ما الذي تخبرك به نسبة الخسائر فعلًا

It sits under every CFD advert in Europe and at the foot of every regulated CFD provider's homepage, ours included: a percentage, a sentence, no further explanation. It is arguably the most standardised piece of honesty in retail finance: every provider computes its number from the same definition, on the same schedule. That makes it genuinely useful, and genuinely easy to over-read. This article traces the sentence to its source, walks through the required calculation, and is precise about what a single percentage cannot tell you.

Where the sentence comes from

The wording comes from the European Securities and Markets Authority. In March 2018, ESMA announced EU-wide restrictions on the marketing, distribution and sale of CFDs to retail clients ¹. It was the first time ESMA had exercised its product-intervention powers under Article 40 of MiFIR ². The package had five parts: leverage limits from 30:1 down to 2:1, a 50% margin close-out rule, negative balance protection, a restriction on incentives offered to trade CFDs, and a standardised risk warning that must include the percentage of the provider's own retail investor accounts that lose money. How ESMA reshaped retail CFD leverage covers the leverage caps; this article is about the warning.

[insert percentage per provider]% of retail investor accounts lose money when trading CFDs with this provider.

Mandated warning template, ESMA Decision (EU) 2018/796, Annex II, Section B

The intervention did not come out of nowhere. Analyses by national regulators across the EU had found that 74–89% of retail CFD accounts lose money, with average losses per client ranging from €1,600 to €29,000 ¹. France's AMF published one of the starkest in October 2014: of 14,799 active retail clients at the firms it surveyed, more than 89% ended a four-year observation window negative, down €10,887 on average ³.

Loss-making accounts, EU regulators' analyses

74–89%^†

Average loss per client, same analyses

€1,600–€29,000^†

Losing clients over four years (AMF, France)

89.4%^†

The formal text is ESMA Decision (EU) 2018/796 of 22 May 2018, published in the Official Journal (OJ L 136) on 1 June 2018 and renewed at three-month intervals; Article 40 measures are temporary by design ⁴. Permanence came from national regulators. In the UK, the FCA's policy statement PS19/18 made ESMA's temporary measures permanent from 1 August 2019, keeping the warning and deliberately aligning its calculation method with ESMA's ⁵. That is why the same sentence, carrying different numbers, appears across the EU and the UK.

How the percentage must be calculated

The decision's Annex II does not leave the number to marketing judgement. It prescribes the calculation, and the definition rewards close reading ⁴.

Three consequences follow. First, the unit is the account over twelve months, not the trade: a client who loses on sixty positions and recovers everything on the sixty-first is not loss-making. Second, open positions count: the test sums realised and unrealised profits, so an account holding a losing open position is loss-making even though nothing has been closed. Third, the test is net of every cost, so an account flat on price but paying spread and swap charges all year lands in the losing column. The FCA added one boundary clarification when it adopted the method: an active account with exactly zero change counts as profit-making ⁵.

A provider with no trading history cannot invent a number. Under ESMA's decision it must display the standard range (between 74% and 89%) drawn from the regulators' analyses ⁴; the FCA instead requires new firms to state, in wording aligned with Austria's regulator, that the vast majority of retail client accounts lose money ⁵. Absence of data is disclosed as the industry base rate, never as silence.

What the number does not measure

It is not the probability that a trade loses. The warning compresses every trade an account made in a year into one signed number. A provider's percentage says nothing about the odds on any single position: win rates and account outcomes are different quantities, connected only through costs and position sizing.

It is not the size of anyone's loss. The test is binary: an account down €5 after costs and an account down €50,000 both count once. The magnitudes regulators saw were severe: the AMF found a mean result of −€10,887 against a median of −€1,843, a gap that says the average was dragged by a minority of large losses ³. None of that distribution survives into the headline figure. Nor does the other tail: accounts that did not lose money include accounts that made almost nothing.

It is not a census of people. The unit is an account at one provider, and the annex excludes any account with no open CFD in the trailing twelve months. A client who lost heavily and walked away over a year ago has left the statistic; a trader with accounts at three brokers is three data points; the figure cannot say whether the same people keep losing or a rotating cast each loses once. Why most traders lose money looks at what longer studies add.

And it is not a lifetime outcome. Twelve months is a short window. The AMF's data made the point precisely: in any given year, roughly 84% of the clients it studied lost money, yet across the full four years more than 89% did ³. Extend the horizon and the share of accounts that have been net losers grows: a 12-month figure is closer to a floor for long-run experience than an estimate of it.

It is a base rate for an activity, measured at one provider over one year — not your odds on a trade, and not the size of anyone's loss.

Why the number varies between brokers

Put two regulated brokers side by side and their percentages will usually differ. The calculation is identical, so the difference lives in the inputs. Client mix: a book of experienced, well-capitalised accounts aggregates differently from a book of first-year accounts. Instrument mix: major FX pairs at 30:1 and crypto CFDs at 2:1 produce different P&L distributions. Costs, because the test is net of them: wider spreads and heavier financing tip marginal accounts into the losing column.

Timing matters too. The number is a quarterly snapshot of a trailing twelve-month window: a volatile quarter rolls in, and a year later rolls out, moving the published figure without any change in client behaviour. Two firms recalculating on different dates describe different slices of the same market, and a new firm shows standard wording rather than a number. The honest conclusion: small gaps between providers are weak evidence of anything. The percentage is a property of a client base and a window as much as of a broker.

How to actually use it

Used with its limits in view, it is still one of the most informative sentences a broker publishes. Four readings that respect what it measures:

• Treat it as the base rate of the activity. The honest prior is that most accounts like yours ended the year negative. Ask what, specifically, would make yours the exception (a cost edge, a sizing discipline, a longer horizon) rather than assuming one.
• Track one provider through time rather than ranking providers. A firm's own number drifting over several quarters says something about conditions; a small gap between two firms mostly says their client bases differ.
• Read it next to costs. The test is net of all charges, so the percentage already contains the spread, commissions and overnight financing, costs a practice account may not fully show. The full cost of a CFD trade itemises them.
• Never read it as a per-trade probability, a loss size, or a verdict on you. It is an aggregate of other people's accounts, over a window that closed before you arrived.

North Stone Capital publishes the same class of warning, computed the same way: account-level, net of all costs, over a rolling twelve-month window, recalculated quarterly. Deliberately, our current figure is not quoted in this article: a quarterly number belongs next to the products it describes, where it is kept current, not in an editorial page where it would quietly go stale. What does not change is the definition above: at ours as at every regulated provider, that is what the number means, neither more nor less.

ESMA's warning did something rare in financial marketing: it forced every provider to publish an unflattering statistic about its own customers, calculated to a shared public definition. Take it seriously, as a base rate, read in full knowledge of what it is not.