The textbook portfolio
Textbook diversification assumes the portfolio is built from cash. The optimization is unconstrained. The output is a clean efficient frontier, and the recommended allocation is a function of risk tolerance.
Real portfolios are almost never built from cash. They are inherited from a prior adviser, accumulated through employer equity grants, structured around a concentrated founder position, or shaped by tax basis that makes liquidation expensive. The constraints are not edge cases. They are the substance of the engagement.
Diversification under constraint
The honest framing is that diversification is a process, not a state. A portfolio that holds a single position worth a meaningful fraction of total net worth is not diversified, even if the rest of the assets are spread across a hundred securities. The remediation is multi-year, takes account of tax consequences, and depends on the client's view of the underlying position.
There are well-understood techniques for managing concentration risk without triggering immediate liquidation: exchange funds, completion portfolios, hedging overlays, charitable structures, and disciplined trim schedules. Each has trade-offs in cost, complexity, and time horizon. None of them are appropriate by default. All of them require the client's authorization, because the choice between them reflects the client's preferences, not the adviser's.
What diversification is for
Diversification is not a goal in itself. It is a means of reducing the dispersion of outcomes. A perfectly diversified portfolio that is not aligned with the client's actual liabilities, time horizon, and obligations is not well-constructed. A less diversified portfolio that is aligned with those things may be exactly right.
The work, then, is not to optimize a portfolio in isolation, but to design a portfolio that fits the client's life. That requires more conversation than calculation. It is the part of advisory work that does not delegate well to a screen.
