Apr 22

Sample Regression Function (SRF) - 8 minute screencast

by David Harper, CFA, FRM, CIPM

FRM | Quant Methods |

srfThumb

 

This is a brief screencast to explain the terminology around the sample regression function (SRF). The stochastic population regression function (PRF) is theoretical: we don't access it directly. We infer it based on samples; each sample produces its own sample regression function (SRF). So, there is a many-to-one relationship. Note the PRF error term corresponds to the SRF residual. The residual is an key part of the SRF: the OLS produces a line that minimizes the sum of squared residuals:

prtDetail

 

Here is the screencast:

Comments

  1. Be the first to leave a comment!

Leave a Comment