Charge comparison tables assume a customer who is not you

This is written to be used rather than admired. Each section below is a decision about comparing platform charges, and each one has a default.

Every table has a hidden customer

A comparison ranks platforms using assumptions about balance, number of holdings, trading frequency and the mix of account types. Change any one of those and the ranking changes, sometimes reversing entirely between the top of the table and the middle. The assumptions are usually stated somewhere, and they are almost never the assumptions that describe your own situation.

A table built around a large portfolio traded often will favour completely different providers from one built around small monthly contributions. This is not a criticism of the tables, which have to assume something, but it is a reason not to read the ranking as an answer.

The four numbers you actually need

Your balance, your number of holdings, your trades per year and the account types you use between them determine your real cost. Those four are enough to compute a total under any published charging structure in a few minutes with a calculator. Doing that for three shortlisted providers is far more informative than reading any number of general comparisons.

Put simply, the exercise also reveals which of the four is driving your cost, which is frequently not the one you expected. For most regular contributors the driver is the percentage fee; for frequent traders it is the per-deal charge instead.

What the tables leave out

Interest retained on cash rarely appears, and it can represent a substantial part of what a provider earns from you. Currency conversion margins rarely appear either, and they matter to anybody holding foreign-listed assets at all.

Exit and per-holding transfer charges are usually omitted because they are one-off rather than annual costs. Fund availability is not a charge but determines whether the cheapest share class is even reachable on that platform. A ranking built purely on annual percentage fees is measuring one of at least five things you actually pay.

Service is a real variable

Transfer speed, reliability of the interface and the ability to reach a person matter far more during a problem than during a normal year. These things are hard to quantify, which is exactly why they rarely appear anywhere near the numbers in a comparison. They are also what people cite when they leave a provider, rather than the charges they originally compared so carefully.

Longevity matters too, since a provider likely to be acquired or repriced is imposing future administration on you.

None of this is measurable, and pretending the decision is purely arithmetic ignores the part that most often causes regret.

The cost of comparing too often

Charges change slowly and transfers are slow and effortful, which makes an annual comparison sufficient for very nearly everybody. Continuous comparison produces a stream of small potential savings that are rarely worth the disruption required to capture them. A switch justified by a saving of a few units of currency a year will cost far more than it recovers.

On an ordinary week, set a threshold in advance, expressed as a saving worth some stated proportion of the balance, and act only above it. A rule decided before you look protects you from the version of yourself who has just read a persuasive comparison.

A ten-minute version

Write your four numbers down once and keep them alongside your written plan where you will find them again. Once a year, apply them to your current provider and to the two most plausible alternatives you would actually consider.

Compare the resulting totals rather than the rates, and include the cost of exiting in that calculation. If the difference falls below your stated threshold, do nothing at all and record that you checked. The record of having checked is what stops the question resurfacing every time somebody publishes a new ranking.

The takeaway

Build the shortlist from the table. Do the arithmetic with your own four numbers.

Pick the one that costs you least, and let the rest wait.