Test anfragen
Two companies, same industry, same leverage, same growth outlook. One has a market cap of 50 billion, the other 500 million. Should an investor demand a higher return from the smaller one, just because it's smaller?
Most practitioners say yes. The intuition holds up reasonably well: smaller companies typically have weaker access to capital, especially in a downturn, when a large company can lean on committed credit lines or public markets that simply aren't available to a smaller, often private, business. Worth a quick distinction: this is about operating risk, not marketability, that's a separate discount, DLOM. Smaller companies also tend to run less diversified operations and a thinner management bench, fewer product lines, fewer markets, more reliance on a small number of people to keep things running. Each of these is a genuine source of risk that CAPM alone doesn't fully capture. Size premium is meant to fill that gap.
So how does the size premium actually get measured? The standard approach is backward looking: rank public companies into buckets by size, market cap most commonly, though some studies use revenue, profitability, or other measures instead, then track their realized historical returns, sometimes across several decades, in some cases as far back as 1926, and measure how much each bucket's actual return exceeded what CAPM alone would have predicted using that bucket's own beta. Whatever's left over becomes the size premium. It's a number built from real historical market data.
Market risk premium doesn't have one standard answer either. Some practitioners still rely on historical averages or surveys, but forward looking, implied models, backed out of current prices and consensus forecasts, have become a credible, well established option alongside them. Size premium hasn't really been given the same treatment. It's still built almost entirely the historical way. Why?
The building blocks for a forward looking size premium exist elsewhere in valuation practice already. In principle, the same implied cost of capital logic used for index level ERP work could be applied by size bucket instead, giving a premium built from where prices sit today rather than decades of realized history.
One concern we came across while exploring the subject: reliable analyst coverage and forecast data thin out quickly once you move down in size, right at the point where the premium is supposed to matter most. Is a forward looking size premium even achievable with usable data at the smaller end, or does the data simply run out before the model does?
None of this makes the forward looking model an obvious choice. It trades one set of problems for another, decades of realized returns swapped for current prices and consensus forecasts that carry their own assumptions about growth and payout. The question isn't which approach is right. It's whether looking at both would tell us something the historical number alone doesn't.
When you're pricing a private company today, how do you approach size premium?
Off a decile table, a rule of thumb, something else? And is that number grounded in data you trust, or more of an inherited market convention?
Do you want to see us explore the forward looking option further?
%20Loans%20in%20Private%20Credit.png)
