Identification of Relationships between Economic Policy Variables in Colombia through Crossed-Correlation Functions

Abstract

To study and analyze the relationship between economic and financial variables, diverse methodologies have been used. However, a robust tool is used to analyze these relationships, which is known as the crossed correlation proposed since 1970 by Box and Jenkins and later retaken in 1977 by Pierce and Haugh. This methodology consists in identifying and estimating the ARIMA model that best fits each series and obtaining the residues identified as white noise; these constitute the pre-whitened series. Finally, the crossed correlation is estimated among the pre-whitened series and the causality relationships are determined between each series pair. This research relates four fundamental economic variables with great impact on a company’s finances: General Index of the Colombian Stock Exchange, Emerging Markets Bonds Index, Representative Market Rate, and Rate of Inflation, finding evidence of instantaneous, unidirectional, and bidirectional relationships.