PipWardenStart free

Why most forex traders lose money: what the 74-89% figure means

Why do most forex traders lose money? What ESMA's 74-89% figure measures, how leverage, costs and losing streaks drive it, and what automation cannot change.

Key takeaways

  • On 27 March 2018, ESMA reported that national regulators found 74-89% of retail CFD accounts typically lose money, with average losses per client of €1,600 to €29,000.
  • The percentage on an EU broker’s website is that broker’s own figure, the share of its retail CFD accounts with a net loss over the previous 12 months, recalculated every three months.
  • ESMA singled out excessive leverage, because it magnifies costs relative to margin, shrinks the adverse price move that depletes the margin and goes with more frequent trading.
  • A 50% drawdown needs a 100% gain to recover, and at an assumed 50% loss rate, 100 trades hold a run of 6 losses in a row 54.6% of the time.
  • Automation can enforce sizing, loss limits and news rules on every trade, but it cannot give a strategy an edge, remove costs or promise anyone a place in the minority.
On this page
  1. The number: 74-89%
  2. What the warning on a CFD broker’s site measures
  3. Why do most forex traders lose money? Leverage, costs and complexity
  4. The limits ESMA set
  5. The maths that works against traders
  6. Losses need bigger gains to recover
  7. Losing streaks are normal
  8. Costs multiply with trade frequency
  9. What automation can change
  10. What automation cannot change
  11. Questions to ask before trading CFDs at all
  12. How PipWarden handles this

Most forex traders lose money because every trade starts behind by its costs, leverage magnifies those costs and every price move against the position, and losing streaks that are statistically normal do lasting damage at large position sizes. The primary figure comes from the EU regulator ESMA, which reported on 27 March 2018 that 74-89% of retail CFD accounts typically lose money.

Below: what that number measures, what drives it and what software can and cannot change. Forex and CFD trading is high risk and you can lose money. Nothing here is financial advice.

The number: 74-89%

On 27 March 2018, the European Securities and Markets Authority (ESMA) announced restrictions on contracts for differences (CFDs) sold to retail clients. Its press release gives the evidence behind them:

“NCAs’ analyses on CFD trading across different EU jurisdictions shows that 74-89% of retail accounts typically lose money on their investments, with average losses per client ranging from €1,600 to €29,000.”

Three details in that sentence matter:

  • NCAs are the national competent authorities, the financial regulators of each EU country. The range comes from their separate studies, not from one EU-wide dataset.
  • Retail has a legal meaning. Under MiFID II, a retail client is “a client who is not a professional client” (Article 4), and professional status depends on the criteria in Annex II.
  • Forex is included. A CFD is a cash-settled derivative that tracks the price of something else. ESMA’s analysis behind the measures counts rolling spot forex, a leveraged currency position that rolls over instead of settling, as a CFD, and its Q&A (Q5.12) confirms the measures apply to it.

ESMA’s own measures were temporary. They applied from 1 August 2018 and were renewed in three-month steps. Between March 2019 and April 2020, national regulators adopted measures of their own, which replaced ESMA’s.

What the warning on a CFD broker’s site measures

The percentage in an EU broker’s risk warning comes from these rules, now carried in national form. The wording for a website, from Annex II of ESMA’s CFD decision, reads:

“CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. [insert percentage per provider]% of retail investor accounts lose money when trading CFDs with this provider.”

The same annex sets how the percentage is worked out:

Rule in the decision What it means for the number
Provider-specific Each broker reports on its own clients, so the figure differs by broker.
Every three months, over the previous 12 months The figure moves from quarter to quarter and describes the last year.
Realised and unrealised net profits on CFDs An account counts as losing if its net result, open trades included, is below zero.
All charges, fees and commissions included Costs are part of the result.
Excludes accounts with no open CFD, non-CFD products, deposits and withdrawals Only CFD trading results count, not cash moved in or out.
No open retail CFD in the period The broker uses the standard “Between 74-89%” wording instead.

Two things follow. The unit is the account, not the person: someone with two accounts at the same broker is counted twice, and an account down €5 counts the same as one down €5,000. And the figure describes that broker’s past clients. It is not a forecast for any one trader, in either direction.

Why do most forex traders lose money? Leverage, costs and complexity

Leverage is the size of a position relative to the money needed to open it, the margin. At 30:1, €1,000 of margin controls a €30,000 position.

ESMA gave its reasons in the same press release: “complexity and lack of transparency” and “the particular features of CFDs – excessive leverage”. Its product intervention analysis of 1 June 2018 states that, as national regulators’ studies have shown, “high leverage causes poor outcomes for investors”. It names three ways:

  1. It magnifies costs, such as spreads, commissions and financing charges, relative to the margin put up.
  2. It amplifies losses. The higher the leverage, the smaller the adverse price move needed to deplete much or all of the margin.
  3. It goes with more trading, and every extra entry and exit pays the costs again.

