Size premium
DCFAn increment added to a CAPM cost of equity for smaller companies, justified by returns that beta alone did not explain.
Also written: small company premium
CAPM prices exactly one thing: exposure to the market, scaled by beta. The size premium is an extra increment bolted on to its output for smaller companies, so that cost of equity becomes the risk free rate, plus beta times the equity risk premium, plus the premium. It is empirical rather than theoretical. Over long runs of data, smaller companies returned more than their measured betas predicted, and if that holds, beta understates what their shareholders require and the premium adds the difference back by hand.
Several distinct things sit inside the one number, which is both why it is used and why it resists precision. The beta of a thinly traded share is measured downward, because a price that does not update every time the market moves looks less correlated with it than the underlying business is. The model assumes the marginal investor is diversified, which describes a pension fund and not the founder or family who typically own a small company. Illiquidity is absent from the model altogether. And smaller companies fail more often, with a loss that is not symmetrical with the upside.
The evidence is weaker than the practice around it, and saying so is the mark of somebody who has read past the summary. The effect is concentrated in the very smallest companies rather than sloping smoothly across sizes, and it has been less consistent in recent decades than in the studies that established it. Survivorship compounds the doubt: companies that fail stop producing returns, so a series assembled from survivors can overstate what a representative portfolio would have earned.
The data problem is sharper in Europe than most textbooks admit. The published series practitioners quote, the Ibbotson series and the Duff and Phelps studies now published under Kroll, are built on United States returns, and there is no European series of comparable standing and coverage. A European valuation using one is importing another market's history. That is defensible if you say so, and weak if the number is presented as a fact.
Worked example
Illustrative. A 3.0% risk free rate, a beta of 1.2 and a 5.5% equity risk premium give a CAPM cost of equity of 9.6%.
Add an illustrative size premium of 3 points and the rate becomes 12.6%. On cash flow growing at 2% forever, that removes about 28% of the value.
Cutting every cash flow by a fifth, forever, would have removed 20%. The premium is the larger lever, which is why it deserves more argument than it usually receives.