The Herfindahl-Hirschman Index (HHI) compresses "how concentrated is my portfolio?" into a single number between 0 and 1. Low means your money is spread across many positions. High means one or two holdings dominate — and your outcome depends on them far more than you probably realise.
Counting positions is a terrible measure of diversification. A portfolio holding ten stocks sounds diversified. But if 85% of the money sits in one of them and the other nine split the remaining 15%, you do not have a ten-stock portfolio. You have a one-stock portfolio with nine decorations, and it will behave like one.
HHI is the standard fix. Instead of counting holdings, it weighs them — and it weighs them in a way that deliberately punishes concentration.
Take each position's share of the portfolio as a decimal (a position that is 40% of your money is 0.40). Square each one. Add them up. That sum is the HHI.
Squaring is the whole trick. It makes big positions count for disproportionately more than small ones: a 50% position contributes 0.25 to the total, while ten 5% positions contribute 0.0025 each, or 0.025 in total — one tenth as much. The measure is engineered to make one dominant holding impossible to hide behind a long tail of small ones.
The scale runs from near 0 to 1. An HHI of 1.0 means everything is in a single stock. The lowest achievable value for a portfolio of N equal-weighted positions is 1/N — so ten equal positions give an HHI of 0.10, and twenty give 0.05.
This is not a metric we invented. HHI has been used for decades by competition regulators (including the US Department of Justice) to measure how concentrated an industry is among a few firms. The maths is identical; only the thing being weighed changes.
One divided by HHI gives you the "effective number of names" — how many equally-weighted positions your portfolio actually behaves like. This is the number we show as a hint, and it is far more intuitive than the raw HHI.
Example (illustrative). A portfolio with one 80% position and four 5% positions has HHI = 0.80² + 4 × 0.05² = 0.64 + 0.01 = 0.65. Its effective number of names is 1 / 0.65 ≈ 1.5. You believe you hold five stocks. Your portfolio behaves as though you hold about one and a half. That gap between what you think you own and what you effectively own is precisely what this metric exists to expose.
By contrast, ten equal 10% positions give HHI = 10 × 0.10² = 0.10, and an effective count of exactly 10 — what you see is what you own.
We show the HHI to three decimals and colour it amber above 0.40, alongside the effective-names hint. Above roughly 0.40 you are, in effect, running a portfolio of two-and-a-half names or fewer — a single holding is dominating your outcome, whatever the position count says.
That threshold is a prompt to look, not a verdict. Deliberate concentration is a legitimate strategy — some very good investors concentrate on purpose, because diversification dilutes your best ideas along with your worst. The failure mode is not concentration. It is ACCIDENTAL concentration: drifting into a single dominant bet because one winner grew and you never rebalanced, and then being surprised when that one holding decides your year.
If the number is higher than you intended, the levers are straightforward: trim the dominant position, add to the smaller ones, or size new entries deliberately — which is what position sizing is for.
The critical blind spot: HHI sees position SIZES and is completely blind to what the positions actually are. Ten equal positions in ten megacap technology stocks produce a beautiful HHI of 0.10 while being, economically, one enormous bet on one sector. When that sector falls, all ten fall together. The metric that catches this is correlation, not HHI, and you genuinely need both.
The corollary is that a good HHI can be actively reassuring in the worst possible way — it can certify a portfolio as "diversified" on the exact day that the single shared risk running through all of it is about to show up. Read HHI as "is my money spread out?", never as "am I safe?"
Second, HHI ignores cash and it ignores volatility. A portfolio of ten equal positions where one is a quiet utility and another swings 8% a day is not really equally weighted in terms of RISK, even though HHI sees ten identical slices. Equal dollars are not equal exposure.
The Herfindahl-Hirschman Index: the sum of the squares of each position's share of the portfolio. It runs from near 0 (money spread widely) to 1 (everything in a single stock).
There is no universal answer, but as a reference, N equal positions give an HHI of 1/N — so ten equal holdings give 0.10. We flag values above 0.40 in amber, because at that point a single holding is effectively driving your outcome.
It is 1 divided by the HHI — the number of equally-weighted positions your portfolio actually behaves like. A portfolio of five stocks with one huge position can easily have an effective count of about 1.5.
Not necessarily, and this is the most important caveat. HHI only measures position sizes. Ten equal positions in ten stocks from the same sector give an excellent HHI while remaining a single concentrated bet, because they all fall together. Check correlation as well as concentration.
Browse all S&P 500 tickers to see this metric applied to individual companies.
Educational research only — not investment advice.