Co-integration analysis of economic variables

Abstract

The motivation for the current undertaking is to learn and exercise the multivariate co-integration analysis techniques, to explore and identify useful relationship(s) between India’s GDP and other representative indicators as proxies for various segments of the Indian economy, and thus use the identified relationship(s) to predict the country’s GDP.It is known that certain macroscopic indicators, such as GDP, are very difficult to measure accurately due to laborious and extensive data collection process. However, certain datasets can be and are accurately and timely determined by the government or other relevant agencies. The intention is to choose such datasets judiciously to include the breadth of sectors of the economy whilst calculating the national output. For instance, national energy consumption data to chart the country’s annual industrial sector output (or growth).Our analysis suggests that it is indeed possible to identify relevant indicators covering the three slices of the Indian economy –agriculture, industry & services. Further, it is noted that a co-integral relationship does exist between the national output and the lagged indicators identified. Results predicting the GDP using the said relationship have been presented and discussed.