How to Calculate Aggregate Risk Across Multiple Positions: Correlation, Concentration and Margin

Answer

The answer: start with the sum, then decompose down to the factors

Multiple position risk always begins the same way: convert each position’s estimated stop-loss into your account currency, then simply add the amounts together. That figure is your gross open risk, the worst-case loss budget if every position simultaneously reaches its stop. But this simple sum is only a starting point. The real skew in a book appears only once you also separate the shared factors such as the same currency or the same index, the direction of each position, correlation, and simultaneous margin use.

The point most easily missed is that even when each trade is sized to risk 1% on its own, overlapping factors mean a single price move can stop out several positions at once. Three positions that should total 3% behave very differently if the book is skewed USD-long: one adverse USD move hits two of them together, so about 2% moves in a single shock. How you decide the trade size itself is covered in the FX & CFD Lot Size Calculation Guide; this article focuses on taking those already-sized positions and reviewing them across the whole account.

Every number and chart below is fictional educational data, not a real price, contract specification or track record. The point is not the size of the amounts themselves but the workflow of summing and then decomposing risk.

Terms and assumptions

Terms and assumptions: gross risk, shared factors, net / gross exposure, correlation

Before the aggregation, here are the words used in this guide, with their units. Definitions vary by broker and context, so treat these as generalized educational labels rather than universal rules.

• Gross open risk (JPY): each position’s estimated stop-loss converted into the account currency and summed without regard to direction. An upper-bound view of the worst-case loss budget.
• Shared factor: something that moves several positions together, such as USD strength or weakness, risk-on / risk-off sentiment, or the direction of a particular index.
• Net exposure: the residual skew in a currency or factor after offsetting long against short.
• Gross exposure: the total added up without offsetting. Reading both prevents the illusion that a book is hedged when it is not.
• Correlation: the degree to which two instruments move together, expressed from -1 to +1, but a variable value that shifts with period, timeframe and market regime.
• Conversion rate: the factor that restates a foreign-denominated loss into the account currency. Getting the direction wrong misstates the size of the loss.

The single-trade question of what percentage of the balance each position should risk is covered in the article on risk per trade. This guide picks up where that leaves off, looking at what happens when several of those 1% or 2% trades are held at the same time.

Summing

Sum each position’s loss in your account currency

The first step in aggregation is to bring every position’s estimated stop-loss into the same account currency (JPY here). The per-position estimated loss is found with the formula below.

What matters here is the direction of the conversion rate. A foreign-denominated loss is converted in the direction that restates it into the account currency. For example, if selling EUR/USD produces a loss of about 66.7 USD in the quote currency (USD) and the account is JPY-denominated, multiply by USD/JPY = 150.00 (1 USD = 150 JPY), so 66.7 USD × 150 ≈ 10,000 JPY. Reverse this direction and the loss comes out orders of magnitude wrong, so always compute “loss in the quote currency × the rate that restates that currency into JPY.” For pairs already quoted in the account currency (such as USD/JPY), the conversion rate is 1.

The formula for deriving each individual lot from per-unit value and stop distance is covered in the forex lot size formula. In this article’s aggregate tables, each position’s “per pip/point value (JPY, lot and conversion folded in)” is shown as a pre-folded figure, so multiplying by the stop distance alone yields the estimated loss.

Reading the skew

Currency and index concentration: read it as net / gross exposure

The total alone (gross open risk) does not reveal which factor a book is skewed toward. This is where you tally currency and index concentration with direction attached. Even across different symbols, breaking positions into their component currencies can reveal the same factor stacking up.

A classic example is holding long EUR/USD and long GBP/USD at the same time. The symbols differ, but both sit on the USD-short side, so the USD factor is doubled. If the USD rises (EUR and GBP fall), both move against you together, a more one-directional state than the simple sum suggests. Conversely, if one is long and the other short so the USD directions cancel, net exposure shrinks: gross (total) can be large while net (residual) stays light.

That is exactly why showing net (after offset) and gross (total) per currency helps you notice a book that is more concentrated than it looks, or, conversely, better hedged than you feared. When an equity index CFD is itself denominated in a specific currency and tied to a regional risk-on move, treat it as the same factor. The way contract size and point value work for CFDs is explained in the CFD position sizing guide. Note that this tally is a conceptual review to notice skew, not a trade recommendation.

