Portfolio Diversification: Spot Hidden Concentration Risks and Spread Risk Properly

“Don’t put all your eggs in one basket” – everyone knows the saying. Yet plenty of portfolios hold ten stocks that will all fall together in the next downturn. Diversification is more than “owning many positions”: the point is to spread risks that are actually different from one another. This guide shows what diversification really does for you (and what it does not), where concentration risks hide, how many positions make sense – and why too much diversification is a mistake as well.

What diversification really delivers – and what it doesn’t

The risk of a stock has two components. Unsystematic risk is company-specific: an accounting scandal, a failed product, a lost major customer. This risk can be diversified away – the scandal at company A does not touch company B. Systematic risk (market risk), on the other hand, hits everyone: recession, an interest-rate shock, war. It does not disappear, no matter how many stocks you hold.

For a sense of scale: a typical individual stock swings by 30–50% a year, a broad equity market by roughly 15–20%. As few as 20 to 30 genuinely different stocks push a portfolio’s volatility close to the market level – what remains is market risk, and it stays. Diversification therefore protects you from being ruined by a single company. It does not protect you from a 30% drawdown in a crash. Investors who confuse the two sell in horror during their first bear market – one of the classic investing mistakes.

Also important: in expected-value terms, diversification costs you no return; it reduces dispersion. It is the only “free lunch” in the market – but only if the positions are genuinely exposed to different risks. Which brings us to the core problem.

Hidden concentration risks: 10 tech stocks are not a diversified portfolio

The number of positions tells you little. What matters is whether the positions hang on the same drivers.

Sector and factor concentration

Apple, Microsoft, Nvidia, Amazon, Meta, Alphabet, Tesla, AMD, Palantir, Shopify – that is ten names, but at its core a single bet: US technology, high valuations, interest-rate sensitivity, AI expectations. In 2022 this “diversified” portfolio lost 40–60% while the broad market gave up around 20%. Concentration also builds up through factors: ten cyclical value names (autos, banks, chemicals) all hang on the economic cycle together, even if the industries sound different. So audit your portfolio by drivers, not by names: what would have to happen for more than half of your positions to lose 30% at the same time?

ETF investors are not automatically safe either: the MSCI World is roughly 70% US, and its ten largest holdings – mostly tech – make up over 20% of the index. If you add a Nasdaq ETF and three individual US tech stocks on top, you have bought the same concentration three times over.

Currency concentration

As a euro-based investor with 80% US stocks, you carry a dollar risk on top of the equity risk: if the dollar falls 10% against the euro, you lose 10% of return without a single stock going down – in 2025 exactly that caught many portfolios off guard. For equities, currency risk is secondary over the long run, but it should be a conscious decision, not a byproduct of chance.

The most dangerous concentration: your employer’s stock

If you hold shares in your own employer, you double your risk: should the company get into trouble, your portfolio and your paycheck are threatened at the same time. The Enron collapse in 2001 is the textbook example – employees lost their jobs and their retirement savings on the same day. Rule of thumb: consistently reduce employer stock above 5–10% of your net worth as soon as holding periods and tax rules allow. The fact that you “know the company from the inside” is not a counterargument – the Enron employees thought so too.

How many positions – and how large?

For private investors holding individual stocks, a few simple guardrails have proven their worth:

15 to 25 stocks from at least 5–6 sectors and several regions capture most of the diversification benefit. Beyond about 30 names, each additional stock barely reduces risk any further but keeps demanding attention.
Starting weight per position: 3–7%. At a 5% weight, a total loss in one position costs you 5% of the portfolio – painful, but survivable.
Upper limit per position: about 10%, even if it grew there through price gains. Beyond that point, a single thesis determines your outcome.
Avoid mini-positions below 2%: they move nothing, but they clutter the portfolio.

If you have less time, a core-satellite approach works well: a broad world ETF as the core (say 70–80%), plus a few deliberately chosen individual positions – for instance quality stocks with a thoroughly tested thesis. Position size may well depend on your level of conviction: a stock with a limited bear case can carry 6%, a speculative thesis more like 2–3%. Our guide to bull and bear scenarios shows how to think a bear case through properly.

Correlation, explained in plain terms

Correlation measures, on a scale from −1 to +1, how much two investments move in step. At +1 they move in lockstep, at 0 independently of each other, at −1 exactly in opposite directions. The diversification benefit only kicks in at correlations well below +1: two assets, each with 30% volatility and a correlation of +0.9, combine to almost 30% again – nothing gained. At a correlation of +0.3 the pair’s volatility drops noticeably toward 24%, at 0 to about 21%.

For context: two US tech stocks often sit at +0.7 to +0.9. Stocks from different sectors and regions are more like +0.4 to +0.6. High-grade government bonds versus stocks have historically been around 0, at times negative – which is why a bond allocation dampens a portfolio so effectively. Gold versus stocks is close to 0. The uncomfortable truth: in phases of panic, correlations between stocks rise toward +1 – in a crash, almost everything falls together, as in 2008 and in March 2020. Genuine diversification across asset classes (stocks, bonds, gold, cash) therefore carries you further than diversification within the stock market alone.

Overdiversification: when spreading turns into “diworsification”

Fund manager Peter Lynch coined the mocking term “diworsification”: diversification that no longer improves anything and merely dilutes. Typical symptoms:

50, 80, 100 individual positions – nobody can seriously follow that many theses. The result is an expensive, poorly maintained near-index; at that point an ETF is the more honest and cheaper solution.
Five ETFs with the same contents: MSCI World, S&P 500, Nasdaq 100 and a “Global Tech” theme fund overlap massively – perceived diversification, real concentration.
Positions without a thesis: if you no longer remember why a stock is in your portfolio, you are not holding an investment, you are holding a souvenir.

The art lies in the middle ground: concentrated enough that good decisions pay off – diversified enough that a single mistake cannot knock you out of the game. Every position should have a nameable reason and a weight you can actually feel.

Rebalancing: maintaining your diversification

A portfolio drifts. Example: you start with 60% stocks and 40% bonds. After five strong years for equities you are at 75/25 – your risk has grown substantially without you doing anything. Rebalancing means restoring the target weights: sell what has risen, add to what has lagged. That forces countercyclical behavior and keeps risk where you wanted it in the first place.

Two rules are practical, and they combine well: review once a year on a fixed date, and additionally act whenever a position misses its target weight by more than a quarter (a 5% target has moved above 6.25% or below 3.75%). Trading more often adds little benefit but costs fees and – outside tax-sheltered accounts – taxes on gains with every sale. A more tax-efficient route is often fresh money: direct new contributions specifically into the underweighted positions instead of selling winners.

Bottom line: diversify risks, not line items

Good diversification does not count portfolio rows; it counts independent sources of risk: different sectors, regions, factors, currencies, asset classes – and never your livelihood and your portfolio tied to the same company. 15–25 well-reasoned positions at 3–7% weight, an annual rebalancing and an honest look at hidden concentrations beat any randomly filled 60-position portfolio. Analyzing each individual stock – business model, valuation, risks, for example with a structured report like the one from aktienanalyse.ai – remains mandatory: diversification does not turn bad individual decisions into a good portfolio, it only ensures that no single decision ruins you. You will find more fundamentals in our knowledge hub.