher than others when formulating composite earnings expectations. This suggestion is predicated on the assumption that capital markets distinguish between analysts who are ex ante superior, and that they utilize this information when formulating stock prices. Our study provides an ex ante framework for identifying those analysts who appear to be superior. When constructing weighted forecasts, a one-year estimation period should be used because we obtain the strongest results of persistence in this case.     

Systemic Risk and International Portfolio Choice

Returns on international equities are characterized by jumps; moreover, these jumps tend to occur at the same time across countries leading to systemic risk. We capture these stylized facts using a multivariate system of jump‐diffusion processes where the arrival of jumps is simultaneous across assets. We then determine an investor's optimal portfolio for this model of returns. Systemic risk has two effects: One, it reduces the gains from diversification and two, it penalizes investors for holding levered positions. We find that the loss resulting from diminished diversification is small, while that from holding very highly levered positions is large.     

HEALTH CARE IN THE UNITED STATES: WHY IS SUPPLY SO PRICE INSENSITIVE?

Health spending as a percentage of gross domestic product in the U.S. economy is growing, from 5% in 1960 to about 16% in the current period, and it is predicted to grow to as much as 30% in 2050. Then why is the supply of health care in the United States so insensitive to steeply rising prices? This paper conducts an econometric study to show that high health‐care costs have an adverse impact on labor productivity, causing a negative production externality in all industries. So, can the rising cost of health‐care affect the U.S. comparative advantage? The paper seeks answers to these questions in a general equilibrium model and finds that the labor productivity shock is responsible for the sluggish or declining supply of health care. Consumers are able to afford less health care due to a possible decline in real wages. U.S. comparative advantage becomes a nonissue, provided that the equilibrium is stable in spite of a negatively sloped health‐care supply curve. Negative externality, leading to market failure, may be addressed in two alternative ways. (JEL F11, I11, I12, I18)     

FURTHER ANALYSIS OF THE TWIN DEFICITS

In a 1989 Contemporary Policy Issues article Miller and Russek published findings of a causal relation between the fiscal deficit and the trade deficit. However, they found no overwhelming support for reverse causation between the twin deficits. The authors of the analysis here gathered annual data on U.S. federal budget deficits and net exports for 1950–1988 and deflated the nominal values by the GDP deflators to examine the causal relation based on real values. They made a distinction between structural and actual budget deficits. Instead of an arbitrary choice of lag structure, they used Hsiao (1979, 1981) minimum final prediction error criterion to determine the optimum lag lengths of the explanatory variables. The analysis reveals a unidirectional causal relation running from structural budget deficits to net exports, confirming some of Miller and Russek's findings. Contrary to Miller and Russek's conclusions, however, findings here indicate a bi-directional causal relation between actual budget deficits and net exports. These findings suggest important policy implications.     

NON‐COOPERATIVE ENTRY DETERRENCE IN LICENSE AUCTIONS: DYNAMIC VERSUS SEALED BID*

We examine the impact of potential entry on incumbent bidding behavior in license auctions, in both dynamic and sealed bid formats. Unlike sealed bid auctions, dynamic auctions reveal information about the identities of potential winners and allow bidders to revise their bids. This helps incumbents to coordinate their entry deterrence efforts. If entry is sufficiently costly for each incumbent, only the dynamic auction has an equilibrium where entry is deterred for sure. Numerical calculations suggest that, regardless of how costly entry is for each incumbent, sealed bid auctions can generate a higher probability of entry as well as a more efficient allocation.     

Analysts' Selective Coverage and Subsequent Performance of Newly Public Firms

This study examines the ability of analysts to forecast future firm performance, based on the selective coverage of newly public firms. We hypothesize that the decision to provide coverage contains information about an analyst's underlying expectation of a firm's future prospects. We extract this expectation by obtaining residual analyst coverage from a model of initial analyst following. We document that in the three subsequent years, initial public offerings with high residual coverage have significantly better returns and operating performance than those with low residual coverage. This evidence indicates analysts have superior predictive abilities and selectively provide coverage for firms about which their true expectations are favorable.     

Did Houston Clients of Arthur Andersen Recognize Publicly Available Bad News in a Timely Fashion?*

Despite the allegations of audit failure and the enormous publicity surrounding Arthur Andersen's indictment, there is no systematic empirical evidence on characteristics of accounting information of clients of Arthur Andersen vis‐à‐vis other Big 6 auditors. I examine whether earnings of Andersen's Houston‐based clients are timely in reporting bad news about future cash flows. I find that relative to a control group consisting of Houston‐based clients audited by other Big 6 auditors, earnings of Andersen clients are less timely in reporting bad news. Further, it appears that operating accruals of Andersen clients are less effective in accelerating the timely recognition of bad news than operating accruals of non‐Andersen clients. The findings suggest that the clients of Andersen's Houston office engaged in aggressive accounting practices, including delayed recognition of publicly available bad news.     

Common Failings: How Corporate Defaults Are Correlated

We test the doubly stochastic assumption under which firms' default times are correlated only as implied by the correlation of factors determining their default intensities. Using data on U.S. corporations from 1979 to 2004, this assumption is violated in the presence of contagion or “frailty” (unobservable explanatory variables that are correlated across firms). Our tests do not depend on the time‐series properties of default intensities. The data do not support the joint hypothesis of well‐specified default intensities and the doubly stochastic assumption. We find some evidence of default clustering exceeding that implied by the doubly stochastic model with the given intensities.