The mathematics of forex trading connects a market view to a financial result. A forecast that EUR/USD will rise says nothing about how much an account will gain, how much it could lose, or whether the trade is worth taking after costs. Those questions require position size, entry and exit prices, currency conversion and an estimate of execution costs.
The calculations themselves are mostly straightforward. The harder part is using inputs that match the trade actually being placed. A pip value quoted in dollars may be wrong for a sterling account. A stop distance measured from a chart midpoint may omit part of the spread. A profitable backtest may depend on fills that were never available.
A mathematical approach also separates quantities that can be calculated from quantities that must be estimated. Contract size and commission can usually be established before an order. The probability of winning, future slippage and the largest forthcoming drawdown cannot.
The examples below use hypothetical prices, costs and account balances. Their purpose is to show how the calculations fit together, from a single currency movement to the results of a trading strategy.
Currency Quotes, Lots and Pips
What the exchange rate measures
In EUR/USD, the euro is the base currency and the US dollar is the quote currency. A price of 1.1000 means one euro is worth $1.10. Buying EUR/USD creates exposure that benefits from a rise in the euro relative to the dollar. Selling the pair reverses that directional exposure.
Position size is measured in units of the base currency. A 25,000-unit EUR/USD position represents €25,000 of exposure. At 1.1000, its dollar notional value is $27,500. The notional value measures the position being controlled, rather than the amount deposited as margin.
Under the usual retail forex convention, a standard lot contains 100,000 base currency units, a mini lot contains 10,000 and a micro lot contains 1,000. These relationships appear in OANDA’s explanation of lot sizes. The broker’s contract specification and order increments still need checking, particularly where a platform accepts units instead of lots.
Converting pips into money
For most commonly traded currency pairs, one pip is 0.0001. For pairs quoted in Japanese yen, it is usually 0.01. A platform may display an extra decimal place representing a tenth of a pip, as explained in OANDA’s guide to pip conventions. Confusing the final displayed digit with a whole pip creates a tenfold error.
Pip value in the quote currency equals the number of base currency units multiplied by the pip increment. For 25,000 units of EUR/USD, that is 25,000 × 0.0001, or $2.50 per pip. A 40-pip movement therefore changes the position’s value by $100 before separately charged costs.
For 100,000 units of USD/JPY, the calculation is 100,000 × 0.01, or ¥1,000 per pip. The result is in yen. It is not automatically $10 simply because the position is one standard lot.
Pips also differ from percentage returns. A move in EUR/USD from 1.1000 to 1.1100 is 100 pips, but the exchange rate has risen by approximately 0.91%. The percentage calculation is 0.0100 divided by 1.1000. Keeping price changes, pip counts and account returns separate prevents several common reporting errors.
Calculating Forex Profit and Loss
For a long position, profit or loss in the quote currency equals position size multiplied by exit price minus entry price. For a short position, reverse the subtraction. Use the executed prices for both transactions. This is the basic structure used in OANDA’s account statement calculations.
A trader buys 25,000 EUR/USD at 1.1000 and sells at 1.1040. The calculation is 25,000 × (1.1040 − 1.1000), producing a $100 trading profit. Selling at 1.0960 instead produces a $100 trading loss.
If those are actual fills, their difference already reflects the bid and ask prices at which the trader transacted. Subtracting another spread charge from that result would count the same cost twice. Separately booked commission, financing and conversion charges still need to be reconciled with the broker’s statement.
When the account currency is different
A position’s quote currency determines the currency of its initial profit calculation. The account currency determines how that result appears in the account.
Suppose a dollar account holds a long position of 100,000 USD/JPY, opened at 150.00 and closed at 151.00. The gain is 100,000 × 1.00, or ¥100,000. Converting at an assumed rate of ¥151 per dollar gives approximately $662.25, before conversion charges.
