How do I choose an ETF: what should I look at first?

Cost (TER) and size (AUM)

The TER (Total Expense Ratio) is the annual management cost, deducted automatically from the ETF's value: for an MSCI World ETF it typically ranges from 0.12% to 0.20% a year. Over 20 to 30 years, even small TER differences add up significantly. AUM (assets under management) instead indicates the size of the fund: larger ETFs (ideally above £500 million) offer better trading liquidity and a lower risk that the provider decides to close the fund.

Tracking difference: the real cost beyond the TER

The TER isn't the only cost that affects actual returns. Tracking difference measures the real gap between the ETF's return and that of the index it tracks, over time — and it can differ from the stated TER, in either direction. Some ETFs, particularly physically replicated ones that temporarily lend out the securities they hold to other market participants (securities lending), generate extra income that can partly offset the TER, pushing the real tracking difference below the stated cost. Other ETFs, due to management inefficiencies or internal transaction costs, can have a tracking difference higher than the TER.

For a beginning investor, the TER is still the first practical criterion to look at — tracking difference requires directly comparing the ETF's historical performance with that of the underlying index, data that isn't always easy to find, but worth knowing for large ETFs with several years of history.

Replication type and domicile

A physically replicated ETF actually holds the securities in the index it tracks, while a synthetically replicated ETF uses derivative instruments (swaps) to replicate its return. For beginners, physical replication is generally preferable for its simplicity and transparency. Also check the domicile: UCITS ETFs domiciled in Ireland or Luxembourg are almost always preferable for a UK investor, for reasons of tax efficiency on foreign dividends.

Trading liquidity: volume and spread, not just AUM

AUM is a decent broad indicator, but an ETF's practical liquidity is better measured with two other metrics: daily trading volume on the exchange and the bid-ask spread — the difference between the price you can sell at and the price you can buy at, at any given moment.

For regular purchases of modest amounts, even a spread of 0.1-0.3% has little impact. For larger orders, or sales during periods of market stress, a tight spread (ideally under 0.1% for the most liquid global-index ETFs) and high trading volume reduce the implicit cost of the transaction. Most brokers show spread and volume in real time before you confirm an order — worth checking, especially for less widely held ETFs.

Number of holdings and geographic and sector diversification

The number of holdings shows how many different companies make up the ETF: an MSCI World ETF, for instance, includes over 1,400 companies across 23 developed markets — very broad diversification that reduces the risk tied to any single company. More concentrated ETFs (a few dozen or a few hundred holdings, or focused on a single country or sector) offer less diversification and more exposure to specific risks.

It's also worth checking how the weight is distributed: if the top 10 holdings already account for 25-30% of the fund (common for tech-heavy indices like the Nasdaq 100), the real diversification is lower than the total holding count suggests. Similarly, a "global" ETF can still be geographically concentrated — MSCI World, for example, has historically carried a substantial weighting towards the US (roughly 60-70%), reflecting that country's market weight rather than an active allocation choice.

Fund history and provider: established player or new trend?

Two less technical but equally important factors: how long the ETF has existed and who manages it.

Launch date: an ETF with 10+ years of history has been through at least one full market cycle (growth, correction, recovery), giving you concrete data on how it's actually behaved — not just theory. An ETF launched recently around a "trending" theme (artificial intelligence, niche themes, whatever's currently popular) doesn't yet have that track record, and often comes with higher costs and lower liquidity in its early years.

The provider: established, large providers (e.g. iShares/BlackRock, Vanguard, Xtrackers/DWS, Amundi, SPDR/State Street) have been running ETFs for decades, with well-tested processes and lower operational risk. A new or niche provider isn't necessarily worse, but it comes with less history to base a judgement on — an extra reason for caution, especially for beginners.

Accumulating or distributing

Accumulating (Acc): dividends are automatically reinvested, accelerating capital growth and deferring any tax until you sell — preferable during accumulation. Distributing (Dist): dividends are paid out periodically to your account — more suited if you want a regular cash flow, typical of the income phase. For a more detailed comparison, see the dedicated FAQ Should I choose accumulating or distributing ETFs?

Simplicity: buy what you understand

The most underrated criterion is probably the most important one: only buy what you genuinely understand. If you can't explain in a single sentence what an ETF tracks, how it works, and why you're buying it, it's probably not the right instrument for you right now — however promising it looks. A single broad, diversified ETF like MSCI World, simple to understand and keep track of, almost always beats a pile of complex instruments that nobody ever reviews carefully.

Before buying an ETF, a practical 5-step checklist:

• Check TER and AUM on the fund's factsheet (KID/KIID) or the provider's website.
• Check the replication type and UCITS domicile.
• Look at the number of holdings and the top 10 by weight, to understand the real diversification.
• Check how long the fund has existed and who the provider is.
• If you can't explain in a single sentence what the ETF does, don't buy it — research it further first.