How Much Diversification Is Enough

The shape of the answer

Kabir accepts the case for diversification and wants to know when to stop. The answer has a distinctive shape, and knowing the shape is more useful than any number.

The benefit is front-loaded and then flattens. Going from one holding to a few removes a great deal of the risk that is specific to individual holdings. Going from a few to many removes some more. Going from many to very many removes almost nothing, because what remains is the risk they all share, and no amount of adding removes that — how diversification works.

So the curve rises steeply, bends, and then runs nearly flat. Everything practical follows from where the bend is.

Why we quote no number

There are widely circulated figures for how many holdings are enough. We are not going to repeat one, and the reason is not caution for its own sake.

Any such number depends entirely on how similar the holdings are, and that is not a constant. A few holdings in genuinely different businesses achieve more than many in one industry. A number that was measured in one market over one period does not transfer to another, and the ones in circulation are generally quoted without either.

We hold no constituent or correlation data for Indian shares, so we cannot compute the point at which the curve bends here, and we will not import somebody else's answer. What we can give is the principle, which is more robust anyway: count sources of risk, not holdings.

Counting sources of risk instead

The useful question is how many genuinely different things your outcome depends on. Run down this list and count honestly.

Asset class. Shares, bonds, cash, property. This is the largest division available and the one that does most of the work.

Geography. Whether all of it depends on one economy.

Currency. Related to geography and not the same thing — a holding can be in foreign businesses and carry currency exposure that behaves separately.

Company size and industry, within shares.

And whether your own income depends on the same things. This is the one people miss entirely. If you work in an industry and hold shares in that industry, and your city's property market depends on it too, you have concentrated far more than any portfolio analysis will show — because the biggest holding on your balance sheet is your future earnings, and it is not in the portfolio.

Most people who feel diversified because they hold many funds turn out, on this count, to have two or three genuine sources of risk.

What the extra holdings cost

The benefit flattens; the costs do not.

Charges accumulate. Every holding carries its own annual charge, certain and compounding, whether or not it adds any diversification — how investment fees reduce wealth.

Rebalancing gets harder. More holdings means more decisions and more transactions, each potentially a tax event, and rebalancing is the mechanism that actually controls risk over time — how rebalancing controls risk.

And you stop being able to see it. This is the cost that matters most and appears in no calculation. A portfolio you cannot hold in your head is one whose concentration you will not notice accumulating, whose drift you will not correct, and which you will eventually stop reviewing properly. Complexity beyond what you will actually maintain is itself a risk, and it is the one that most reliably damages real portfolios.

A workable stopping rule

Stop when you can no longer say what a new holding adds that you do not already have.

That test does the work of any number, and it fails safe: if you cannot answer it, the addition is buying a charge. It also correctly permits a small number of holdings when they are genuinely different, and correctly refuses a large number when they are not.

Two supporting checks. Look at the combined portfolio, not the list of funds — the practical method is measuring concentration and overlap. And get the asset allocation right first, because it dominates everything decided inside the equity portion: choosing an asset allocation.

The other direction: when you are not diversified enough

The failure is not symmetric, and undershooting is worse than overshooting.

The clearest case is a single large holding you did not deliberately choose — inherited, or accumulated through an employer. Maya's position is exactly that, and it carries a risk that no amount of care elsewhere in the portfolio offsets: a company-specific disaster is not something diversification within the rest of the portfolio can absorb when that one holding dominates.

If a single holding is large enough that its failure would change your plans, that is the problem to solve, and it is a different and more urgent problem than whether you own four funds or seven.

What to take away

The benefit of diversification arrives early and flattens quickly, so the practical question is not how many holdings but how many genuinely different sources of risk — and most portfolios have far fewer than they appear to.

Count those, include your own earnings in the count, and stop adding when you cannot say what the next holding contributes. A portfolio simple enough to maintain and honestly diversified across a few real dimensions beats a long list every time, because the long list is the one that quietly stops being looked at.