How Many ETFs Do You Actually Need?
How many ETFs do you need? Fewer than most people hold. Diversification comes from what your funds own, not how many you own — and five can hold the same companies.
How Many ETFs Do You Need? Why 5 Can Be Less Diversified Than 1 (Explained by a Banker)
Prefer to watch? This article is the written companion to the video above.
Most portfolios are not built. They accumulate.
An S&P 500 fund, because everyone starts there. A world fund a year later, because holding only America felt narrow. Then a Nasdaq 100, because technology had been doing well. Somewhere along the way a dividend fund and a fund with “quality” in its name. Five positions on the statement, five different providers, five sets of paperwork. It looks like a fortress.
Read the holdings and it is one room with five doors.
The short answer
Fewer than you have, and the number is the wrong thing to count.
Diversification is a property of what you own. It has nothing to do with how many product names appear on your statement, and a portfolio of five funds can be less spread out than a portfolio of one. What makes a second fund worth holding is that it does something the first one cannot do. That is the whole test, and the rest of this article is what it looks like in practice.
Nine of ten
Take the two funds people most often own together, precisely because they feel like different things. One tracks the largest American companies. The other calls itself a world fund and holds 1,282 companies across 23 developed markets.
The overlap problem
Two funds, one engine room
An S&P 500 fund is American large caps. A world fund is 23 developed markets and 1,282 companies. Here are the ten largest positions of each, joined where they are the same company.
Nine of the ten are the same company. Only Tesla and JPMorgan Chase differ, and both sit at the bottom of their list. What actually separates the two funds is not which companies you own but how much of them: the same ten names are 36.31% of one fund and 26.41% of the other.
Nine of the ten largest positions are the same company. Not similar companies. The same ones: NVIDIA, Apple, Microsoft, Amazon, both share classes of Alphabet, Broadcom, Micron and Meta. Only Tesla and JPMorgan Chase differ, and both sit at the bottom of their respective lists.
What separates the two funds is not which companies you own. It is how much of them. The same ten names are 36.31% of the American fund and 26.41% of the world fund, according to the SPDR S&P 500 fact sheet dated 30 June 2026 and the MSCI World factsheet dated 31 July 2026.
Nobody hid this. Both documents are free, both are one page long, and both print the table on the front.
Why the same names keep turning up
Because almost every large index weights companies by size. The bigger the company, the bigger its slice — in every index it appears in.
That single rule produces the effect. Seven American companies are around 31.53% of the S&P 500 fund on that fact sheet. Six of those same seven already account for 22.38% of the world fund. Add a Nasdaq fund, which is narrower still, and the concentration goes up rather than down.
You never decided to bet a third of your money on seven companies. You bought three funds with three different names, and the arithmetic decided for you.
None of this makes those funds bad. It makes the third one pointless.
The test that fixes it
Banks do not decide what to hold by counting positions. Every position has to justify a line on the sheet, and the justification has a specific form: this holding is here because it does a job nothing else here does.
The test
Does this fund have a job?
Not "is it a good fund" — almost all of them are fine. The question is whether it does something no fund you already own is doing. Send each holding down the track.
A global bond fund
passes
on different asset class
It does not fall when shares fall, which is the entire job.
An emerging markets fund
passes
on different region
Countries a developed-market index leaves out by definition.
A small-cap fund
passes
on different factor
Companies below the size cut of the index you already hold.
A second large-cap fund
fails
Same companies, different label. That job was already taken.
Applied to a private portfolio, there are three honest ways a fund can pass. It reaches a market the others do not. It holds an asset that behaves differently when shares fall. Or it covers a slice of the market the others exclude by size or style.
A bond fund passes, and it passes on the strongest ground of the three, because the point of it is that it does not fall on the day everything else does. An emerging markets fund can pass, since a developed-markets index leaves those countries out by definition rather than by accident. A small-cap fund can pass for the same reason.
A second large-cap fund with a different label does not pass. That job was taken.
Here is the version worth remembering. If you cannot name the job, the fund is not diversification. It is decoration.
How simple is too simple?
That is the fair objection, and it deserves a real answer rather than a reassurance.
The honest limit sits further out than most people expect. A single broad world fund already holds 1,282 companies. The question of how many you need was settled in academic work decades ago: Evans and Archer put the shape of it on the map in the Journal of Finance in 1968, and Meir Statman sharpened it in 1987, concluding that a portfolio of randomly chosen stocks needs at least 30 of them for a borrowing investor and 40 for a lending one.
