Educational content. This article defines market correlation and shows how the coefficient is calculated and how it relates to overlapping exposure. It does not constitute investment advice or a recommendation. CFD trading carries significant risk of loss and may not be suitable for all investors.
Market correlation measures how closely two prices move together, expressed as a coefficient between -1 and +1. A value of +1 means the returns always move in the same direction, 0 means no linear relationship, and -1 means they move in opposite directions. In the Vanto feed snapshot of 10 October 2026, 37 of 77 instruments involve USD, which is why open positions often overlap.
This article gives the formula, a worked calculation, a table showing how correlation changes the combined risk of two positions, and a count of shared currency exposure across the Vanto instrument list. The figures are arithmetic and feed facts, not forecasts.
What Is the Correlation Coefficient?
The correlation coefficient is a number between -1 and +1 that summarises how two series of returns move relative to each other. It is calculated from returns (percentage changes over a period), not from price levels.
| Coefficient | Meaning | Example of the pattern |
|---|---|---|
| +1.0 | Perfect positive: always the same direction | A price and a copy of itself |
| +0.5 to +0.9 | Positive: usually the same direction | Two instruments with a shared driver |
| 0 | No linear relationship | Two unrelated drivers |
| -0.5 to -0.9 | Negative: usually opposite directions | An instrument and its inverse |
| -1.0 | Perfect negative: always opposite | A long and a short on the same symbol |
The coefficient says nothing about size. Two instruments can have a correlation of +0.9 while one moves three times as much as the other. Size is the job of volatility; direction is the job of correlation.
It also says nothing about cause. A high coefficient can come from one instrument driving the other, from a third factor driving both, or from coincidence over a short sample.
How Is Correlation Calculated?
The coefficient is the covariance of the two return series divided by the product of their standard deviations. In practice, with returns A and B over n periods:
r = sum of (A - mean A) x (B - mean B) / square root of [sum of (A - mean A)^2 x sum of (B - mean B)^2]
A worked example with five illustrative daily returns, in percent. These are invented numbers to show the arithmetic, not market data:
| Day | Return A | Return B | A minus mean (0.12) | B minus mean (0.04) | Product | A squared | B squared |
|---|---|---|---|---|---|---|---|
| 1 | 0.4 | 0.3 | 0.28 | 0.26 | 0.0728 | 0.0784 | 0.0676 |
| 2 | -0.2 | -0.1 | -0.32 | -0.14 | 0.0448 | 0.1024 | 0.0196 |
| 3 | 0.6 | 0.5 | 0.48 | 0.46 | 0.2208 | 0.2304 | 0.2116 |
| 4 | -0.5 | -0.6 | -0.62 | -0.64 | 0.3968 | 0.3844 | 0.4096 |
| 5 | 0.3 | 0.1 | 0.18 | 0.06 | 0.0108 | 0.0324 | 0.0036 |
| Sum | 0.7460 | 0.8280 | 0.7120 |
r = 0.7460 / square root of (0.8280 x 0.7120) = 0.7460 / 0.7678 = 0.97.
The result of 0.97 looks nearly perfect, but it rests on five observations. A sample this small can show a strong coefficient by chance, so the window length matters as much as the number itself.
Why Does the Window Length Change the Answer?
The same two instruments can show a strong correlation over one window and almost none over another, because the coefficient only describes the period it was measured on. Charting platforms usually offer rolling windows, such as 20, 60 or 250 daily returns, and each gives a different reading.
A short window reacts quickly to a new regime but is noisy. A long window is stable but slow to notice that a relationship has changed. Neither is the true value, since the coefficient is an estimate of a relationship that itself moves over time.
The stock-bond relationship is a documented example of that instability: the coefficient has changed sign across decades, as covered in why stocks fall when bond yields rise. A relationship described as strong in a textbook is a statement about the past, not a rule the market has to obey.
The Vanto feed used for this article is a snapshot of current quotes and instrument specifications. It holds no price history, so this article does not publish measured coefficients between instruments. Use the correlation tools in a charting platform, or the downloaded history in MT5, for a measured value over a window you choose.
