Join our community of traders FOR FREE!

  • Learn
  • Improve yourself
  • Get Rewards
Learn More

Correlations Within the Forex Market

Correlations within the Forex market

You will learn about the following concepts

  • Why is correlation important in trading?
  • How to evaluate the correlation coefficient?
  • How can these relations be applied in trading?
  • and more

Within the Forex market there is a certain relation between currencies that can be observed. Since currencies trade in pairs, they cannot be isolated from each other. In order to measure the strength of relation between currency pairs, a trader can use different statistical coefficients, as one of the most popular is correlation.

Why is correlation important in trading?

For a trader, who uses a number of currency pairs on his/her account, it is extremely vital to know how these pairs relate to each other. If a trader focuses on one single pair, this might not be that important. If some currency pairs have a strong relation and move together, this means that a trader can be exposed to a significantly high risk, if he/she trades both pairs. Therefore, the correlation coefficient is a milestone in managing risk.

How to evaluate the correlation coefficient?

First, we need quotes for a specific period of time on two different currency pairs. We can evaluate the correlation between daily charts for EUR/USD and GBP/USD for the period since September 1st 2013, for example. Quotes should regard the same time frame – daily EUR/USD to daily GBP/USD, weekly to weekly, etc. Second, we will use the build-in function CORREL in Excel in order to estimate the correlation coefficient during the period September 1st till October 31st. We see that the correlation between these two pairs is 0.8911 or 89.11%. This is a significant correlation.

Sep 02, 20131.31901.5546
Sep 03, 20131.31721.5562
Sep 04, 20131.31991.5620
Sep 05, 20131.31191.5590
Sep 06, 20131.31811.5631
Sep 09, 20131.32561.5698
Sep 10, 20131.32681.5731
Sep 11, 20131.33141.5824
Sep 12, 20131.32981.5803
Sep 13, 20131.32971.5878
Sep 16, 20131.33361.5902
Sep 17, 20131.33561.5903
Sep 18, 20131.35231.6133
Sep 19, 20131.35341.6036
Sep 20, 20131.35231.6005
Sep 23, 20131.34961.6040
Sep 24, 20131.34741.5994
Sep 25, 20131.35181.6074
Sep 26, 20131.34871.6038
Sep 27, 20131.35221.6140
Sep 30, 20131.35231.6187
Oct 01, 20131.35231.6192
Oct 02, 20131.35901.6236
Oct 03, 20131.36231.6157
Oct 04, 20131.35541.6009
Oct 07, 20131.35811.6098
Oct 08, 20131.35991.6119
Oct 09, 20131.35191.5955
Oct 10, 20131.35261.5975
Oct 11, 20131.35401.5946
Oct 14, 20131.35561.5972
Oct 15, 20131.35211.5983
Oct 16, 20131.35221.5946
Oct 17, 20131.36681.6150
Oct 18, 20131.36851.6167
Oct 21, 20131.36731.6132
Oct 22, 20131.37781.6231
Oct 23, 20131.37781.6164
Oct 24, 20131.38021.6198
Oct 25, 20131.38051.6168
Oct 28, 20131.37881.6140
Oct 29, 20131.37421.6047
Oct 30, 20131.37231.6024
Oct 31, 20131.35791.6043
correlation coefficient0.8911621711

It is useful to mention that high positive values of correlation suggest that currency pairs move significantly in one and the same direction, while negative values close to -1 suggest that they move equally, but in opposite directions. If the correlation coefficient is close to 0, there is no relation between currency pairs and they move independently.

Also, correlation coefficients will change, if we use different data array. In our example we use 2 months of data (from September 1st until October 31st), but if we use a 3-month array of data, we would certainly evaluate a different correlation coefficient.

Correlation is usually presented in a correlation table or a matrix. First, let us see the correlation coefficients regarding one and the same time frame (2 months) and multiple currency pairs. We use daily quotes for all pairs.

Daily Time Frame for 2 months


When crossing every column and row, we can observe the correlation coefficient for the corresponding pairs. If we cross EUR/USD and GBP/USD, we can see a coefficient of 0.8911. Since EUR/USD is related with itself, the correlation coefficient is 1. Additionally, EUR/USD to GBP/USD presents the same relation as GBP/USD to EUR/USD, thus their correlation coefficient is the same.

Second, let us see the correlation coefficients regarding one currency pair, but different time frames.

EUR/USD Correlation Coefficients
1 week0.8163-0.9693-0.7767
1 month0.92-0.9882-0.7705
6 months0.5077-0.9585-0.1282

How can this be applied in trading?

correlation eur-usd - gbp-usd
Here we are looking at 1-hour charts of EUR/USD and GBP/USD. They move quite simultaneously. Let’s suppose the following situation. A trader decides to open one position with EUR/USD and another one with GBP/USD, considering that it will be a diversification and a risk reducing move. However, this appears to be wrong, because the high correlation coefficient between the pairs suggests that adding GBP/USD will not provide the trader with any risk reduction. Entering the above mentioned positions will surely increase traders risk, because the EUR and the GBP very often move in one and the same direction. Since we have a correlation coefficient of 0.89, we can imagine it as a replacement of 1 lot of GBP/USD with 0.89 lot of EUR/USD. Opening 1 lot in EUR/USD and 1 lot in GBP/USD will appear to be equal to 1.89 lot of EUR/USD. This means that the trader almost doubled his risk exposure.

If a trader uses highly negatively correlated pairs (EUR/USD and USD/CHF, again 1-hour charts as shown below) and take opposite positions with them, this will appear to be the same as taking one-direction positions with highly positively correlated currency pairs. Or, if he enters a long position with EUR/USD and simultaneously a short position with USD/CHF, this will give him the same situation as entering a long position in EUR/USD and another long position in GBP/USD, or entering a long position with twice the size of EUR/USD.

correlation eur-usd - usd-chf