Masters Thesis

Financial development and output growth: a causal VAR empirical analysis

This study examines empirically the potential causal link between financial development and output growth in each of sixteen countries that have experienced sustained economic growth in the postwar period within a multivariate vector autoregressive (VAR) framework. I use three time series methodologies such as usual Granger causality approach, modified causality test of Toda· and Yamamoto (1995) and vanance decompositions to examine interrelationships between variables in the VAR system for the same data set. I find mixed results on the direction of causality and the results vary with respect to the type of test employed. Using annual time series data from 1960 to 2004 and a Granger causality approach, I find no evidence of causality in either direction in twelve countries, bi-directional causality in two countries, one-way causality from growth to finance in two countries and no one-way causality from finance to growth for any of the countries examined. Although the results are somewhat sensitive to the type of causality tests employed, the general conclusion is that financial development and output growth are causally independent at annual frequencies. This casts doubts on claims that financial development leads output growth. Therefore, more empirical studies and the need for broadening the econometric approach are called for before making any general conclusion about this relationship.