Portfolio concentration: how much in one stock is too much?

How much of a portfolio in one stock is too much? There is no single right number, but practitioners commonly cite a range of roughly 5–10% for any one company, with wider single-name weights treated as a deliberate exception rather than a default. The figure varies because it depends on what else you own, how correlated those positions are, and how large the portfolio is relative to your total wealth. This article explains what diversification can and cannot do, why a plain-looking portfolio can be far more concentrated than its statement implies, and how to measure your own concentration with a weight table and a single summary figure — the effective number of holdings.

What diversification actually removes

Concentration matters because of a distinction at the foundation of modern portfolio theory: a position's total risk splits into two parts. Idiosyncratic risk (also called specific or diversifiable risk) is the part unique to one company — an accounting fraud, a failed drug trial, a lost anchor customer. Market risk (systematic or non-diversifiable risk) is the part every stock shares — a recession, a rate shock, a broad repricing. Since Harry Markowitz's 1952 paper “Portfolio Selection” (Journal of Finance), the core insight has been that spreading capital across imperfectly correlated names cancels much of the idiosyncratic part while leaving the market part intact.

That single sentence sets the ceiling on what any weighting rule can achieve. Diversification removes the risk that the fate of one company decides your result; it does not remove the risk that the whole market falls. Limiting concentration is therefore not about escaping losses — a broadly spread portfolio still drops when the market drops — it is about making sure that no single business failure is decisive. Nothing you do with weights moves that boundary, and nothing needs to.

The ranges practitioners cite — and why they vary

In practice a band has settled around single-name weights of roughly 5–10% of a portfolio, with sector limits often cited around 25–30%. Some written rule sets are more granular — 5% as a standard position, 10% for a high-conviction name, 15% as an absolute ceiling that requires a documented exception. These are descriptive observations of common practice, not a specific study statistic, not a rule that fits every portfolio, and certainly not a limit prescribed for any individual reader.

Why the range rather than a single figure? Because the right number for one person is not the right number for another. A brokerage account sitting beside a paid-off house and a public pension carries a concentrated position very differently from an account that is someone's entire retirement. What the rough bands offer is less a magic percentage than a mechanism: anyone with a written line at all notices concentration before it becomes the portfolio's quiet main bet.

The worked example marked on the ladder — a single name at 22% of the portfolio — sits well past the commonly cited band. That does not make it a mistake. It makes it a position large enough that the rest of this article is worth doing: measuring it precisely, understanding what it does to the whole, and deciding on purpose whether to keep it that size.

The employer-stock special case

One concentrated position deserves separate mention because it recurs so often: shares in your own employer, accumulated through options, restricted units or a share-purchase plan. It is a frequently discussed case precisely because the concentration is doubled. Your salary, your bonus and often your next job already depend on that one company; adding a large slice of your investment portfolio in the same name stacks financial capital on top of human capital that is already fully exposed to the firm.

If the business stumbles, the paycheque and the portfolio can move together — the opposite of what diversification is for. The arithmetic is identical to any single-name concentration; what makes the employer case distinctive is that the correlated risk sitting outside the portfolio is easy to overlook when the shares arrived as compensation rather than as a deliberate purchase. Naming that hidden correlation is the whole point of treating it as a special case.

Hidden concentration: the same mega-caps, several times over

The most common way concentration hides is through funds that appear to diversify but quietly converge on the same handful of names. The instrument sold as the cure — the broad index fund — is itself concentrated. According to the MSCI World Index factsheet (as of 30 June 2026), the index spans 1,283 companies across 23 developed markets, yet the United States alone accounts for 72.45% of its market value, the ten largest positions weigh 25.74% together, and the information-technology sector reaches 30.27%. Two names lead the list: NVIDIA at 5.18% and Apple at 4.77%.

Overlap makes it worse. A familiar pattern: a world-index fund as the base, a US or technology fund added “for growth”, and a direct position or two in the same mega-caps on top. On a statement that is three or four independent-looking lines; economically they contain the same companies. To find your true weight in one name you have to look through the funds — each fund's share of your portfolio times the name's weight inside that fund, summed across every fund, plus any direct position.

