Skip to main content

Table 9 Cross-sectional models with analysts’ forecasts

From: Earnings forecasts: the case for combining analysts’ estimates with a cross-sectional model

  Bias Et+1 Bias Et+2 Bias Et+3
Mean Median Mean Median Mean Median
Panel A: Bias of cross-sectional models with analysts forecasts
CM − 0.006 0.005 − 0.009 0.000 − 0.005 − 0.001
[− 1.08] [1.99] [− 1.06] [0.09] [− 0.38] [− 0.33]
HVZ 0.016 0.003 − 0.008 − 0.001 − 0.024 − 0.009
[4.41]\(**\) [1.48] [1.3] [0.46] [1.76] [1.66]
EP 0.005 0.002 − 0.023 − 0.007 − 0.043 − 0.014
[1.6] [1.07] [2.72]\(*\) [2.11]\(*\) [2.1]\(*\) [1.74]
RI 0.001 0.001 − 0.012 − 0.003 − 0.020 − 0.007
[0.21] [0.46] [2.08]\(*\) [1.32] [1.82] [1.47]
  Accuracy Et+1 Accuracy Et+2 Accuracy Et+3
Mean Median Mean Median Mean Median
Panel B: Accuracy of cross-sectional models with analysts forecasts
CM 0.046 0.015 0.063 0.026 0.070 0.033
HVZ 0.067 0.016 0.067 0.027 0.080 0.036
EP 0.059 0.016 0.077 0.030 0.094 0.040
RI 0.058 0.015 0.072 0.028 0.080 0.036
  ERC Et+1 ERC Et+2 ERC Et+3
ERC Adj R-Squared ERC Adj R-Squared ERC Adj R-Squared
Panel C: ERC of cross-sectional models with analysts forecasts
CM 0.132 0.016 0.130 0.017 0.098 0.009
[13.12]\(**\)   [5.72]\(**\)   [6.75]\(**\)  
HVZ 0.112 0.013 0.112 0.014 0.075 0.006
[7.65]\(**\)   [5.09]\(**\)   [6.73]\(**\)  
EP 0.132 0.016 0.111 0.013 0.070 0.005
[6.53]\(**\)   [5.18]\(**\)   [4.97]\(**\)  
RI 0.129 0.015 0.117 0.014 0.087 0.006
[6.99]\(**\)   [5.68]\(**\)   [10.61]\(**\)  
  1. This table summarizes the mean and median forecast bias (Panel A), accuracy (Panel B), and ERC (Panel C) for the HVZ, EP, and RI cross-sectional models including analysts’ forecasts instead of earnings from Compustat. We also include the results from the CM in order to facilitate the comparison among the models. Bias is defined as the difference between earnings forecasts and actual earnings, scaled by the firm’s end-of-June market equity and we define accuracy as the absolute difference between actual earnings and earnings forecasts, scaled by the firm’s end-of-June market equity. The ERC is estimated by regressing the sum of the quarterly earnings announcement returns (market-adjusted, from day \(-1\) to day \(+1\)) over the next one-, two-, and three-years on firm-specific unexpected earnings (i.e., the forecast bias) measured over the same horizon. We standardize the unexpected earnings and the returns to make the ERC comparable among all models. We estimate one-, two-, and three-year ahead forecast for the periods 1985–2015, 1987–2015, and 1989–2015, respectively. The Newey-West t-statistics are presented in brackets. \(**\) and \(*\) denote significance at 0.01 and 0.05 levels, respectively