In this post we will try to briefly explain the term correlation and how it can be used in Forex trading.
Generally correlation refers to a statistical relationship between two variables. In Forex it is a measure of how two currency pairs move in relation to each other. Correlation is useful because it can indicate a predictive relationship that can be exploited. Correlation is computed into the so called correlation coefficient which takes values between -1 and +1.
A correlation coeffiecient of +1 (perfect correlation) means that as one of the pairs moves in one direction, the second moves in the same direction as well. An example of positively correlated pairs is EURUSD and GBPUSD. Alternatively, a perfect negative correlation (coefficient of -1) means that when one of the pairs moves up, the other one moves down. A good example of negatively correlated pairs are EURUSD and USDCHF. A correlation coefficient of 0 means that there is no relation between the two variables. When talking about correlation coefficient, usually the Pearson correlation coefficient is assumed, which is sensitive only to a linear relationship between two variables or pairs in our case. If we have a series of n measurements of X and Y (X being the close prices of the first pair and Y being the close prices of the second pair for example), written as xi and yi and x and y are the sample means of X and Y, then the Pearson correlation coefficient r can be computed using the formula:
How to interpret the value of the correlation coefficient?
In practice we usually don’t see a perfect correlation. It should be noted also that even if we get a result different from zero, it doesn’t necessarily mean that a relationship between the two pairs is reliable. In order to interpret the result more objectively, we should take into account the sample size n (or the period). It is considered that there is a relationship between the pairs if the following rule of thumb is satisfied:
where |r| is the absolute value of the correlation coefficient. In the download section you can find a free MT4 indicator computing the Pearson correlation and showing the areas of insignificant correlation in other color.
The most basic trading strategy is to watch highly correlated pairs and wait for periods when the correlation between the pairs drops. Usually these periods don’t happen quite often and for a long time. That means that after some time the correlation between the two pairs will increase again to its usual levels. It is these periods of non-correlation that can be exploited. It should be determined which of the two pairs is overvalued/undervalued in order to determine the direction of the trades to be taken. Trades are opened on both pairs but opposite direction (if usual correlation is positive) or same direction (if usual correlation is negative).
Correlation is usually applied in pairs trading or statistical arbitrage strategies. These topics will be discussed in more detial in the next posts.
References:
Brandywine