The limits ESMA set

ESMA’s 2018 measures capped leverage for retail clients by the type of underlying:

Underlying Maximum leverage
Major currency pairs 30:1
Non-major currency pairs, gold, major equity indices 20:1
Commodities other than gold, non-major equity indices 10:1
Individual equities and other reference values 5:1
Cryptocurrencies 2:1

ESMA’s major pairs are any two of the US dollar, euro, Japanese yen, pound sterling, Canadian dollar and Swiss franc (Annex I of the decision). AUD/USD is therefore a non-major pair, capped at 20:1.

Three more protections came with the caps:

  • Margin close-out. The broker must close one or more positions when the funds in the account plus open profit or loss fall below half of the initial margin the rules require for them. In MT5, the margin level, equity ÷ margin × 100 in the MT5 help, shows how close an account is to the broker’s stop out.
  • Negative balance protection. Losses on CFDs are limited to the funds in the account. ESMA’s chair said the measures would “for the first time ensure that investors cannot lose more money than they put in.”
  • A restriction on incentives to trade, plus the standardised warning above.

The national measures that replaced ESMA’s carry the same elements, as ESMA restated on 24 February 2026; the exact figures are in each national regulator’s rules. They shrink positions and cap what a retail client can lose at the money in the account. They do not make any trade more likely to win. The margin calculator shows the margin a position ties up at a given leverage.

The maths that works against traders

Losses need bigger gains to recover

A drawdown is a fall from a previous high. A drawdown of d needs a gain of d ÷ (1 − d) to get back, because the gain has to come from a smaller balance.

Drawdown Gain needed to recover
10% 11.1%
20% 25.0%
30% 42.9%
50% 100.0%
75% 300.0%

Losing streaks are normal

Assume each trade loses 50% of the time, independently of the others. Over 100 trades, the exact chance of at least one run of 6 losses in a row is 54.6%. For 8 in a row it is 17.0%, and for 10 it is 4.4%.

What the streak costs depends on the size of each trade:

Risk per trade After 6 losses in a row Gain needed to recover
1% −5.9% +6.2%
2% −11.4% +12.9%
5% −26.5% +36.0%
10% −46.9% +88.2%

Each loss here is a share of the current balance, and the numbers are illustrative. Real trades are not fully independent: losses can cluster around news or a change in the market, so real streaks can run longer. The drawdown calculator runs both calculations with your own inputs.

Costs multiply with trade frequency

Every trade pays the spread, the gap between the buy and sell price, and often a commission. A position held past the daily rollover pays or receives swap. In MT5, right-click a symbol in Market Watch and choose Specification to see its contract size, swap long and swap short. The swap rates there carry a multiplier for each weekday, and the example in the MT5 help triples the swap on Wednesday.

Frequency shows up in the data too: ESMA’s analysis cites a 2014 study by France’s regulator, the AMF, which found that investor performance fell as leverage and the number of trades rose. The risk-reward calculator shows the break-even win rate for any stop and target, before and after spread and commission.

What automation can change

A rule-based Expert Advisor (EA), a program that trades inside MetaTrader 5, does not get tired, bored or angry. That helps with the mistakes that come from the trader rather than the market:

  • Consistent sizing. The lot size comes from the same risk rule on every trade, worked out from the balance and the stop distance, not from a feeling about the setup. See how to calculate lot size and the position size calculator.
  • A daily loss limit. Once the day’s losses reach a set share of the account, no new trades open. A person can argue with that rule after three losses. Code does not. See daily loss limit and how to set one.
  • No revenge trading. Code raises the size after a loss only if it is written to. Martingale systems are written to do exactly that, which is how they can end accounts.
  • Standing aside. Code can apply a news blackout, pausing new entries around high-impact releases such as NFP, and skip entries whenever the spread is wider than a set maximum.

What automation cannot change

  • An edge. An edge is a positive average result after costs, and discipline does not create one. An EA takes every signal its rules produce, with no second thoughts. If those signals lose on average after costs, automation only makes sure every one of them is taken. A strategy tuned too closely to past data (overfitting) can look strong in a backtest and fail on new prices.
  • Costs stay. A bot pays the same spread, commission and swap as a person, and one that trades often pays them more often.
  • Prices gap. MT5 triggers a stop loss when the bid (for a buy) or the ask (for a sell) reaches it, and then closes the position at market (MT5 help). After a weekend gap or during a fast news move, the fill can be worse than the stop (slippage). ESMA’s analysis recalls 15 January 2015, before the EU required negative balance protection, when the Swiss franc rose about 15% against the euro in one morning and some investors became liable for tens of thousands of euros. A limit checked before each order stops new trades; trades already open run to their own stops.
  • Machines stop. An EA runs only while the desktop MT5 terminal it is attached to is open and connected, which is why EAs are usually run on a PC or VPS that stays on.

