Capital at risk. CFDs and leveraged forex are complex instruments and most retail accounts lose money trading them. This article explains a regulatory disclosure; it is not a recommendation to trade, and nothing here is advice.
Every regulated CFD broker in the UK and the EU has the same sentence on its homepage, differing only in one number: a percentage of retail investor accounts that lose money when trading CFDs with this provider.
It is the most-seen and least-understood number in retail trading. People read it as a probability of failure, compare it between brokers as though it were a quality score, and take a lower one as reassurance. All three readings are wrong, and the rule that created the number says so explicitly.
The rule, and what it actually requires
The wording is not the broker's. It is prescribed by the regulator, and in the UK it lives in the FCA Handbook at COBS 22.5.6R (handbook.fca.org.uk), which requires a firm marketing leveraged CFDs, spread bets or rolling spot forex to retail clients to include a specific risk warning containing "[insert percentage per provider]% of retail investor accounts lose money when trading CFDs with this provider".
The same rule then defines the calculation with unusual precision, and the definition is where all the interesting parts are.
- It is recalculated every three months, covering the twelve-month period immediately before the calculation date (22.5.6R(4)). The number you see is a rolling annual figure refreshed quarterly, not a lifetime statistic.
- An account counts as losing if the sum of all realised and unrealised net profits on the relevant investments over that period is below zero (22.5.6R(5)).
- All costs count. The calculation must include all costs, fees, commissions and any other charges (22.5.6R(6)). Spreads and swaps are inside the number.
- Dormant accounts are excluded. An account with no open position during the period is not counted (22.5.6R(7)(a)).
- Deposits and withdrawals are excluded (22.5.6R(7)(c) and (d)). This is the big one.
Why excluding deposits and withdrawals changes everything
Read that last exclusion again, because it is the difference between what the number says and what people think it says.
The calculation measures profit and loss on positions, not the change in your account balance. Money you paid in and money you took out are both invisible to it.
So the figure cannot tell you how much anyone lost. An account that finished the year down a rounding error is one tick. An account that was funded five times and lost every deposit is also one tick. The two are identical to the calculation, which counts accounts, not amounts.
It is a headcount of unprofitable accounts, not a measure of money lost. Nothing in the disclosure tells you the size of the losses behind it.
That cuts both ways, and it is worth being honest about both. The number does not prove the losses were catastrophic. It also gives no comfort whatsoever that they were small.
Why comparing brokers on it is a mistake
This is the most common misuse, and it is worth being blunt: the percentage describes a broker's client base, not its product.
Two firms with identical spreads, identical execution and identical regulation will publish different percentages if their clients differ — and their clients always differ. A firm whose marketing reaches inexperienced traders with small accounts and high leverage will show a worse figure than one serving experienced clients trading conservatively, without a single thing about the service being better or worse.
There is even a perverse case. A broker with a stricter onboarding process, rejecting more unsuitable applicants, ends up with a client base that fares better — so the firm doing the most to protect people can end up with a number that looks similar to one doing very little.
What it is genuinely useful for
The disclosure was introduced as a product intervention, and read as one it is doing exactly its job. Its message is about the instrument, not the firm.
Leveraged CFDs and rolling spot forex are products where the typical retail outcome, across every regulated firm publishing a figure, is a loss. Not a marginal edge lost to costs — a majority of accounts, at every provider, every quarter, for years.
That consistency is the signal. If one firm published a bad number it would say something about that firm. When every firm publishes a bad number, it says something about the product, and the correct conclusion is the one the regulator intended: understand what you are trading and size it so that being in the majority is survivable.
The costs sitting inside the number
Since the calculation includes all costs, fees, commissions and charges, part of what it measures is not bad trading at all. It is the cost of trading, which applies whether you were right or wrong.
Three of them do most of the work.
- The spread, paid on entry and exit of every position, which our guide to spread vs commission breaks down properly.
- Swap or rollover, charged or credited each night a position is held, explained in forex swap and rollover fees.
- Slippage, the difference between the price you asked for and the price you got, covered in what slippage costs.
A strategy that is exactly break-even before costs is a losing strategy after them, and it appears in the percentage as one more losing account.
What to do with it
- Take it as a statement about the product. That is what it is for, and on that reading it is accurate and worth heeding.
- Do not rank brokers with it. Use regulation, cost structure and execution quality instead — how to choose a forex broker works through what actually differs.
- Check the date and the entity. The figure belongs to a specific legal entity and is refreshed quarterly; a group can run several entities with different numbers.
- Understand the protections that come with the same rulebook. Leverage caps, the margin close-out rule and negative balance protection arrive from the same intervention — see ESMA rules explained and the margin close-out rule.
- Do not read a lower number as permission. Every one of them is a majority.
Regulatory detail in this article is from the FCA Handbook, COBS 22.5, read on 19 August 2026. Rules differ outside the UK; check the regime that applies to the entity you trade with.
Frequently Asked Questions
Does a lower loss percentage mean a better broker?
Not reliably. The figure describes the broker's client base over the last twelve months, not the quality of its execution or pricing. A broker whose clients are mostly experienced and trade conservatively will show a lower number than one that markets to beginners, with identical spreads and identical execution. Compare brokers on regulation, costs and execution; treat the percentage as a warning about the product, which is what it was designed to be.
Does the percentage include my deposits and withdrawals?
No, and this is the detail most people miss. Under FCA COBS 22.5.6R(7) the calculation must exclude any deposits or withdrawals of funds. It looks only at realised and unrealised net profit on the positions traded during the period, including all costs, fees, commissions and charges.
How often is the number updated?
Every three months. COBS 22.5.6R(4) requires the calculation to be performed quarterly and to cover the twelve-month period immediately preceding the calculation date, so the figure on a broker's website is a rolling annual measure refreshed four times a year.
Does an account that lost one cent count the same as one that was wiped out?
Yes. COBS 22.5.6R(5) says an account counts as having lost money if the sum of its realised and unrealised net profits over the period is below zero. There is no severity weighting, so a client down a trivial amount and a client who lost their entire balance are one tick each.