Suppose, for illustration, that half a portfolio sits in a broad world-index fund. Apple is 4.77% of the MSCI World (the sourced figure cited above), so that one fund alone places 0.50 × 4.77% ≈ 2.4% of the whole portfolio into Apple — before counting a US-large-cap fund that contains the same name at a higher weight, and before any direct position. Add a hypothetical 3% direct Apple position and the true figure is at least 5.4%, close to double what the single “Apple 3%” line on the statement implies. Only the 4.77% index weight is sourced; the 50% and 3% allocations are illustrative round numbers used to show the arithmetic.

The diagram makes the point visually: two funds that look independent on a statement share their largest names. The overlap is not a rounding error — for broad-market and large-cap funds it is precisely the mega-caps that dominate both, which is why the honest question is never “how many funds do I own” but “how many distinct economic bets do they add up to.”

Measuring your own: a weight table and the effective number of holdings

An honest concentration check needs no bank connection and no special software — just your own position data and a little arithmetic. Export your holdings from your broker as a CSV (almost every provider now has an export or reporting area), compute each position's share of the total, and resolve fund positions into their largest components using the public index factsheets. That weight table — every position, its market value, its share of the total, sorted high to low, with funds looked through — answers the three questions that matter most: how heavy is the largest single name, how heavy the largest sector, and how large the true regional tilt once the funds are unpacked.

To compress the whole distribution into one figure, take the Herfindahl-Hirschman Index (HHI): the sum of the squared position weights, HHI = Σ w_i². Squaring is the whole idea — large weights count far more than small ones, small ones almost not at all. Its reciprocal, 1 ÷ HHI, is the more intuitive number: the effective number of holdings, the count of equally sized positions that would carry the same concentration.

For contrast, an equal-weight portfolio of eight positions would have HHI = 8 × 0.125² = 0.125 and an effective number of holdings of exactly 8. The illustrative portfolio in the table nominally also has eight positions, but because the largest is 22%, its HHI works out to 0.1386 and its effective number of holdings to about 7.2. The list says eight; the concentration behaves like about seven. As the top position grows, that effective number falls — long before anything on the statement looks different from the month before.

Thresholds as personal policy, monitoring as fact

Detecting and measuring are snapshots; concentration is a process. A portfolio that is evenly spread today can carry a 22% position again in eighteen months on price movement alone — no order required. That is why a written policy tends to work better than a one-off cleanup: self-chosen ceilings per position, sector and region, a fixed review cadence, and for any deliberate exception a dated, written reason. Concentration then stops being an accident in slow motion and becomes a decision with a signature.

Software can do exactly one thing here, and do it reliably: make the facts visible. Acutic's risk dashboard reads a portfolio imported by CSV and reports the weight of each position, sector and region — looking through funds, from a file, with no account linking. A rule monitor can state that a position stands at 12.3% against a self-set 10% line, as an observation, not an instruction. What follows from that fact is the reader's decision, not the tool's.

When concentration is a deliberate decision

None of this makes concentration a fault in itself. Concentrated portfolios have built fortunes and destroyed them; that wide spread of outcomes is exactly what concentration means, mathematically. A large position that was entered on purpose, sized deliberately, written down and reviewed on a schedule is a risk decision. A large position that surprises its owner the first time the portfolio is honestly measured is not. The difference is not the percentage — it is whether the number was chosen.

That is where a decision record earns its place. Writing down why a position is the size it is — the reasoning, the date, the conditions under which the view would change — turns a drifting weight into a documented choice you can revisit. The companion piece on the investment decision journal covers how to keep such a record; the German sister article Klumpenrisiko im Depot works the same measurement in a DACH-specific context; and if your question is how the whole portfolio stacks up against a broad benchmark rather than any single name, how to benchmark your portfolio takes it from there.

Further reading: the Acutic methodology page documents the full seven-factor construction, and the product page shows the CSV import and risk dashboard in detail. Create free account to run the numbers on your own portfolio.

Acutic provides investment research and educational analysis under MAR Art. 20 / § 85 WpHG. Acutic does not provide investment advice (Anlageberatung per § 1 Abs. 1a S. 2 Nr. 1a KWG / Art. 4(1)(4) MiFID II), portfolio management, or any other licensed investment service. No content in this article constitutes a personal recommendation.