Portfolio Heat Calculator: Your True Combined Risk

This portfolio heat calculator shows the real combined risk of every open trade at once, after correlation. Add up to five positions, set the risk percent and the direction of each, and the tool returns two numbers. The first is the naive total you get by adding percents. Next comes the correlation-adjusted figure that treats overlapping currency legs as what they really are. The tool also flags your largest single-currency exposure and scores the book against the common 6% heat convention. Everything runs in your browser, and nothing you type is sent anywhere.

What portfolio heat actually measures

Portfolio heat is the total capital you have exposed across every open position at the same time. Most traders build it the simple way. Three trades at 1% risk each equals 3% heat. That figure answers one clean question: if every stop is hit, what does the account lose? It is a worst-case number, and it is worth knowing.

The problem is what that number implies. Adding percents suggests three separate bets. It reads like a diversified book. In a forex account it usually is not. Currency pairs share legs. EURUSD, GBPUSD and AUDUSD all quote against the dollar. Buy all three and you have placed one directional bet three times.

So heat has two faces. Naive heat tells you your maximum loss. Correlation-adjusted heat tells you how the book behaves day to day. This tool shows both, because you need both. Size your book with the first, and judge its structure with the second. Neither number predicts anything. Both are arithmetic on the inputs you supply.

Why adding risk percents understates your true risk

Adding percents understates risk in one specific way. It says nothing about whether your positions move together. Two trades at 1% each always read as 2% heat, whether they are the same trade twice or genuine opposites.

Consider a long EURUSD and a short USDCHF. Both are dollar shorts. The published long-run correlation between EURUSD and USDCHF is -0.90, and shorting the second flips that sign to +0.90. Thus the two trades behave almost like one position of double size. The correlation-adjusted heat comes out at 1.95%, barely below the naive 2.00%.

Now flip it. Long EURUSD and long USDCHF at 1% each. The same -0.90 correlation now works against you as a hedge. Adjusted heat drops to 0.45%. The naive figure is still 2.00% in both cases. The naive figure cannot tell those two books apart. That is the gap this calculator closes. Check the raw numbers on the forex correlation matrix.

How to use this portfolio heat calculator

1. List every open or planned position, one per row. Pick the pair from the dropdown, then set Long or Short. Direction matters here, because it flips the sign of every correlation for that row.
2. Enter the risk percent for each row. This is the loss on that single trade if its stop is hit, as a share of the account. If you sized the trade with the position size calculator, it is the same number you typed there.
3. Leave unused rows on "No position". Blank or zero risk rows are ignored, so a two-trade book works fine.
4. Add your account balance if you want the answer in money. Leave it blank for percentages only.
5. Read the two headline figures. The adjusted heat is the structural number. The naive heat is your absolute worst case.
6. Check the largest single-currency exposure line. If one currency dominates, the book is a bet on that currency, whatever the pair names say.
7. Act on the traffic light. Green means the adjusted heat sits below 3%. Amber means it sits between 3% and 6%. Red means it is above the common 6% convention.

A worked example: three trades that are really one bet

Take the default setup. You are long EURUSD, long GBPUSD and long AUDUSD, each risking 1% of a 10,000 USD account. Naive heat is 3.00%, or 300 USD if all three stops are hit. That is the honest worst case, and the tool reports it.

Now look at the structure. The published correlations are EURUSD to GBPUSD at +0.85, EURUSD to AUDUSD at +0.60, and GBPUSD to AUDUSD at +0.60. Feed those into the aggregation and the correlation-adjusted heat comes out at 2.66%. If the three trades were genuinely independent, the same three 1% risks would aggregate to just 1.73%. Your book is running 54% hotter than an independent book of the same size.

Read the currency line and it becomes obvious. Every one of those three positions is short the dollar. Net dollar exposure is 3.00% short. The euro, sterling and Aussie legs sit at 1.00% long each. You did not place three trades. You placed one 2.66% bet against the dollar and split the ticket three ways. That is a valid position to hold. Just hold it knowingly.