The other side
Where adding funds stops paying
Spreading money across more companies lowers the risk that any single one of them ruins your year. It does that fastest at the start, and then it stops. This is the shape of the thing.
The gain is nearly all spent before you reach 50 names. Which is why a second fund holding the same kind of companies changes so little: you are buying more of a benefit you have already collected. A world fund at 1,282 companies is not sitting near the start of this curve. It is sitting at the far flat end of it.
Thirty to forty. Not four hundred, and certainly not four thousand.
Going from one company to fifty changes your risk profile beyond recognition. Going from five hundred to fifteen hundred is a rounding error. This is why the second large-cap fund feels so unsatisfying once you look at it directly: you are paying for more of a benefit you finished collecting a long time ago.
Two footnotes that keep this honest
The first is about what a share fund is not. A stock ETF is one hundred per cent stocks. It answers how your share money is spread. It says nothing at all about how much of your money belongs in shares in the first place, and that second question moves outcomes far more than the first one does. It is the subject of how much of your money should be in stocks, and it is not solved by owning a very good equity fund.
The second is about the word on the tin.
Read the label twice
What "world" pays for
A hundred squares, one for every hundredth of the money you put into a broad world fund. This is where it lands.
- Japan 5.73%
- United Kingdom 3.61%
- Canada 3.41%
- France 2.44%
- the other 18 markets 12.78%
This is not an argument against the fund. It is an argument against reading the word on the tin and stopping there. If you want less than seventy-two cents in the dollar riding on one country, that is a decision to make on purpose — and it is exactly the kind of job a second fund can be given.
That world fund was 72.03% United States on 31 July 2026. Japan came second at 5.73%. “World” is on the label; the engine room is narrower than the label suggests. This is not an argument against the fund, and it is not a scandal — it is what a market-cap-weighted index does when one country’s companies are worth the most. But if you would rather not have seventy-two cents in the dollar riding on a single country, that is a decision to take deliberately. And it happens to be a real job, which a second fund could be hired to do.
What most people get wrong
They treat the number of funds as a score.
Five feels more serious than two. It feels like effort, like research, like someone who is paying attention. The statement is longer, and a longer statement reads as a more considered one.
Then the market falls and all five fall together, because they were all holding the same seven companies. The portfolio behaves like a single bet, which is what it always was, and the only thing the extra funds contributed was four more sets of fees and a rebalancing job nobody enjoys.
There is a second failure mode worth naming, and it is quieter. Complexity discourages maintenance. Two positions get rebalanced. Eight get postponed. The cost of the extra funds is not really the fees, though fees are worth seeing over a lifetime — it is that the portfolio slowly stops being managed at all.
The audit: ten minutes, two questions
Open your holdings. For each fund, pull up the factsheet. They are free, they are one page, and every provider publishes one monthly.
Question one. Write down the top ten positions of each fund, side by side. If the same names keep appearing, you have your answer, and there is nothing further to analyse.
Question two. For every fund, finish this sentence out loud: this fund is here to do the following job that no other fund I own is doing. Region, asset class, or slice of the market. If you can finish the sentence, keep it. If you cannot, you have found the one to look at.
Then do nothing dramatic. Selling to tidy up can trigger tax on gains, and paying a certain cost to fix an uncertain inefficiency is usually a poor trade. The gentle version works: stop adding to the fund without a job, and send new money to whichever holding actually has one. The portfolio simplifies itself over a couple of years without a single realised gain.
Once you know how many, how to pick an ETF covers which one — domicile, UCITS structure, total expense ratio and the tracking questions that decide between two funds that look identical. If you are still deciding whether to hold index funds at all, index funds versus active funds is the prior argument, and the three-bucket structure is where the bond fund earns its job.
Drawing on nearly two decades in the financial industry, the thing I would most like to leave you with is how unglamorous the answer is. Nobody builds a reputation recommending two funds. It is still, for most people, the right number.
This article is educational content. It is not financial, tax or legal advice, and nothing in it is a personal recommendation. Rules and tax treatment differ by country and change over time. If a decision matters to you, it could be worth having it reviewed by a qualified professional before you act.