At the entry rate, a single pip was worth approximately $6.67. Multiplying that entry pip value by the 100-pip gain would give approximately $666.67. The difference arises because the dollar value of the yen profit changed as the exchange rate moved. A current pip value is useful for estimating exposure, but it need not remain fixed throughout the trade.
Crosses introduce another conversion. A 25,000-unit EUR/GBP position has a pip value of £2.50. If GBP/USD is 1.2500, that is $3.125 per pip for a dollar account. For a sterling account, the value remains £2.50.
These conversions matter when comparing trades across pairs. A journal reporting only pips can make a small monetary gain appear larger than a substantial monetary loss. Results should also be recorded in account currency and, where useful, as a percentage of the capital exposed.
Position Sizing From a Defined Loss Budget
Position sizing starts with the amount the trader is prepared to expose to a planned loss. It then works backwards from the exit assumption to the number of units.
Consider an account with $10,000 of equity and a hypothetical loss budget of 1%, or $100. This percentage is an illustration, not a universally suitable risk level. The account’s other positions, the strategy’s behaviour and the trader’s capacity to absorb losses all affect the decision.
Suppose a EUR/USD trade has a planned distance of 25 pips between the entry fill and the expected stop exit. With a dollar account, one standard lot is worth $10 per pip. Ignoring separate charges and slippage, the size would be $100 divided by 25 × $10, or 0.40 lots.
The denominator describes the loss on one lot. The numerator describes the permitted planned loss. Mixing dollars with pounds, or units with lots, breaks the calculation even when the arithmetic is correct.
Including costs without counting them twice
Now assume the broker charges $7 per standard lot for the complete opening and closing transaction. The expected loss per lot becomes $250 from the price movement plus $7 commission, or $257.
The revised size is $100 divided by $257, approximately 0.3891 lots. If the platform accepts increments of 0.01 lots, rounding down produces 0.38 lots. Its planned loss is $95 from the 25-pip movement plus $2.66 commission, totalling $97.66.
That leaves little allowance for a worse fill. Adding a hypothetical one-pip adverse slippage allowance changes the loss per lot to $267. The calculated size becomes approximately 0.3745 lots, rounded down to 0.37. Under those assumptions, the planned loss is $98.79.
The spread is not added again here because the 25-pip distance runs from the assumed executed entry price to the expected executable exit price. A model based on midpoint prices would require a separate spread adjustment. Every calculation needs a clear price convention.
Why a stop does not create a fixed maximum loss
An ordinary stop does not guarantee its execution price. During a gap or rapid movement, the available fill can be beyond the requested level, a limitation described in OANDA’s guidance on volatility and stop orders.
A one-pip allowance therefore describes an assumption, not a contractual ceiling. A useful sizing exercise also calculates the result with several worse exit prices. If an additional ten or twenty pips would produce an unacceptable account loss, the position may be too large for the conditions.
Changing the stop also changes the trade. Tightening it solely to fit a larger position can cause normal price movement to trigger exits. Widening it after entry increases the monetary exposure unless position size is reduced. The stop location and position size need to be evaluated together, using the same assumptions throughout.
Margin, Leverage and Account Exposure
Margin is collateral required to support a position. It is separate from the amount expected to be lost at a stop.
For a simple margin arrangement, required margin equals notional exposure multiplied by the margin rate. A 5% margin rate corresponds to maximum leverage of 20:1. Actual requirements can depend on the instrument, account, position size and broker rules, as illustrated by OANDA’s margin explanation.
Take a separate example of 38,000 EUR/USD units at 1.1000. The dollar notional value is $41,800. At a 5% margin rate, the required collateral is $2,090.
With $10,000 of account equity, the position’s effective leverage is $41,800 divided by $10,000, or 4.18:1. The account is not using 20:1 merely because that maximum is available.
If the same position could instead be opened at 30:1, its initial margin would be approximately $1,393.33. Its pip value would still be $3.80. A 25-pip adverse movement would still produce a $95 trading loss before separate costs. Changing the permitted leverage changes the collateral requirement; it does not change the price exposure of an unchanged position.
