Understanding how trading instruments move and behave relative to one another is a simple way to start identifying relationships and patterns in price action that can be taken advantage of in the pursuit of higher returns. The relationship between two assets is mathematically termed a “correlation.” These correlations can be applied to a number of different trading strategies depending on the nature of the relationship and how the instruments have behaved historically. While past behavior is not necessarily indicative of future price action, it is a great benchmark to work off of when planning and executing trades.

##
**What is a Correlation?**

A correlation is most commonly defined as a relationship or association between two or more variables. More specifically, it is a statistical measure of the relationship between the changes that are associated with the variables. With respect to trading, correlations tell us the degree to which movements in two instruments are related. There are two main types of correlations that are applied to financial trading, namely the direct correlation and inverse correlation.

A direct correlation would imply that a rise in one asset would be mirrored by a rise in another asset. Conversely, an inverse correlation would suggest that a movement higher in one instrument would be accompanied by a movement lower in another asset. How they are related depends not just on whether or not there exists a direct or indirect correlation, but the strength of the correlation itself.

##
**How Strong is the Relationship**

One of the beauties of calculating correlations between instruments is the ability to determine just how related two assets are. Direct correlations range in strength from 0.0 to 1.0. The closer the correlation trends towards 1.0, the stronger the relationship in directional movements between instruments. Conversely, the closer the correlation is to 0.0, the looser the relationship between the two instruments.

At the 0.0 level, there is exists no measurable relationship between the two assets being compared. Inverse correlations range between 0.0 and -1.0. The strongest inverse correlation would be nearest to the -1.0 level whereas a reading closer to 0.0 implies a looser correlation and weak relationship between the two measured assets.

##
**Applying a Correlation to Trading**

When using a correlation to a make a trading decision, the first step is to determine what two trading assets are going to be compared. Once decided, the correlation coefficient is applied and resembles an oscillator, similar to the relative strength index or stochastic oscillator, trending above and below the zero (0.0) line. The sample below gives a strong view as to how the indictor window appears when applied.

In the example above, the correlation is measuring the relationship between the gold (XAUUSD) and the US dollar index (DXY). One way to show the relationship is to overlay the two price charts (below) of the instruments to determine the returns of each asset over the time period measured. However, the correlation coefficient is a more mathematical representation of this relationship. What the correlation implies is that an inverse relationship exists between the gold and the US dollar. As the US dollar rises, gold falls. Conversely, as the US dollar falls, gold rises. However, as the correlation window shows, the relationship does not stay static but changes over time.

##
**Pair Trading**

One area in particular that correlations are coveted for is pair trading. Typically this involves trading two similar or related assets that have diverged from their historical correlation in a bet on their future convergence and reversion to more historical price trends and relationships. In the case of the earlier example with gold and the US dollar, although not a perfect inverse relationship, from the correlation we can infer that if there is strong momentum higher in the US dollar due to fundamental news or events, gold prices are likely to react negatively. However, if the inverse correlation loosens, gold prices might not be as sensitive to momentum in the US dollar.