The use of small-stock premium data for companies outside USA

A while ago I wrote an article concerning this topic and I'd like to share some of my findings and proposals from that article.

Empirical studies have shown that smaller companies on average tend to earn a higher rate of return, i.e. investors demand a higher rate of return when investing in small companies as compared to large companies. Such findings have been well documented (and sometimes even linked to the so-called January effect). So, data providers (such as Ibbotson or Duff&Phelps) provide data on small-stock premiums for different size groups, whereas these data are based on data from the USA. 

These size groups (and limits for belonging into each group) may be OK for companies in the USA and even some larger and more internationalized equity markets. However, for smaller countries and markets most of the companies would automatically fall in the group of smallest companies, and we would have to use (in some cases an absurdly) high small-stock premium, even for companies that are generally among the largest in a certain market. The argument in favour of using these premiums directly is that equity markets are globally linked and there is a complete and rather cheap way to move capital around, so investors should use the same criteria when choosing to invest in any market. But really, in a very small market - how many international investors are there and what's their share as compared to domestic investors? Furthermore, I haven't seen many research reports by local analysts following the stock market that use small stock premium in their calculation of the rate of return at all.

This is why I believe that the use of data on small-stock premiums based on US data should be adjusted to reflect the characteristics of a local market. Of course, it is up to the appraiser to decide what he thinks is the right approach. So we have the following options:

1. Simply use US data and their size limits to define size groups.

2. Arbitrarily decide the size of the small-stock premium we will use in each case, although this is a completely subjective approach.

3. Redefine the size limits using the data from the local stock exchange. 

The latter approach is the one I believe to be the most appropriate. Of course, as far as the market value is concerned it is finally up to the market participants to define which approach they will use and this is what will determine the market value in the end, be it the "right" way or not.