Analyst characteristics, timing of forecast revisions, and analyst forecasting ability

Abstract

a b s t r a c tWe first examine whether analysts with certain characteristics that prior research has identified are related to superior forecasting ability systematically time their forecast revisions later in the fiscal quarter. We then examine whether this superior ability persists after controlling for the timing advantage by using relative forecast error, a measure that largely eliminates the timing advantage of recent forecasts. Using a sample of quarterly earnings forecast revisions over the 20-year period from 1990 to 2009, we find that analysts with more firm-specific and general experience and more accurate prior-period forecasts, analysts employed by larger brokerage firms, and analysts who follow fewer industries and companies tend to revise forecasts later in the quarter. We also find that analyst characteristics that are positively correlated with revision timing are negatively related to relative forecast errors. These results are consistent with analyst characteristics being useful proxies for analyst forecasting ability and analysts with greater ability revising forecasts later in the quarter.

Do analysts who understand accounting conservatism exhibit better forecasting performance?

“…A significant and positive coefficient on MATCH L (0.008, t ‐statistic = 2.11) suggests that analysts’ timeliness‐matching performance has incremental explanatory power beyond control variables such as past forecast accuracy and past forecast bias in explaining forecast accuracy of the current earnings forecasts. Most of the estimated coefficients on other control variables are in line with prior studies (e.g., Kim, Lobo, & Song, ; and Kumar, ). For instance, we find a negative and significant coefficient (−0.055, t ‐statistic = −11.69) on HORIZON , which indicates that analysts provide more accurate forecasts as the actual earnings announcement date comes closer.…”

Ambiguity Aversion, Company Size and the Pricing of Earnings Forecasts

“…Several authors have documented an aggregate cross sectional size effect (Banz, 1981; Reinganum, 1981; Fama and French, ; Amel‐Zadeh, ). Under our hypothesis this size effect may also be partly related to ambiguity .…”

“…This notion of resolution implies that analyst forecasts issued later in the quarter are more informative, helping the market form more accurate earnings expectations. Kim et al () suggest that this is indeed the case because analysts who delay their forecasts observe a larger information set that includes the forecasts of other analysts, reports or news releases, conference calls or even management access (see also Ivkovic and Jegadeesh, ). To examine whether late forecasts are relatively more informative we divide our sample into quartiles according to forecast horizon and run the regression shown by Equation 5 separately for each group.…”

The case for humble expectations: CEO humility and market performance

“…Analysts' EPS expectations : Analysts are important stakeholders in the marketplace and evaluate firms' future growth potential by setting their own performance forecasts for individual firms. Since analysts set market expectations for firms, extant finance and accounting literatures typically capture market performance expectations by aggregating all analysts' ratings following a given firm (e.g., Hong & Kubik, ; Kim, Lobo, & Song, ). Consistent with prior research, we operationalize analysts' EPS expectations as the I/B/E/S last analysts' consensus forecast for our focal firm's EPS for the year ( t ) and scaling it by the year‐end firm's stock price for a given year.…”