The math behind correlation-adjusted heat

The formula is standard portfolio volatility aggregation, applied to risk percents instead of position values. Write each position's risk percent as w, and its direction as s, where long is +1 and short is -1. Then:

adjusted heat = sqrt( sum over i, sum over j of r(i,j) x s(i) x s(j) x w(i) x w(j) )

The diagonal terms, where i equals j, use r = 1 and contribute w squared. Those alone give the independent baseline, sqrt(sum of w squared). The off-diagonal terms add or subtract depending on the correlation sign and the two directions. Positive pairs that you hold in the same direction add. Positive pairs held in opposite directions subtract.

Three checks confirm the formula behaves. With all correlations at zero, three 1% positions give sqrt(3), which is 1.732%. With all correlations at 1, the same three give sqrt(9), which is 3.000% and matches the naive sum exactly. Doubling every risk percent doubles the adjusted heat, which is what a risk measure should do. The quantity inside the square root can never go below zero for a valid correlation matrix, and the tool clamps it at zero anyway.

Reference table: pre-computed heat for common baskets

These figures come straight from the correlation values published on this site. Use them as a sanity check on your own inputs.

Positions (risk each)	Naive heat	Independent baseline	Adjusted heat	Largest currency leg
Long EURUSD, GBPUSD, AUDUSD (1%)	3.00%	1.73%	2.66%	3.00% short USD
Long EURUSD, GBPUSD, NZDUSD (1%)	3.00%	1.73%	2.63%	3.00% short USD
Long AUDUSD, NZDUSD, XAUUSD (1%)	3.00%	1.73%	2.61%	3.00% short USD
Long EURUSD, short USDCHF (1%)	2.00%	1.41%	1.95%	2.00% short USD
Long EURUSD, long USDCHF (1%)	2.00%	1.41%	0.45%	1.00% long EUR
Long EURUSD, long USDJPY (1%)	2.00%	1.41%	1.22%	1.00% long EUR
Long AUDUSD, long NZDUSD (1%)	2.00%	1.41%	1.92%	2.00% short USD
Long EURUSD, GBPUSD, AUDUSD, NZDUSD, XAUUSD (1%)	5.00%	2.24%	4.07%	5.00% short USD
Long EURUSD, GBPUSD, AUDUSD (2%)	6.00%	3.46%	5.33%	6.00% short USD

The last row is the one worth staring at. It passes the 6% naive test on the nose. Structurally it is a single 5.33% directional bet against the dollar, held in a 10,000 USD account as a 600 USD worst case.

The 6% convention, and why it is a convention

The 6% figure comes from popular risk-management writing, most often quoted as a monthly drawdown stop. The logic is simple. Risk 2% per trade, cap yourself at three open trades, stop trading for the month once you are 6% down. It is memorable and it keeps beginners alive.

It is not a law, and nothing derives it. A scalper closing positions inside an hour carries heat very differently from a swing trader holding through two weekends. An account with an 8% historical drawdown tolerance is a different animal from one that has never been 3% underwater. The right cap is the one that keeps you trading your plan after a bad week.

Adjusted heat	Traffic light	Reading
Below 3.00%	Green	Well inside the common convention
3.00% to 6.00%	Amber	Inside the convention, upper half
Above 6.00%	Red	Above the common convention

Set your own bands if you have the data to justify them. Model the effect of different caps with the drawdown calculator and the risk of ruin calculator before you raise the ceiling.

Why correlated stops all hit in the same hour

Correlation is an abstraction until the stops start filling. Then it gets very concrete. A dollar-positive surprise, say a hot CPI print, lifts the dollar against everything at once. If you are long EURUSD, GBPUSD and AUDUSD, all three positions turn red in the same minute.

Stops cluster in time as well as in direction. Because the three pairs move together, the price levels that trigger your exits arrive together too. You do not lose 1%, then 1% an hour later, then 1% tomorrow. You lose 3% inside one candle. Your equity curve takes the hit as a single step down.

This is why heat matters more than any single trade's risk. One 3% loss feels different from three 1% losses spread across a week, even though the arithmetic is identical. It hits harder psychologically, and it is more likely to break your process. Logging these clusters in a trade journal is the fastest way to see how often it happens to you.

