Where diversification actually pays

A portfolio of twenty businesses that all depend on the same construction cycle is a single position expressed twenty times. Diversification only reduces risk where the underlying drivers of performance are genuinely different, and the number of holdings tells you almost nothing about that.

We think about it in terms of what would have to be true for two holdings to disappoint at the same time. If the answer is a shared customer base, a shared input cost, a shared regulatory regime or a shared funding market, the two are correlated regardless of how different their industries appear on paper.

Where diversification genuinely pays is in cash flow timing. Businesses that generate cash at different points in their own cycles allow the group to fund investment internally, which is when capital is cheapest and most useful. That is a structural advantage, not a statistical one, and it is the main reason we hold across sectors rather than concentrating in the one that looks best today.

It also pays in knowledge. Operating across industrial, real asset and services businesses surfaces practices in one holding that transfer usefully to another. Procurement discipline, maintenance planning and pricing architecture are not sector specific, and a diversified group sees more variations of each than a specialist ever will.

What diversification does not do is substitute for quality. Spreading capital across weak positions produces a reliably mediocre result. We would rather hold a smaller number of businesses we understand completely, chosen so that their fortunes do not rise and fall together.