Effective diversification: why 11 holdings can behave like 3

Eleven holdings, six asset classes — behaving like just 3.1 independent assets. Line counts lie; effective assets tell the truth.

Open almost any retail portfolio: a dozen holdings, several asset classes, a few geographies. On paper, it is diversified. Measured, it is often a different story: weights concentrated in two or three positions, and holdings that rise and fall as one block. The line count is the most intuitive diversification measure there is — and the most misleading.

There is a measure that does not get fooled: the number of effective assets. It answers the only question that matters: how many genuinely independent bets does your portfolio actually contain? Sometimes the answer is flattering. Often it is a much smaller number than the one on your broker's screen — and knowing it changes how you read every other risk measure.

The line-count trap

Counting holdings is counting labels. Diversification does not care about labels: it depends on two things the count ignores entirely. Weights, first — if one position is half the portfolio, the other nine are extras on the set. Correlations, second — three tech stocks, a Nasdaq ETF and two cryptos add up to six lines and, quite often, a single bet: global risk appetite. When it turns, everything falls together.

Adding a holding can even make a portfolio more concentrated rather than less, if the new line moves like the old ones. More lines has never meant less risk.

The mental test is simple: think back to your last big red day. How many of your holdings finished down that day? If the answer is “almost all of them”, your portfolio has already shown you its true diversification — all that was left was to put a number on it.

Effective assets: counting bets, not labels

The intuition fits in one example. A portfolio where one holding weighs 90% and nine holdings share the remaining 10% behaves, in practice, like a portfolio of barely more than one asset. At the other extreme, ten equally weighted holdings that are independent of each other genuinely count as ten. The number of effective assets sits between those poles: it penalises concentrated weights, then shrinks further when holdings are correlated — two positions that always move together count as one.

It reads at a glance: “your 11-holding portfolio behaves like X assets”. When X is far below the line count, your diversification is cosmetic. It is also a number you can compare with yourself over time: the same portfolio, measured every week, tells you whether your trades and the market's moves are widening or narrowing your true spread of bets.

A real example: 11 holdings behaving like 3.1

OplynQ's public sample portfolio — the one anyone can open without an account — holds 11 positions worth €84,615, split across stocks 32%, ETFs 28%, crypto 13%, cash 12%, bonds 8% and commodities 7%. Six asset classes, eleven holdings: every appearance of a prudent portfolio. Measured with the method above — each holding's weight, refined by the correlations between holdings — it behaves like 3.1 effective assets.

Why so few? Two causes stack up. Weights, first: a handful of positions concentrate a large share of the value, and the small ones — bonds, commodities — weigh too little to truly count. Correlations, second: the tech stocks and the crypto in this portfolio largely move together, forming a single directional block. Cash and bonds do diversify, but their weight caps the effect. The result: eleven labels, three bets.

And the number moves without you touching anything. When one position climbs faster than the rest, its weight swells, concentration rises and the effective-asset count erodes — without a single order placed. When correlations tighten, same effect. A portfolio that counted as five bets can count as three a few months later, with exactly the same holdings. It is a measure to revisit regularly, not a box to tick once and forget.

The hidden duplicates

The line count also misses duplicates. Holding an Apple share and a World ETF means holding Apple twice: the World ETF carries the big US large caps at the top of its basket, and your direct share stacks on top. The same logic applies to a sector ETF held alongside its own top constituents, or to two “world” ETFs from different issuers — two lines, nearly the same basket. Without looking inside the ETFs, these duplicates stay invisible, and your true exposure to a single stock or sector can far exceed what the list of holdings suggests. The mirror image exists too: dropping a directly held share does not remove the corresponding exposure as long as an ETF in the portfolio still carries it — the list of holdings changes, the measured exposure much less.

Diversifying for real — in measurement terms

Let us be explicit up front: nothing here is a recommendation to buy or sell anything. It is a reading grid — diversification is something you measure, not something you declare:

• The test is not the label, it is the measured correlation. Two asset classes only diversify each other if they do not move together — and that is verified in a correlation matrix, not in a brochure.
• Weights matter as much as lines. A position that has grown to dominate the portfolio crushes diversification, even surrounded by ten exemplary small holdings.
• Geographies only diversify if they genuinely behave differently. Again: something to check in the correlations, not to assume.
• Look inside the ETFs. Real exposure per stock, sector and region is computed with ETFs unfolded — it is the only way to see the duplicates.

And after every trade, one useful question: did my number of effective assets go up, or did I just add a label? This measure belongs inside a full tracking routine — we laid it out in our portfolio tracking method.

What OplynQ does

OplynQ displays your consolidated portfolio's effective-asset count, the correlation matrix between your holdings (computed on 5-day rolling windows), and ETF transparency: your real exposure per sector and region with ETFs unfolded — which is where the duplicates show up. An AI copilot can then explain the number: which holdings form a block, and what genuinely diversifies. Diversification is only one facet of risk — volatility, VaR and drawdown are the others, explained in our portfolio risk analysis article.