Portfolio concentration risk: what it is and how to measure it
Concentration risk is the share of a portfolio's outcome that rides on a small number of positions, sectors or shared drivers. A book of twenty stocks can carry it; a book of five can avoid most of it. What matters is how the weight is distributed and what the holdings have in common, not the count.
What concentration risk means
Diversification works because unrelated holdings rarely all go wrong at once. Concentration is the opposite condition: enough of the book depends on one thing that a single bad outcome — one company's earnings miss, one sector's rerating, one change in interest rates — moves the whole portfolio.
It comes in three forms, and a portfolio can pass one test while failing another:
One holding is a large share of the total. Its own news becomes the portfolio's news.
Several holdings sit in the same industry or story, so they tend to be repriced together even when each weight looks modest.
Holdings in different sectors that answer to the same driver — long-duration growth, the dollar, oil — and move as one when that driver moves.
Most concentration is never chosen
Few investors decide to put a quarter of their money in one company. It usually arrives by drift: the holdings that do well grow, the rest do not, and nobody trades.
An illustrative book starts with ten equal positions of 10% each. One of them triples and the other nine stay flat. That one holding is now 25% of the portfolio — two and a half times its starting weight — without a single purchase.
Drift is why concentration is worth measuring on a schedule rather than only at the moment of buying. The weight on the day of purchase says little about the weight a year later.
Four ways to put a number on it
Each measure answers a slightly different question. None needs more than the current weights of what you hold.
| Measure | How it is calculated | What it tells you |
|---|---|---|
| Largest position | The biggest single weight | How much one company's result can move the whole book |
| Top-five share | The five largest weights, added | Whether a handful of names carry most of the portfolio |
| Herfindahl index (HHI) | Every weight squared, then summed | One number for the whole distribution; larger means more concentrated |
| Effective number of holdings | 1 divided by the HHI | How many equal-sized positions the book behaves like |
The last one is the easiest to read. Squaring each weight makes large positions count for much more than small ones, so the effective number falls quickly as one position grows.
Each weight is 10%. The HHI is ten times 0.1², or 0.10. The effective number of holdings is 1 ÷ 0.10 = 10.
The large position contributes 0.4², or 0.16; the other nine add about 0.04 together. The HHI is 0.20, and the effective number is 5 — ten names that behave like five.
Both books are illustrative. The arithmetic is the same for any portfolio.
Correlation: concentration you cannot see in the weights
All four measures above treat every holding as independent. Real holdings are not. Three semiconductor makers and two cloud-software companies look like five positions spread across two industries, yet much of the time they are repriced by the same thing — expectations for technology spending and the level of interest rates.
Correlation measures how closely two holdings move together, from −1 (opposite) through 0 (unrelated) to +1 (in lockstep). A portfolio whose holdings mostly correlate strongly with each other is more concentrated than its weights suggest, and the effective number of holdings overstates its real diversification. Correlations are not fixed: they tend to rise in sharp sell-offs, which is when diversification is most wanted.
How regulated funds define "diversified"
There is no single correct level of concentration for a private investor; it depends on goals, time horizon, taxes and conviction. Regulated funds do have to meet written limits, and they give a sense of scale:
- US diversified funds (Investment Company Act of 1940): for at least 75% of the fund's assets, no more than 5% may sit in any one issuer, and the fund may not hold more than 10% of any issuer's voting shares.
- EU UCITS funds: no more than 10% in any one issuer, and the positions above 5% may not add up to more than 40% of the fund.
These are rules written for pooled funds sold to the public. They describe how regulators draw the line for those products, not a target that fits an individual portfolio.
What investors weigh when a position has grown large
A concentrated position is not automatically a problem. Some investors hold one deliberately, because their conviction in it is high and they have decided the risk is worth carrying. The measures above say how much depends on a few outcomes; they do not say whether that is acceptable to the person who holds the book.
When investors do decide to reduce concentration, the approaches commonly considered include selling part of the position, directing new money to other holdings so the large one shrinks as a share over time, and writing down in advance the price or event at which they would revisit it. Each has trade-offs — selling can realise a taxable gain, and redirecting new money works slowly. Which, if any, fits depends on circumstances only the investor knows.
Where concentration shows up in the product
TensionLine's Portfolio Fit lens reads concentration, correlation to what you already own, and how far each position has drifted, alongside four lenses that read the company and its environment. A holding that has grown large can read Weakening on concentration alone, with the reasoning shown, even while the business itself is doing well. The reading is analysis for you to interpret; the decision, and any trade, stay with you.
See it read your own book.
Public beta — FREE, opens on 15 October. No card. An analytical tool — not investment advice.