Tosh5457 said:Is correlation between 2 time-series a useful indicator? In currency pairs for example, sometimes the correlation between 2 pairs (e.g. EUR/USD and GBP/USD) for the past x days is strong, but it can weaken very fast. Should I just take the average over time and go with it, or is there a better indicator to measure a relation between 2 time-series?
A time series correlation is a statistical measure that indicates the extent to which two or more time series variables are related. It measures the strength and direction of the linear relationship between the variables.
Time series correlation is calculated using a mathematical formula called the correlation coefficient. This coefficient is calculated by dividing the covariance of the two time series by the product of their standard deviations.
A high correlation coefficient, closer to +1 or -1, indicates a strong positive or negative linear relationship between the time series variables. A low correlation coefficient, closer to 0, indicates a weak or no linear relationship.
Correlation is a useful tool for identifying patterns and relationships between time series variables, but it should not be solely relied upon for predicting future trends. Other factors such as external influences and underlying patterns should also be considered.
Yes, there are some limitations to using time series correlation as an indicator. It assumes a linear relationship between variables, and may not capture non-linear relationships. It also does not account for causation, and a high correlation does not necessarily mean causation between the variables.