Portfolio Concentration Risk: Why I Underwrite the Shape of the Book

Concentration is the risk that quietly does the damage.

I put short-term capital into many small businesses and am repaid out of their daily sales. On its surface the model is simple, and the instinct is to judge it one deal at a time. Is this business sound? Will it repay? Those are fair questions for a single position. They are the wrong frame for the thing I actually hold, which is a book. The property that most determines how that book behaves is portfolio concentration risk, and it is the exposure I watch before almost anything else.

Why Portfolio Concentration Risk Hides in Plain Sight

Concentration is the risk that quietly does the damage. It never shows up in any single underwriting decision. Every position can look reasonable on its own while the book as a whole carries a fragility that none of them reveals individually. The exposure lives in the distribution, not in the deal, which is precisely why it gets underweighted. Underwriting attention naturally goes to the file in front of you, and concentration is invisible at that level.

Consider two books of the same total size. One is built from many small positions. The other from a handful of large ones. They are not two versions of the same risk. They are different instruments. When one business stumbles in the first book, it is immaterial, lost in the aggregate. When one business stumbles in the second, it leaves a mark I can measure. Same event, two entirely different outcomes, decided by how the capital was distributed.

Put plainly, I can lose one position out of a hundred and the book absorbs it. Lose one out of ten and I feel it. The arithmetic is obvious once it is stated, and yet it is routinely set aside in favor of deal-level conviction.

This is why I treat the questions that matter as structural ones. How many positions. How large is the largest relative to the total. How correlated are the names, by industry, by geography, by the conditions that would pressure them at the same time. A book can be full of sound individual deals and still be poorly built. A book of more ordinary deals, well distributed, can be far more durable.

I still read each business honestly. That work matters. But it sits inside a larger discipline, and the larger discipline is portfolio construction. In the end I am not underwriting individual businesses so much as the shape they make together.