The US Commodity Futures Trading Commission titled its customer advisory on the subject AI Won’t Turn Trading Bots into Money Machines. It states that “AI technology can’t predict the future or sudden market changes.” Fixed rules cannot either.

No bot, ours included, can promise that you end up among the accounts that do not lose. For how that promise is usually sold, read how to spot a forex robot scam, and for what a bot does day to day, read what a forex trading bot is.

Questions to ask before trading CFDs at all

The warning ends with its own test: “You should consider whether you understand how CFDs work and whether you can afford to take the high risk of losing your money.” In practice:

  • Can you afford to lose this money? All of it, not a bad month’s worth. Negative balance protection, where it applies, limits the loss to the account; it does not protect what is in it.
  • Do you understand leverage and margin? Where does a margin call warn you, where is the stop out, and what does one losing trade cost at your usual size?
  • Do you know your costs? The spread, commission and swap for each pair, against your stop distance.
  • What does your broker’s own percentage say? An EU-regulated CFD broker has to show it with its marketing.
  • Have you tried it on demo? A demo account shows how the platform, the sizing and the rules behave with no real money at stake, though demo fills and spreads can differ from live ones. See how to test an EA on a demo account.

How PipWarden handles this

PipWarden does not change these odds, and it does not claim to. Its job is narrower: it checks the limits you set before every order and records why each signal was taken or skipped. It runs as an EA in MT5 on a Windows PC or VPS, needs a broker that allows EAs and does not run on a phone.

  • Risk per trade, 1% by default. The lot size comes from that share and an ATR-based stop distance, capped by your maximum lot.
  • A daily loss limit, 5% by default. Once the day’s losses reach it, no new trades open until the next trading day. Trades already open run to their own stops, so a day can end past the limit.
  • Position caps, by default 3 open positions per account and 1 per pair.
  • A news blackout. By default, no new entries from 60 minutes before to 30 minutes after a high-impact release, for the currencies involved.
  • A maximum spread for each pair, if you set one.
  • A kill switch that cancels pending entries and closes every position the bot opened, leaving your manual trades alone.
  • Email and Telegram alerts when a limit is hit.

Every skipped signal is logged with its reason, such as “Daily loss limit reached”, “High-impact news” or “Spread too wide”. The strategy is rule-based trend-momentum, with no martingale and no grid. The optional AI filter, when available and switched on, can only veto a trade, never create or enlarge one. The full list is on the features page.

Forex and CFD trading is high risk, and you can lose money. Read the risk disclosure before you trade.

Frequently asked questions

What percentage of forex traders lose money?
The primary figure is ESMA’s: in March 2018 it reported that national regulators’ analyses found 74-89% of retail CFD accounts typically lose money, and rolling spot forex counts as a CFD under its rules. Each EU-regulated CFD broker also publishes its own percentage for its retail accounts over the previous 12 months, so the figure differs from broker to broker.
Why do CFD brokers show "X% of retail accounts lose money"?
EU product intervention rules require it. The warning must include the share of the broker’s own retail CFD accounts that lost money, calculated every three months over the previous 12 months with all fees and commissions included. It describes that broker’s past clients, not the chances of any one trader.
Does leverage cause most losses?
ESMA named excessive leverage as the feature of CFDs behind its concern. Its June 2018 analysis says high leverage hurts in three ways: it magnifies costs relative to margin, it shrinks the adverse move that depletes much or all of the margin, and it goes with more frequent trading. A strategy that loses on average after costs loses at any leverage; high leverage allows larger positions, so the losses can come faster.
Can a trading bot stop me from losing money?
No. A bot can apply the same position size, loss limits and news rules on every trade, which removes some human mistakes. It cannot give a strategy an edge, remove spreads and swaps, or stop prices from gapping past a stop. The CFTC’s advisory on AI trading bots says “AI technology can’t predict the future or sudden market changes.”
Is the "90% of forex traders lose money" figure true?
We could not find a primary source for a flat 90%. The regulator figure is ESMA’s range of 74-89% of retail CFD accounts, from national regulators’ analyses published in 2018, plus each EU broker’s own percentage, which changes by broker and by quarter. The top of the range is close to 90%, but no single number fits every broker and period.

Educational content, not financial advice. Forex and CFDs are traded on margin and are high risk: you can lose money, and more than your deposit with some brokers. Examples use made-up numbers and show no real results. Read the risk disclosure.