Home

Correlation Analysis is a way of determining any relationship between two seperate entities (e.g a price and a particular indicator. The result of this analysis shows if changes in the first entity coincide with changes in the second entity. The first item is known as the 'independent' entity, and the trick is to see whether changes in it affect the 'dependent' (second) entity (usually price). The results are usually called a 'correlation coefficient' and vary between plus one and minus one. A coefficient of +1.0, is a 'perfect positive correlation' - changes in the independent entity result in an exact change in the dependent entity. A coefficient of -1.0, is a "perfect negative correlation," and mean that changes in the independent entity cause an identical change in the dependent entity, but in the opposite direction. A coefficient of zero implies no relationship between the two entities. Low correlation coefficients (less than 0.1) suggests the relationship is weak or non-existent. High correlation coefficients (over 0.9) indicates that the dependent entity is likely to react predicably to changes in the independent entity. In day trading schools, this technique is often used to try and limit the risk by making simultaneous 'bets' on two strongly correllated entities that have an inverse relationship. A common use for this technique is to coompare 2 securites or indices (e.g. the DOW and the FTSE) to see if any useful reliable relationship exists that can be exploited. This forms the basis of 'pairs trading'.