Fictional case

A three-position fictional case: 1% each, 3% total, and the factors overlap

One fictional dataset is reused all the way through the tables, figures and worksheet that follow. The assumptions: an account balance of 1,000,000 JPY, with each position sized so its stop-loss risks 1.0% (10,000 JPY) of the balance. Everything below is a fictional educational example, not a real price or contract specification.

Gross open risk is 30,000 JPY (3.0% of balance). Each row’s “value (JPY)” already folds per-unit value, quantity and conversion into a single figure, even for the foreign-denominated EUR/USD and US30 (for example, the 10,000 JPY on short EUR/USD is a loss of about 66.7 USD in the quote currency USD, converted at USD/JPY = 150). Up to here it is a simple sum and does not yet reflect overlapping factors.

The problem starts here. Reviewing the three positions by shared factor, long USD/JPY and short EUR/USD are both USD-long, while long USD/JPY and long US30 are both Risk-on. Aggregating the losses per factor produces the concentration map below.

Gross open risk is 3.0%, but a single USD move against you moves 2.0% (the two USD-long positions) at once, and a single risk-off move moves 2.0% (the two Risk-on positions) at once. The feeling that “spreading across three trades makes each one light” differs sharply from the amount that actually moves on a single factor. This is why stacking up isolated single-position calculations never reveals the full multi-position picture.

Correlation

Correlation is not a fixed value: do not auto-discount because it looks low

One tool for seeing factor overlap numerically is correlation. But correlation is a variable that moves between -1 and +1 and shifts with period, timeframe and market regime. Even when it is low in calm conditions, many assets tend to move the same way during stress, and it is not unusual for correlation to rise precisely when risk is most concentrated. Below is an educational example of a correlation matrix for three instruments.

Beyond color (blue = positive, red = negative), the sign and value are always shown together. The caveat here is that the matrix describes the relationship between instruments; it only becomes overlapping risk once you multiply in the direction of the positions you actually hold. In this example, USD/JPY is long and EUR/USD is short, so the -0.60 negative correlation between them, once direction is considered, means “both work in the same USD-long direction.” Reading a correlation sign mechanically as a hedge leads to error.

Most important of all: do not automatically subtract losses from gross open risk because correlation looks low. Low correlation in the past does not guarantee it in the future, and if correlation jumps during stress you understate losses by exactly the amount you discounted. Use correlation as material to notice overlapping factors, and view the loss budget itself conservatively at the gross level. A way to test a strategy’s loss characteristics and how it behaves when conditions change is illustrated in the strategy risk and robustness workbench.

Stress

Keep the simultaneous-shock, gap and slippage stress row separate

It is important not to mix normal stop-loss losses with losses in fast markets. A stop order does not guarantee execution at the requested price, and gaps, fast markets, low liquidity and slippage can widen the loss beyond the intended stop distance. So the aggregate table keeps a normal-stop row and a stress row separate.

Normal gross open risk is 30,000 JPY (3.0%), but the stress row that assumes gaps and slippage swells to 50,000 JPY (5.0%). The larger a product’s stress-time gaps, as with index CFDs, the wider this difference grows. Stress loss is not a “guaranteed number” but a separate line under an assumption. Use normal stop-loss for designing the risk budget and stress loss for checking spare capacity, and do not mix the two. How much the margin side can withstand as floating loss grows is confirmed in the combined margin section next.

Combined margin

Check required margin and free margin on a combined basis

Separately from the loss budget, multiple positions also require you to view combined required margin. Add up the required margin for each open position and check the whole account’s spare capacity through total usage and free margin.

Total required margin is 400,000 JPY, so total usage against the 1,000,000 JPY balance is 40.0% and free margin is 600,000 JPY. What to watch is that a book skewed toward one currency or index moves together, swinging marked P/L negative all at once and eroding this free margin faster than expected. Floating loss is the unrealized P/L marked at the current price; it drives margin level and liquidation checks. It plays a different role from the estimated stop-loss (the risk budget), so review them separately.

The calculation of single-position required margin, usage and effective leverage itself is covered in detail in required margin, margin usage and effective leverage. The combined usage shown here is only a simple-sum estimate and does not replace a broker’s official margin-level calculation or liquidation threshold. Always confirm the definition of margin level, the marked basis and the liquidation level against the official specification of the account you use.

Verify the steps

Check it in the aggregate worksheet (educational mini-calculator)

Get a feel for the aggregation flow so far with the educational worksheet below. Enter the estimated loss, required margin and shared-factor tags for three positions, and it displays gross open risk, balance ratio, total required margin, margin usage, free margin, and simultaneous loss aggregated by factor. It applies no automatic correlation discount and makes no trade judgment. Even with JavaScript disabled, the static calculation table immediately after shows the same input example, formulas and answers.