This distinction becomes more important with several open trades. Free margin is generally equity less used margin, subject to the broker’s definitions. Floating losses reduce equity and can bring the account towards forced liquidation. An individual trade’s stop may never be reached if losses elsewhere cause the broker to close positions first.
Regulatory limits also depend on jurisdiction and client classification. The FCA’s rules for retail CFD providers include leverage restrictions and account protection requirements, and cover rolling spot forex within their scope. A leverage figure advertised by an overseas entity should not be assumed to apply to every customer.
Margin calculations establish whether the broker will support the position. Loss calculations establish what adverse movement would do to the account. Both are needed before an order is placed.
Broker Costs and Their Effect on the Numbers
A strategy trades through a particular account, with particular contract terms. Its results depend on those terms as well as on the currency market.
When comparing brokers, ForexBrokersOnline.com reviews can provide a starting point for researching available accounts. The numbers used in a trading model should then come from the relevant broker’s current contract specifications, fee schedule and execution records. A comparison page cannot establish the exact cost of a future trade.
The legal entity serving the account also matters. Accounts carrying the same brand can have different pricing, product availability and margin rules. Calculations should match the entity and account type actually being used.
Putting spreads and commissions on the same basis
Suppose Broker A offers a 1.2-pip spread with no separate commission. Broker B offers a 0.2-pip spread and charges $7 per standard EUR/USD lot for opening and closing combined.
For one standard lot in a dollar account, Broker A’s spread cost is $12. Broker B’s spread cost is $2, with $7 commission added, producing a $9 total. Under these hypothetical conditions, Broker B is $3 cheaper for the completed trade.
The comparison assumes the same position size, an unchanged midpoint and spread, and no slippage or financing. It also assumes the $7 charge covers both sides. If a quoted commission applies separately to entry and exit, it must be doubled before making the comparison.
An unchanged 1.2-pip spread ordinarily creates a 1.2-pip loss on an immediate purchase and sale, rather than 2.4 pips. Buying at the ask and selling at the bid crosses the spread once in total. When using midpoint-based calculations, the opening half spread and closing half spread produce that result.
Advertised minimum spreads provide little information about the average cost of a strategy trading during different sessions. A broker comparison becomes more useful when it uses prices observed at the times, sizes and order types the strategy actually trades.
Measuring slippage in cash
Slippage is the difference between an expected execution price and the actual fill. For a buy order, a higher fill is adverse. For a sell order, a lower fill is adverse.
Suppose the expected EUR/USD ask is 1.10005 and a purchase fills at 1.10011. The difference is 0.00006, or 0.6 pips. On 25,000 units, with a pip value of $2.50, the adverse impact is $1.50.
The slippage cost calculator on forex.ke lets readers enter expected and executed prices, position size and pip value to express the difference in money. Its currency conversion input can also translate a dollar amount into Kenyan shillings. The inputs should match the actual instrument and conversion rate.
Compare fills with the executable ask for a purchase or bid for a sale at the chosen measurement time. Comparing a purchase with a chart midpoint wrongly attributes part of the spread to slippage. If slippage is already reflected in actual entry and exit prices, do not deduct it again when calculating realised profit.
Financing and holding time
Positions held across a broker’s funding cutoff can incur financing charges or credits. The applicable rate depends on the pair, direction and account terms. OANDA’s financing documentation provides examples using an annualised rate, a day fraction and conversion into account currency.
For a hypothetical position with a $40,000 funding basis, a 6% annual charge and a 365-day convention, one day costs approximately $6.58. The calculation is $40,000 × 0.06 ÷ 365.
That amount relates to the stated funding basis, not the margin deposit. Actual schedules can include multiple days of funding for weekends or holidays. A strategy holding for several days therefore needs a different cost model from one closing before the relevant cutoff.
