Weighted earnings estimate revisions
factor.formula
First, calculate the size of each institution's last earnings estimate revision released in the past three months.
For each institution, the earnings forecast revision is defined as: the percentage change in the current forecast net profit (the forecast value for a specific reporting period published at the current time) relative to the latest forecast value for the same reporting period published by the institution one month ago within the past six months.
Then, the Accwt2 method is used to take the weighted average of the last earnings forecast revisions of each institution in the past three months to obtain the final weighted earnings forecast revisions.
Detailed explanation:
- :
The earnings forecast value of institution i at time t for quarter q (or reporting period).
- :
The earnings forecast value for the same quarter q (or reporting period) released by institution i one month before time t (i.e., t-1m). This value is the latest forecast value for quarter q released in the past 6 months.
- :
The revision of the earnings forecast of institution i for quarter q at time t is calculated as (E_{i,t,q} - E_{i,t-1m,q})/E_{i,t-1m,q}
- :
The weight of institution i is determined by the Accwt2 method, which is a weighting method based on the historical performance of analysts' forecasts, giving greater weight to analysts with higher forecast accuracy.
- :
The weighted earnings forecast revision at time t is calculated as sum(R_{i,t,q} * w_i) , where sum represents the sum of all analysts
factor.explanation
Traditional earnings revision measures are usually based on changes in consensus forecasts, while this factor can better reflect the heterogeneity of analysts' views and more comprehensively reflect changes in market expectations of company earnings prospects by examining the extent of forecast revisions of each institution and weighting them. Disagreements between analysts may indicate future stock price volatility. Using the Accwt2 weighting method can give higher weights to analysts with better historical forecast performance, thereby increasing the information content of the factor.