Another thought on small-stock premiums

There is another issue I'd like to raise regarding the small-stock premiums. Appraisers usually choose between data from Ibbotson (now Morningstar) or Duff&Phelps, at least the majority do. Although the same logic applies to other data sources as well. Ibbotson calculates their data on equity risk premium and small-stock premiums using long-term average of returns dating back to 1926. Up till last year the long-term equity risk premium was app. 7% (now it is down to 6,5%!). This was then used to calculate CAPM expected rates of return, and the return in excess of CAPM was attributed to small-stock effect, since companies were grouped into size deciles. Similar approach was used by Duff&Phelps, whereas they only use data going back to 1963. Therefore their average equity risk premium was app. 5%. And of course also small-stock premiums were different. 

The point I want to make here is that we need to be consistent when using either of these data sets. We cannot simply use Ibbotson's ERP of 7% and Duff&Phelps data on small-stock premiums and vice versa. This much is usually clear. However, some appraisers heavily rely on Ibbotson for small-stock premiums, while on the other hand disagree with them on the equity risk premium part and decide to use a different one, somewhere in the 4-6% range. I have my doubts whether this is then consistent, since a smaller ERP would also yield higher small-stock premiums, if you observe the past data.