How Many Vanto Instruments Share a Dollar Exposure?
A large share of the instrument list contains USD as the base or the profit currency, so positions that look unrelated often carry the same dollar exposure. The count below comes from the Vanto feed snapshot, 10 October 2026.
| Class | Instruments | With USD as base or profit currency |
|---|---|---|
| Forex | 42 | 14 |
| Indices | 18 | 6 |
| Metals | 2 | 2 |
| Energies | 3 | 3 |
| Crypto | 12 | 12 |
| Total | 77 | 37 |
All 12 crypto symbols in the snapshot, both metals and all three energy symbols are priced in USD. Six of the 18 indices are USD-quoted (CN50U, US100, US2000, US30, US500 and VIX). Being quoted in USD does not make two instruments correlated in the statistical sense, but it does mean their profit and loss are both converted from the same currency, and that a change in the dollar can affect them together.
Which Currencies Appear in the Most Forex Pairs?
The more pairs a currency appears in, the more likely two forex positions share a leg. The count below is from the same snapshot, for the 42 forex symbols.
| Currency | Forex pairs containing it |
|---|---|
| USD | 14 |
| EUR | 11 |
| GBP | 9 |
| AUD | 8 |
| CAD, CHF, JPY, NZD | 7 each |
| SGD | 4 |
USD is in 14 of 42 pairs, one third of the forex list. Taken as a ratio, a trader who opens three random pairs from the list has a meaningful chance that at least two of them contain the same currency.
What Is a Structural Overlap and Why Does It Matter?
A structural overlap is a shared currency or shared constituent between two instruments, and it exists regardless of what a correlation estimate shows. It is different from a statistical correlation, which can weaken.
Three examples:
- Shared quote currency. A long on EURUSD and a long on GBPUSD are both short USD. If the dollar rises broadly, both lose.
- Cross pair decomposition. EURJPY is, in price terms, the product of EURUSD and USDJPY. A long on EURJPY behaves like a long on EURUSD plus a long on USDJPY. Holding all three counts the same exposure more than once. See forex cross pairs explained.
- Shared constituents. US500 and US100 hold many of the same large companies, so they overlap by construction. The Nasdaq 100 and S&P 500 guides cover the overlap.
A structural overlap is also not the same thing as causation. The link between gold and the dollar is a documented mechanism, covered in why gold rises when the dollar index falls, but even that relationship changes in strength over time.
How Does Correlation Change the Combined Risk of Two Positions?
The risk of two equal positions grows with the correlation between them. For two positions of equal size and equal volatility, the combined standard deviation is the single-position risk multiplied by the square root of 2 x (1 + r).
| Correlation r | Multiplier | Combined risk of 2 equal positions |
|---|---|---|
| +1.0 | 2.00 | Twice one position |
| +0.5 | 1.73 | About 1.7 times one position |
| 0 | 1.41 | About 1.4 times one position |
| -0.5 | 1.00 | Same as one position |
| -1.0 | 0.00 | Cancels out |
A worked example on forex with pip values from the feed. EURUSD and GBPUSD both have a contract size of 100,000, so one pip on 1 lot is worth USD 10 (100,000 x 0.0001). With 1 lot of each, a 50-pip adverse move in both pairs loses 50 x USD 10 x 2 = USD 1,000. If only EURUSD had been open, the loss would be USD 500. When two positions carry the same direction on the same currency, the realistic worst case is closer to the doubled figure than to the diversified one. How size is chosen is covered in what is position sizing, and pip values per pair are in pip value for every forex pair at Vanto.
Leverage amplifies this in both directions. Overlapping positions use margin separately, so the account carries the combined exposure on one balance. The mechanics are in what is margin in trading.
How Can a Trader Use Correlation to Check Open Exposure?
A trader can use correlation as a checklist for overlap, not as a signal. The steps below are a way to count exposure, not a recommendation to open or close anything.
- List each open position by currency. Write each position as the currencies it is long and short. A long EURUSD is long EUR and short USD.
