We consider two discrete signals that are discrete functions over time. We introduce a method to approximate one signal in a linear form of the other. The coefficient of the linear expression are obtained by applying a least square method. We present an example by using daily data of exchange rates. If an exchange rate of a currency is in a linear form of the other and there is a time leg, then one may predict the dynamics of a currency based on the dynamics of the other. Investors are interested in predicting the dynamics of exchange rates, stocks and commodity prices, if such prediction exists.
Copyrights © 2014