Single-currency exposure: the simpler mental check

Correlation math is precise but abstract. Currency-leg counting is cruder and faster, and it catches most of the same problems. Every forex position is two bets: long one currency, short the other. Long EURUSD means long euro and short dollar. Short USDJPY means short dollar and long yen.

Count the legs across your book, keeping the signs. Three 1% dollar-short positions give a 3.00% net short dollar exposure. That single number tells you what you actually own. The tool reports the largest net leg for exactly this reason.

A useful habit is to cap the largest single-currency leg rather than the trade count. Many traders keep it at or below 3%. It stops the classic mistake of holding five "different" trades that are one dollar bet. The forex heatmap shows which currency is driving the session, and the interest rate tracker shows the policy backdrop pushing those legs around.

Reduce heat by sizing down, not by hedging

When the number comes back red, there are two ways to fix it. Only one of them is cheap. The reliable fix is to size down. Cut each position's risk percent, or close the weakest idea. The adjusted heat scales linearly with position size, so halving every risk percent halves the adjusted heat. Yet the correlation structure does not change at all.

The tempting fix is to open an offsetting position. Long EURUSD looks hot, so you add a long USDCHF to damp it. On paper the adjusted heat drops sharply. In practice you now pay two spreads, hold two sets of swap charges, and manage two tickets. You have also capped the upside of the original idea. The hedging calculator lets you price that trade-off before you place it.

There is a second problem with hedging by correlation. The correlations you are relying on are historical averages, and they shift. A hedge sized off a -0.90 reading stops working when the relationship drifts to -0.60. Sizing down carries no such model dependency. Check the margin cost of extra tickets with the margin calculator, and the per-pip cost with the pip value calculator.

What this calculator cannot tell you

Be clear about the limits. The correlation values here are long-run averages, published on this site and refreshed periodically. Real correlations move constantly. They also tend to converge toward 1 during a panic, which is exactly when you would want them low. A crisis book is hotter than this tool reports.

Second, the tool assumes each risk percent is real. It is not. Slippage, gaps and weekend openings can push an actual loss past the planned stop. Treat the naive heat as a floor on the worst case, not a ceiling.

Third, the adjusted figure is a volatility-style aggregation, not a loss forecast. It describes how the book tends to move together. It does not say what will happen next week, and it says nothing at all about your winning percentage or your edge. For that, see the expectancy calculator and the Kelly criterion calculator.

Fourth, the tool covers eight instruments. Exotic pairs, indices and crypto are excluded because reliable long-run correlations for them are harder to publish honestly. Finally, it ignores correlation between the pairs and any equity or commodity positions you hold elsewhere. Every tool on this site is built and tested under the Editorial and Testing Policy, and the full set sits on the forex tools page alongside the best MT4 indicators guide.

Want the full toolkit? Get the complete database.

FAQ

What is a safe level of portfolio heat?

There is no single answer, and anyone giving you one is guessing. The common retail convention caps total heat at 6%, built from three trades at 2% each. Many traders run far lower. Pick a cap you can sit through during a bad week, then hold to it.

Why is my adjusted heat lower than the naive total?

Because the adjusted figure is a quadratic aggregation, and it can only equal the naive sum when every correlation is exactly 1. Any imperfect correlation pulls it below the sum. The useful comparison is not adjusted against naive. It is adjusted against the independent baseline, which the tool also reports.

Does going short really flip the correlations?

Yes, and the math is exact. A short position is the negative of a long one, so its covariance with everything else changes sign. That is why the tool asks for a direction on every row. Long EURUSD with short USDCHF is a concentrated dollar bet; long EURUSD with long USDCHF is close to flat.

Should I count pending orders in my heat?

Count them if they can trigger while you are away from the screen. A resting buy-stop is heat the moment it fills, and it usually fills exactly when the correlated positions are moving. Traders who ignore pending orders are the ones who get surprised by a 5% day.

Can I use this for a funded-account challenge?

Yes, and it is a good fit, since most programs enforce a hard daily and total drawdown limit. Set your heat cap below the program's daily limit with room to spare for slippage. Remember that the adjusted figure is a model estimate built on historical correlations, not a prediction of any outcome. Results are not guaranteed; past performance is not indicative of future results.