Expectancy and the Break-Even Win Rate
The proportion of winning trades does not determine profitability by itself. Profitability also depends on the size of wins, the size of losses and the costs incurred to obtain them.
For a simplified strategy with winning and losing outcomes, expected net profit per trade equals the probability of a win multiplied by the average gross win, minus the probability of a loss multiplied by the average gross loss, minus average trading costs.
Written compactly, the calculation is E = pW − (1 − p)L − C. Here, p is the win probability, W is average gross profit on winning trades, L is average gross loss on losing trades expressed as a positive amount, and C is average cost across all trades.
Assume a 45% win rate, a $150 average gross win, a $100 average gross loss and $5 average cost. Expectancy is 0.45 × $150 − 0.55 × $100 − $5, or $7.50 per trade.
This is an average implied by the assumptions. It does not mean the next trade, or the next hundred trades, will deliver that amount. The inputs describe a distribution whose realised results can vary.
Calculating the required win rate
Setting expectancy to zero and rearranging gives the break-even win probability: p = (L + C) ÷ (W + L).
Using the same figures, the required rate is ($100 + $5) divided by ($150 + $100), or 42%. The assumed 45% win rate clears that threshold by three percentage points.
If average cost rises to $10, the break-even rate becomes 44%, and expectancy at a 45% win rate falls to $2.50. At $15 average cost, the break-even rate becomes 46%, leaving expected profit at negative $2.50 per trade.
This shows why apparently small execution differences can matter to a strategy with a narrow expected advantage.
Using measured outcomes rather than targets
A stated target of twice the stop distance does not establish an average realised win of twice the average realised loss. Partial exits, early closures, trailing stops and slippage can change both figures. The calculation needs actual outcomes, or clearly identified modelling assumptions.
Cost treatment must also be consistent. If average wins and losses already include all charges, the cost term should not be subtracted again. If the inputs come from midpoint price changes, spreads and execution costs still need incorporating.
A strategy winning 60% of its trades can lose money. With $50 average gross wins, $100 average gross losses and $5 average costs, its expectancy is $30 − $40 − $5, or negative $15. A high win rate does not repair an unfavourable payoff distribution.
Statistical Uncertainty in Trading Results
A trading record provides estimates of future behaviour. It does not reveal an unchanging win probability or a permanent average payoff.
Suppose a strategy wins 45 of its first 100 trades. The observed win rate is 45%, but the sample contains substantial uncertainty.
Under a simplified model of independent trades with a constant win probability, a 95% Wilson confidence interval is approximately 35.6% to 54.8%. The calculation follows the method described in the NIST guidance on confidence intervals for proportions. A break-even rate of 42% lies within that range.
That interval is not a prediction range for the next hundred trades. Nor does it establish profitability, because average wins, average losses and costs are also estimated. It illustrates how imprecisely a win probability can be measured from a seemingly useful record.
Why the observations may not be independent
Five entries taken during one sustained dollar move may share much of the same underlying exposure. Counting them as five wholly independent tests of a strategy overstates the amount of separate evidence.
A record concentrated in one market condition also says little about conditions absent from the sample. A system tested during persistent trends has not necessarily demonstrated an ability to handle repeated reversals. More observations help only to the extent that they provide relevant information.
The same issue appears when a trader repeatedly changes parameters after inspecting results. If hundreds of combinations are tried, the best historical result may owe much of its performance to chance. Evaluating that chosen version on the same observations rewards the selection process twice.
A more informative test fixes the rules before evaluating a later, unused period. Any changes made after seeing those results should be treated as further development, requiring fresh evaluation.
The record should preserve losing trades, rejected signals and execution differences. Removing inconvenient observations makes the statistics cleaner while making the evidence worse.
Uncertainty can also be tested directly. Recalculate profitability with a lower win rate, smaller average wins, larger losses and higher costs. A strategy that becomes unprofitable after a minor change in any one input depends heavily on assumptions that may be estimated only loosely.
