Is Forex Trading Profitable? Read the Evidence Before the Claims
A forex account can finish a period in profit. Whether it did so after costs, whether the trader could withstand the losses along the way, and whether the result is likely to repeat are three different questions. To assess a claim, start with the period, the account type and the definition of profit. Those details tell you what the result can support.
What the retail loss figures measure
In March 2018, ESMA reported that national regulators had found that 74–89% of retail CFD accounts typically lost money. This was historical account-level evidence across jurisdictions. It does not give a current probability for every forex trader, and it does not describe ordinary currency exchange. If a broker publishes a different percentage, check the entity, jurisdiction, product and period before comparing it.
Also check the unit being counted. One person can hold more than one account, and an account that broke even during one window has shown little about its next window. A vendor can present a genuine profitable account while leaving you with no evidence about typical outcomes across customers. To assess a claim about a group, ask how the sample was selected and whether failed or closed accounts were included.
Calculate what the account kept
In a hypothetical example, trades generate £600 before expenses. Trading costs are £220 and the software and hosting cost £100. The resulting £280 is before any applicable tax. If trading costs were already included in the statement, subtracting them again would understate the result. Depending on the setup, the cost rows can include spread, commissions, slippage, financing and currency conversion. Keep those rows separate so you can see which assumptions matter.
Then compare closed trades with equity. A statement may show £500 in closed profit while open positions are losing £800. At that moment the combined trading result in this simplified example is a £300 loss, before any further costs. For a meaningful return percentage, record what capital was exposed and how deposits or withdrawals affected the account. A percentage calculated on required margin answers a different question from one calculated on the total capital available to absorb losses.

A high win rate can still lose money
Suppose a hypothetical set of 100 trades includes 60 wins at £50 and 40 losses at £100. Gross gains are £3,000 and losses are £4,000, leaving a loss of £1,000 before costs. With 40 wins at £200 and 60 losses at £100, gross gains are £8,000 and losses are £6,000, for a gain of £2,000 before costs. The examples show how outcome size changes expectancy. They do not predict that changing a stop or take-profit setting will produce either result.
Use the same convention for every input. Multiply the average win by the probability of a win, subtract the average loss multiplied by its probability, then deduct average costs that have not been accounted for. Identify whether the result is based on one volatile period or several distinct conditions. If many tickets depend on the same price movement, their large count does not give you that many independent observations. Read any expectancy figure alongside drawdown and open exposure.

Separate the rule from the evidence
Automation can apply a defined rule consistently. It can also repeat a bad rule, fail to recover from an execution error or accumulate exposure you did not intend. Before treating a test as evidence, write down the EA version, settings, broker, account type, symbols and reporting period. Then check whether the reported spread and fills resemble what you can observe in that setup. Keep simulations, demo observations and real-money results clearly labelled.
When comparing an EA such as Fortune MT5, use the listing to identify the product and its documented configuration. Evaluate any selected account history on its own terms. Ask whether it used the same version and settings, whether costs and open positions are visible, and whether you can reconcile the reported result with independent account history. A listing and a live record have different roles. Neither tells you what another account will earn.
Record a review you can repeat
Start with a fixed reporting window and save the inputs before checking the outcome. Record realised P&L, open exposure, commissions and financing, deposits and withdrawals, and the largest equity drop in the period. Mark missing fields explicitly. If you run a forward test, compare its costs and trade behaviour with your prior assumptions before changing settings. Two rolling three-month results that share two months of trades provide overlapping evidence.
Pay particular attention to concentration. Check how much of the result came from a handful of days or one market move. Decide in advance what would stop an evaluation, such as a repeated execution failure, a cost assumption that no longer holds or a loss larger than the amount you can afford to absorb. A short favourable test can justify collecting more evidence. A required income figure cannot establish an edge.
Where independent reporting is available, reconcile what it actually verifies. An account may show that trades took place at specified times and prices while still leaving uncertainty about strategy attribution or how the account was selected. Record that uncertainty alongside the return and drawdown. If you can verify only one account during one period, describe only that result.
Leveraged forex and CFD trading can lead to substantial losses. Historical and hypothetical results do not predict future returns. Read the full SmiteFX Insights article for more on capital basis, selection bias and the limits of account-level evidence.


