Size Premium is Real and it is Small

Today Block, the company that produces CashApp and Square, was announced to being added to the S&P 500®. As I am writing the stock is up over 10%. Several weeks ago, Coinbase was added and its stock reacted similarly to the announcement. The reason is in large part because large index funds that attempt to mimic the performance of the S&P 500® need to add the new entrant's stock to their holdings to remain true to their purpose.

What does this have to do with size premium? The comparative group is the Center for Research on Securities Prices large company stock index. That index now tracks the S&P 500®. CRSP has capitalization deciles based on the capitalization of New York Stock Exchange listed domestic companies. Size premiums have been developed comparing the performance of the Large Company Index to the decile that most closely corresponds to a company that is being valued.

Some say that size premium disappeared because of articles written about it and thus causing markets to adjust, because efficient markets will not allow outsized returns to continue. They are wrong. The premiums published by Ibbotson® in their Stocks, Bonds, Bills, and Inflation® (SBBI®) Yearbooks used arithmetic means to compute their recommended premiums for many years. They may be appropriate for investments expected to be held one year or less, but for almost all private companies, the expected holding period is multiple years. The comparison should use geometric or regression means that ALWAYS produce lower premiums than arithmetic means.

A second part of the problem is with the data. The pre-World War II data for the smallest decile shows rates of return that are not representative of the now 80 years of post-war returns. The decline in premiums is not the result of some article written by an academic, just an artifact of early data not being truly representative of the current condition in the markets.

On the other hand, the argument that no size premium exists denies the obvious reality that some does. Block and Coinbase in 20 years will be valued on their performance and prospects. The returns that they obtained due to their UPCOMING addition to the S&P 500® is in effect an acceleration of returns that will ultimately be realized by those companies while in the index and will, all other things being equal, reduce the performance of the S&P 500®. The denial of the existence of a size premium by academics with big followings does not enhance their credibility.

The truth is in the data. My computation of Decile 10 size premium based on 1946-2024 stock returns is 1.27%, not ridiculously large numbers proposed by some data providers, but not zero either.

S&P 500® is a registered trade mark of S&P Global Ratings

Ibbotson®, Stocks, Bonds, Bills, and Inflation® and SBBI® are registerd trademarks of Kroll, LLC