Primary sources
- 01MSCI World Index (USD) factsheet — reports 1,282 constituents across 23 developed markets, the United States at 72.03% of the index, the ten largest constituents at 26.41% of the index in total, and coverage of approximately 85% of the free float-adjusted market capitalisation in each country — MSCI Inc., factsheet dated 31 July 2026
- 02SPDR S&P 500 ETF Trust fact sheet — reports 504 holdings and the ten largest positions with their weights, led by NVIDIA at 7.50% and Apple at 6.57%, with the ten together at 36.31% of the fund — State Street Investment Management, fact sheet as of 30 June 2026
- 03How Many Stocks Make a Diversified Portfolio? — concludes that a well-diversified portfolio of randomly chosen stocks must include at least 30 stocks for a borrowing investor and 40 stocks for a lending investor — Meir Statman, Journal of Financial and Quantitative Analysis 22(3), September 1987, pp. 353–363
- 04Diversification and the Reduction of Dispersion: An Empirical Analysis — the study that established the shape of diminishing returns to adding holdings; cited here for the finding's origin, not for a figure — John L. Evans and Stephen H. Archer, The Journal of Finance 23, December 1968, pp. 761–767
Questions people actually ask
How many ETFs should I own?
For most long-term investors, one to three broad funds does the whole job. The number itself is not the point though, and chasing a number is how people end up with five funds that hold the same companies. Count jobs instead of tickers. One fund covering global shares is a job. A bond fund is a second job, because it does not fall at the same moment shares do. A fund covering markets your first one excludes by definition, such as emerging markets, can be a third. Beyond that you are usually buying more of something you already own, which costs you rebalancing work and gains you nothing.
What is ETF overlap and why does it matter?
Overlap is the share of your money that ends up in the same companies through different funds. It matters because it is invisible from the outside: the funds have different names, different providers and different fees, so the portfolio looks spread out while the money is not. Two of the most widely held index funds on the planet make the point. On the SPDR S&P 500 fact sheet dated 30 June 2026 and the MSCI World factsheet dated 31 July 2026, nine of the ten largest positions are the same company. Only Tesla and JPMorgan Chase differ, and both sit at the bottom of their list.
Can you over-diversify with ETFs?
You can over-complicate, which is the practical version of the same problem. Adding funds that hold what you already hold does not reduce risk in any measurable way, and it does add cost, paperwork and more places for a mistake to hide. There is also a subtler harm: with eight positions instead of two, rebalancing becomes a chore, so it stops happening. The honest framing is not that extra funds are dangerous. It is that they take real effort and give nothing back.
Is one world ETF enough?
For the equity part of a portfolio, often yes. A broad world fund held 1,282 companies across 23 developed markets on 31 July 2026, and the research on diminishing returns settled decades ago: the fall in risk is nearly all collected before you reach fifty names. Two honest footnotes though. A stock fund is one hundred per cent stocks, so it answers how your share money is spread, not how much of your money belongs in shares at all. And that same factsheet puts the United States at 72.03% of the index, so the word on the label is broader than the engine room.
Do more ETFs mean more diversification?
No, and this is the single most common misreading in personal finance. Diversification is a property of what you own, not of how many product names appear on your statement. Five funds that all hold large American companies leave you exposed to exactly one thing happening. One fund holding 1,282 companies across 23 countries does not. If you want to know how diversified you are, ignore the number of positions entirely and look at what sits inside them.
How do I check whether my funds overlap?
Two questions and about ten minutes. First, open the factsheet of each fund you hold and write down the top ten positions. If the same names keep appearing, you have your answer, and no further analysis is needed. Second, for every fund, finish this sentence: this fund is here to do the following job that no other fund I own is doing. If you cannot finish the sentence for a fund, that is the one to look at. Both factsheets are free, a page long, and published monthly.
Should I sell the funds that overlap?
Not automatically, and not this week. Selling can trigger tax on gains, and a tax bill is a certain cost paid to fix an uncertain inefficiency, which is rarely a good trade. The gentler route is to stop adding to the redundant fund and direct new money to whatever actually has a job. Overlap is untidy rather than urgent. If a position is small and the tax cost is nil or trivial, tidying it up is reasonable, but the order matters: work out the target first, then decide what it costs to get there.
Is a two-fund portfolio too simple?
Simple and simplistic are not the same word. A global share fund plus a bond fund covers the two questions that actually move outcomes, which are how much you own of the growth engine and how much sits somewhere that does not fall with it. What such a portfolio lacks is not diversification but decoration. The genuine limits are worth naming: it holds no emerging markets if the share fund tracks developed markets only, and it leans heavily towards the United States. Both are fixable with one more fund each, if you decide those jobs are worth doing.