A common question we get: "If Buffett were starting today with one million dollars instead of Berkshire's 1 trillion, what would he buy?" The honest answer is that he would buy small caps. Buffett has said this multiple times across his career, most directly in a 1999 Forbes interview where he noted that "I could make you 50 percent a year on a million dollars" using a small-cap-tilted strategy that Berkshire's size now precludes.
The reason is straightforward: Berkshire cannot meaningfully deploy capital into a 500-million-EUR market-cap company. A 5 percent position would be a 25-million-USD purchase, which is nothing relative to Berkshire's 600-billion-dollar equity portfolio but is also a 5 percent ownership stake in the small company, which is too much for Berkshire to acquire without distorting the price.
For a retail investor with 50,000 EUR to 5 million EUR in capital, the small-cap universe is not just accessible; it is the part of the market where the most inefficient pricing exists. This post walks through the Buffett-style small-cap screen with a worked example.
Why small caps are inefficient
Three structural reasons institutional money cannot fully exploit small-cap pricing:
1. Scale constraints. A 10-billion-dollar fund cannot take a meaningful position in a 500-million-dollar company without becoming a 10+ percent shareholder, triggering filing requirements and reducing flexibility. Most large funds simply do not bother.
2. Coverage gaps. A typical 5-billion-USD-market-cap company has 6-10 sell-side analysts covering it. A typical 500-million-USD company has 0-2 analysts. The sparser the coverage, the more dispersed the pricing.
3. Liquidity penalties. Small caps have wider bid-ask spreads and shallower order books, so institutional flow has higher market impact. This compresses what institutions are willing to pay even for high-quality small caps.
These three forces collectively produce the small-cap value premium that Fama and French documented in their famous 1992 paper and that has been replicated in many subsequent studies (with caveats about how robust the premium is in any given decade).
The 5 filters of a Buffett-style small-cap screen
Translating the broader Buffett framework specifically for small caps requires emphasising certain criteria more than others, because small-cap data is sparser and the risks differ.
Filter 1: ROIC above 15 percent for at least 5 of the last 7 years
The single most important filter. Sustained high ROIC is the fingerprint of a real moat, and the moat is what makes a small cap a candidate for long-term compounding rather than a value trap. Small caps without a moat get competed away within 5-10 years.
The 5-of-7 threshold (rather than 7-of-7) allows for one or two cyclical down years that are forgivable. Below 4-of-7 is a warning.