Drawdowns, Compounding and Combined Exposure
A strategy’s average result does not describe the path an account takes to achieve it. The order and size of losses matter because capital is finite and position sizes may change with equity.
After a 20% drawdown, the account needs a 25% gain to recover. After a 50% drawdown, it needs a 100% gain. The recovery formula is drawdown divided by one minus drawdown, using decimals.
For a 20% loss, that is 0.20 divided by 0.80, or 0.25. The asymmetry exists because the subsequent gain is earned on a smaller amount of capital.
Fixed cash risk and fractional risk
Suppose a $10,000 account loses exactly 1% of its current equity on each of twenty consecutive trades. Its remaining balance is $10,000 × 0.99 to the power of 20, approximately $8,179.07. The drawdown is approximately 18.21%.
At a 5% loss of current equity per trade, the calculation becomes $10,000 × 0.95 to the power of 20. The balance falls to approximately $3,584.86, a drawdown of 64.15%.
These figures assume every loss matches the stated fraction, including any costs. They describe the consequences of the sequence, not its probability. Gaps, minimum trade sizes and departures from the sizing rule can make actual results different.
Fixed cash risk behaves differently. Losing $100 on each of twenty trades takes a $10,000 account to $8,000. As the account shrinks, the unchanged $100 represents a larger fraction of the remaining capital. A sizing policy should state whether it uses starting capital, current equity or another clearly defined basis.
Compounding depends on the whole return sequence
A 10% gain followed by a 10% loss leaves the account at 99% of its starting value. Multiplying 1.10 by 0.90 gives 0.99. The arithmetic average of those two returns is zero, but the compounded return is negative 1%.
This is one reason to examine more than average trade profit. A strategy can have attractive typical outcomes while suffering occasional losses large enough to dominate its compounded results. Increasing position size amplifies both the ordinary variation and the damaging extremes.
Several trades can express the same currency view
A long EUR/USD position and a long GBP/USD position both carry exposure against the dollar. They also contain separate euro and sterling risks, but a broad dollar rally can hurt both together.
If each trade has a planned loss of 1% of equity, both reaching their exits implies roughly 2% of planned account exposure, before any additional slippage or charges omitted from those budgets. Assigning the trades to different currency pairs does not remove that combined possibility.
Historical correlation can help describe how returns have moved together, but it does not guarantee the relationship will persist. A practical review also looks at the currency amounts bought and sold, the economic event affecting the positions and the losses under a shared adverse move.
The same reasoning applies to supposed hedges. Opposing positions may reduce some directional exposure while retaining basis risk, separate costs or mismatched sizes. Their combined cash flows need to be calculated rather than inferred from the labels.
Applying the Calculations to a Trading Process
The useful output of forex mathematics is a trade record that can be reconciled with the account statement and a risk estimate whose assumptions are visible.
Before entry, the record should connect the instrument’s contract size to its pip value, convert that value into account currency and calculate the position size from an explicit exit assumption. The margin calculation then shows how much collateral the position requires and how it interacts with positions already open.
For a dollar account buying 25,000 EUR/USD units, the pip value is $2.50. If the actual entry is 1.1000 and the actual exit is 1.1040, the trading profit is $100. With a hypothetical $7 commission per standard lot for the completed trade, commission is $1.75, leaving $98.25 before any financing or other separate charges. Spread and slippage already embedded in those fills are not deducted again.
After execution, compare the assumptions with the recorded prices and charges. A discrepancy may come from conversion, a commission quoted per side, a larger spread, an unexpected funding debit or an incorrect contract size. Resolving it improves the next calculation.
Across many trades, the record supports estimates of net expectancy, loss severity and drawdown. Those estimates should retain their connection to the period, account conditions and trading rules that produced them.
The mathematics can establish how a position responds to a price change and what assumptions a strategy needs to break even. Establishing whether those assumptions will hold requires evidence. Keeping that distinction visible is the basis of a disciplined mathematical approach to forex trading.