Concentration, diversification, and the honest trade
Portfolio construction · Hedge Fund Manager Blueprint
Every portfolio sits somewhere between one position and the index. Where you sit should be a deliberate decision, not a residue of how many ideas you happened to find.
The argument for concentration
Your best idea is better than your tenth. If you genuinely have an edge, spreading capital thinly across many positions dilutes it towards the index — while you still charge fees as though you were adding something.
Concentration also improves the work. Twelve positions can each be understood properly. Sixty cannot, by anyone, and a portfolio of sixty half-understood positions is riskier than a portfolio of twelve well-understood ones, whatever the volatility numbers say.
The argument against
You are not as good as you think. Nobody is. Position-level hit rates for good managers cluster around 55–60%, which means four in ten of your best ideas will be wrong. Concentration turns being wrong into being ruined.
Concentration also creates behavioural risk. A 20% position that falls 40% takes 8% off the fund and a great deal more off your judgement. Decisions made under that pressure are worse, and the pressure is worst exactly when clear thinking matters most.
Where practitioners land
Most fundamental managers run 15–30 positions, with the top five at 25–40% of the book. That is concentrated enough for the work to matter and diversified enough that one error is survivable.
The sizing rule that follows from the previous module: size against downside, not upside. A position where you lose 15% if wrong can be twice the size of one where you lose 40%, at identical conviction.
Correlation is what actually decides it
A portfolio of twenty positions that all depend on a single macro factor is one position. This is the most common concealed concentration in real portfolios, because it is invisible in a list of holdings and obvious only in a drawdown.
So group holdings by what would have to go wrong for them to fail together — not by sector, by mechanism. Then size the group, and let position sizes fall out of that.
Work this through
Take a portfolio — real, simulated or constructed — of at least fifteen positions.
Group them by failure mechanism rather than sector: what single development would hurt several at once? Compute your true exposure to each mechanism.
If any mechanism carries more than 25% of the book, write down how you would fix it and what it would cost you to do so.
You are previewing this Blueprint
The remaining lessons, the workshops, your progress tracking and notes are included with Professional membership.