- Net the same currency. Add the exposure for each currency across all positions. Three positions that are each short USD are one large short USD exposure.
- Check class overlap. Indices that share constituents and crypto symbols quoted in USD can move together in a broad risk sell-off, as described in what is risk-on and risk-off.
- Compare against a measured coefficient. Use a rolling window in a charting tool and note the window you used.
Because Vanto accounts use hedging mode in MT5, a long and a short on the same symbol are both held open as separate positions. The difference between the two modes is in hedging vs netting in MT5.
When Does Correlation Break Down?
Correlation breaks down when the driver that linked two instruments changes, and it often does so at the moments when the link matters most. The common cases:
- A new driver takes over. A currency pair that followed rate differentials can switch to following risk sentiment, as with the dollar in why the US dollar rises when stocks fall.
- Correlations rise in stress. Instruments that normally move independently can fall together in a broad sell-off, so the diversification measured in calm periods may not be there in a crisis.
- A small sample misleads. The 0.97 in the worked example above comes from five returns. Treat any coefficient from a short window as provisional.
- Time zones and sessions differ. Two markets that are open at different times can show a low coefficient on daily data simply because one closes before the other reacts.
Common Mistakes
The most common errors treat the coefficient as a fixed fact about two instruments rather than an estimate over a window.
- Using price levels instead of returns. Two trending prices can show a high coefficient on levels while their returns are unrelated.
- Reading correlation as causation. A high coefficient says the two moved together, not that one drives the other.
- Assuming a negative correlation is a hedge. A coefficient of -0.5 reduces combined risk in the table above to the same as one position, not to zero. Only -1 cancels out, and it is not stable.
- Counting positions instead of exposures. Four positions in four symbols can be one exposure in one currency.
- Ignoring size. Correlation does not tell how much each position moves; two positions can be strongly correlated and still have very different pip or point values.
Frequently Asked Questions
What is a good correlation coefficient for diversification?
A coefficient near 0 or negative reduces combined risk the most, but no value is reliably "good", because the number changes over time. As the table shows, two equal positions at 0 carry about 1.41 times the risk of one, and at +1 they carry twice.
Is a correlation of 0.7 strong?
A coefficient of 0.7 is usually described as strong, because it means a substantial share of the movements go in the same direction. Squaring it gives about 0.49, which is the share of one series' variance explained by the other in a simple linear model.
Does correlation mean one market causes the other?
No. Correlation only describes co-movement. Two markets can move together because one drives the other, because a third factor drives both, or by chance over a short window.
How many days of data do I need to calculate correlation?
There is no fixed answer, and the result changes with the window. Short windows such as 20 days react quickly but are noisy, while longer windows such as 250 days are stable but slow. Five observations, as in the worked example, are too few to rely on.
Why do EURUSD and GBPUSD often move together?
They share USD as the quote currency, so a broad move in the dollar affects both, and the euro and the pound are also both exposed to European and global risk conditions. The overlap is structural for the USD leg and statistical for the rest, and the statistical part can weaken.
Can I hold a long and a short on correlated pairs?
Yes, the platform allows it, since Vanto accounts are in hedging mode. The two positions still use margin and pay costs separately, and the net result depends on how each moves, which correlation does not fully determine.
Calculate the Numbers Before You Trade
Use the Vanto trading calculator to check pip value, contract size and margin for each symbol in your list before comparing exposures. Running the figures for two overlapping positions side by side shows how much a shared currency adds to the combined position.
Risk warning. Trading securities, futures, options, and contracts for differences are complex financial instruments that require knowledge and understanding. Prices can fluctuate significantly and securities may become valueless. Investors may incur losses exceeding the potential for profits. Trading on margin can result in losses greater than the amount initially deposited. Past performance is not necessarily a guide to future performance. The information in this article is for educational purposes only and does not constitute investment advice, a recommendation, or an offer to buy or sell any financial instrument. Consider whether CFD trading is appropriate for your circumstances and seek independent advice if necessary.