• Learn
• Improve yourself
• Get Rewards

# 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.

 Date EUR/USD GBP/USD Sep 02, 2013 1.3190 1.5546 Sep 03, 2013 1.3172 1.5562 Sep 04, 2013 1.3199 1.5620 Sep 05, 2013 1.3119 1.5590 Sep 06, 2013 1.3181 1.5631 Sep 09, 2013 1.3256 1.5698 Sep 10, 2013 1.3268 1.5731 Sep 11, 2013 1.3314 1.5824 Sep 12, 2013 1.3298 1.5803 Sep 13, 2013 1.3297 1.5878 Sep 16, 2013 1.3336 1.5902 Sep 17, 2013 1.3356 1.5903 Sep 18, 2013 1.3523 1.6133 Sep 19, 2013 1.3534 1.6036 Sep 20, 2013 1.3523 1.6005 Sep 23, 2013 1.3496 1.6040 Sep 24, 2013 1.3474 1.5994 Sep 25, 2013 1.3518 1.6074 Sep 26, 2013 1.3487 1.6038 Sep 27, 2013 1.3522 1.6140 Sep 30, 2013 1.3523 1.6187 Oct 01, 2013 1.3523 1.6192 Oct 02, 2013 1.3590 1.6236 Oct 03, 2013 1.3623 1.6157 Oct 04, 2013 1.3554 1.6009 Oct 07, 2013 1.3581 1.6098 Oct 08, 2013 1.3599 1.6119 Oct 09, 2013 1.3519 1.5955 Oct 10, 2013 1.3526 1.5975 Oct 11, 2013 1.3540 1.5946 Oct 14, 2013 1.3556 1.5972 Oct 15, 2013 1.3521 1.5983 Oct 16, 2013 1.3522 1.5946 Oct 17, 2013 1.3668 1.6150 Oct 18, 2013 1.3685 1.6167 Oct 21, 2013 1.3673 1.6132 Oct 22, 2013 1.3778 1.6231 Oct 23, 2013 1.3778 1.6164 Oct 24, 2013 1.3802 1.6198 Oct 25, 2013 1.3805 1.6168 Oct 28, 2013 1.3788 1.6140 Oct 29, 2013 1.3742 1.6047 Oct 30, 2013 1.3723 1.6024 Oct 31, 2013 1.3579 1.6043 correlation coefficient 0.8911621711

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

EUR/USDGBP/USDUSD/CHFUSD/JPY
EUR/USD10.8911-0.9785-0.7899
GBP/USD0.89111-0.93-0.7889
USD/CHF-0.9785-0.9310.85
USD/JPY-0.7899-0.78890.851

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 EUR/USD GBP/USD USD/CHF USD/JPY 1 week 0.8163 -0.9693 -0.7767 1 month 0.92 -0.9882 -0.7705 6 months 0.5077 -0.9585 -0.1282

## How can this be applied in trading?

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.