Rule design

Rule design to reduce risk (not a trade recommendation)

When aggregate risk and factor skew become visible, how you contain them is not a trade judgment on individual symbols but rule design decided in advance. The following are examples of the approach, not recommendations to place any particular trade.

• Reduce quantity: once you notice factors overlapping, lower the quantity per position so that the simultaneous loss moving on a single factor stays within an acceptable range.
• Cap on concurrent holdings: set a rule for how many positions you hold at once, or a ceiling on gross open risk (for example, keeping the sum of balance ratios below a set level).
• Shared-factor cap: place a ceiling on the amount that can move at once on a single factor (USD-long, Risk-on, and so on). The simultaneous loss on one factor often bites more than the three-position total.
• Period risk budget: decide in advance the total aggregate risk you may take per period, such as a day or a week, and stop new entries once it is used up.

These do not guarantee a “safe lot” or a “risk amount you can always keep to”; they are simply a framework for yourself that reduces oversights. When averaging down or pyramiding into the same symbol and direction, the average entry price shifts and concentration in a shared factor intensifies further. How to find total risk and average price before an additional entry is covered in averaging down and pyramiding: weighted average price and total risk.

Avoid

Common mistakes and how to avoid them

Errors in multi-position risk calculation cluster into a few types. If any sound familiar, that item is the place to start reviewing.

• Feeling safe on the gross sum alone: grasping “3% across three trades” but missing the skew where 2% moves at once on a single factor. Always show simultaneous loss per factor alongside.
• Discounting because correlation is low: subtracting from gross on the basis of past low correlation understates risk when correlation rises during stress.
• Getting the conversion direction wrong: reversing the direction that restates a foreign-denominated loss into the account currency and misjudging the size of the loss.
• Ignoring net exposure: not noticing the same currency concentration across different symbols, so a supposed hedge is actually a doubled one-way bet.
• Mixing normal and stress loss: folding gap- and slippage-inclusive figures into the risk budget and distorting the calm-market design.
• Viewing margin only per position: not looking at combined usage and free margin, leaving room for margin level to deteriorate together as floating losses widen at once.

Practical checklist

Practical checklist for multiple positions

Run through the following items once before building multiple positions. It does not produce a pass/fail or a trade judgment; it is a procedure to reduce oversights in aggregation.

• Did you sum each position’s stop-loss into the same account currency, watching the conversion direction (gross open risk)?
• Is the balance ratio of gross open risk within the concurrent-holding cap you decided in advance?
• Did you break down component currencies and indices and tally simultaneous loss per shared factor (net / gross)?
• Did you keep correlation as reference only and avoid auto-discounting the sum because it looked low?
• Did you check normal stop-loss and stress loss separately?
• Did you sum required margin and review the whole account’s spare capacity through total usage and free margin?
• Did you confirm margin rate, minimum lot, margin level and liquidation threshold against the official specification of the broker you use?

The ladder

Free, Pro and Premium roles: viewed by problem

Of the aggregate risk, concentration, correlation and combined margin covered in this article, here is how far you can check for free and where the higher tiers begin, organized by problem. The design is progressive: free is enough for what a single-position calculation can solve, and you consider a higher tier at the point a problem appears that single-position work cannot solve. Feature names, scope and pricing can change, so treat the plan page as the single source of truth.

The aggregation, concentration, correlation and combined margin in this article are material for considering a higher tier at the point a problem single-position work cannot see appears. While confirming single-position required margin is enough, the free scope is sufficient.

Summary

Summary: multi-position aggregate risk and the next step

Multiple position risk always starts by summing each stop-loss in your account currency to get gross open risk. The key point is that the simple sum is only a starting point for a worst-case loss budget; the real skew appears only once you decompose it by shared currency and index factors, direction, correlation, simultaneous stops and combined margin use. In the fictional case, 1% each and 3% total still meant 2% moving at once when one factor reversed.

In practice, if you (1) sum each loss watching the conversion direction, (2) tally simultaneous loss per shared factor as net / gross, (3) keep correlation as reference and avoid auto-discounting, (4) separate normal and stress loss, and (5) sum required margin to review spare capacity, you can greatly reduce multi-position oversights.

Related

Related reading

Explore other English themes from the article library.