Stats › Options, shorts and positioning › Analyst optimism
Analysts are optimistic at every horizon and a statistical model is unbiased at every horizon
| Horizon | Analyst bias | t | Model error | Price scaled gap |
|---|---|---|---|---|
| One quarter | 0.028 | 6.59 | -0.000 | 0.012 |
| Two quarters | 0.053 | 10.31 | -0.001 | 0.024 |
| Three quarters | 0.072 | 11.55 | 0.002 | 0.031 |
| One year | 0.154 | 6.24 | 0.027 | 0.057 |
| Two years | 0.384 | 8.33 | -0.004 | 0.190 |
Analyst bias is the forecast minus realised earnings per share, 1986 to 2019, with autocorrelation corrected t statistics. The model column is the same quantity for a machine learning forecast built on the same information set. The price scaled gap is the difference between the two divided by price, and all five values are statistically different from zero.
Measured against realised earnings, analyst forecasts are too high at one quarter, two quarters, three quarters, one year and two years, with t statistics from 6.24 to 11.55. A machine learning benchmark on the same data has errors indistinguishable from zero.
Members only
The rest of this page is for members
Below this point there are 6 sections, 1 table and 3 named sources, roughly 900 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.