How we review brokers and ETFs

Each area is scored out of 100 on its own, then weighted. Trust and custody together carry the most weight, because over ten years whether a broker can be trusted with your money, and whether you actually own your shares, matters more than saving a few dollars in commission.

Each of the ten areas is scored out of 100. Those scores are multiplied by the weights above, added together, and divided by 20 to land on a figure out of 5. The result is rounded to one decimal place. Nothing is adjusted afterwards.

The word attached to the score comes from the bands above. It is never chosen by hand.

One thing can override the arithmetic. Where a broker fails on something a long-term holder cannot work around, the score is capped no matter what the calculation produces. Those limits are listed further down, and when a cap applies the review says so and gives the reason.

Points are earned, not deducted. A broker does not start at full marks and lose them; it starts at zero in each area and earns points for things we can verify. That matters, because a broker that discloses very little would otherwise look flawless simply by giving us nothing to mark down.

Below is what carries the most weight inside each area. Where we cannot find something, that item earns nothing, and we do not penalize the rest of the area for it.

Where information is genuinely hard to find, the item is worth few points by design. A broker outside the United States is not marked down for lacking a disclosure that only US regulation requires. Thin evidence is a limit on us, not a failing of the broker.

Every review shows an estimated ten-year cost. To make that figure comparable, we apply the same investor to every broker:

We then add up what the broker charges that investor: commission on each purchase, any custody or annual account fee, and the broker’s currency-conversion markup. Nothing else.

This is an estimate under one fixed set of assumptions, not a forecast of what you will pay. It is not a projection of returns, and it says nothing about what your investments will be worth. Your own costs will differ, sometimes by a lot, depending on how much you invest, how often, in which currency, and which account you use. The figure exists to compare brokers on one basis, not to budget with.

Unlike some corners of the trading industry, a real share-dealing broker is licensed. That makes the strength of the license something we can check rather than guess at, and it is the backbone of the trust score. We group regulators by how strictly they supervise and whether they stand behind a compensation scheme.

We look at the tier of the entity that actually holds your account, and many global brokers run several. A well-known global brand operating through a Tier 3 entity in your country is scored as Tier 3, because that is the license standing behind your shares.

Some problems matter so much to someone holding for a decade that no amount of low fees can make up for them. When one applies, the score is capped no matter what the arithmetic produces:

When a cap applies, the review says so. You will see the reason next to the score breakdown, not just a lower number. Where the facts are still disputed, or the broker has made affected clients whole, we score through the criteria instead of applying a cap.

A score is only useful if you know what it leaves out. This one leaves out a great deal.

We build each score from the broker’s own disclosures, from public regulator records, from reputable financial reporting, and from what long-term customers report publicly. These are checked against each other rather than taken at face value. Where a broker’s marketing claims something its own terms do not support, we use the terms document and note the gap.

We also do not accept a broker’s own summary of its regulatory standing. If a claim about a license, a compensation scheme or an enforcement action cannot be traced to a regulator’s own record or to reputable reporting, it does not earn points.

See every broker we score →

Everything above is about brokers. This part is about the funds themselves, and it works differently.

An ETF never gets a Stockember Score. A broker can be measured the same way for everyone: its license, its custody arrangement and its fee schedule are the same facts whichever customer is asking. A fund is not like that. A US-domiciled S&P 500 ETF and its UCITS equivalent hold the same companies, but which one suits you depends on where you live, which account you hold it in, and how your dividends are taxed. Reducing that to one number would be telling you something we cannot know.

So an ETF review publishes the facts that decide long-term cost and fit, and says who a fund suits. It never tells you to buy one.

What every ETF review shows

These are reported from the fund provider’s own factsheet and prospectus, and dated. Where a figure moves, such as the expense ratio or the size of the fund, the review shows when it was last checked.

What an ETF review is not: it is not a recommendation, a ranking, or a view on whether the underlying index is a good place for your money. We assess cost and fit. Choosing what to hold, and whether to invest at all, is your decision. See the ETFs we cover →