Financial development and economic activity in advanced and developing economies: evidence from panel cointegration

Abstract

The purpose of this thesis is to re-evaluate the causality issue and the long-run relationship between financial development and economic activity. Using a sample of 75 countries -21 developed and 54 developing- over the period 1975-2015 in annual annual data three panels were constructed. Unit root tests (IPS, LLC, ADF-PP, Breitung) were conducted in order to ensure stationarity. Tests (Pedroni, Johansen and Kao) were, also, employed to seek for cointegrating vectors. The long run relationship was estimated using fully modified OLS and causality using vector error correction model. In summary, the results of the panel unit root tests indicate that the variables are stationary in their first differences; the cointegration tests show that regardless the proxies for the financial development there is at least one cointegrating vector for all groups. The findings for causality (VECM) detect short-run bidirectional causality for the world group, unidirectional short-run causality from economic activity to financial development for the developing group and unidirectional long-run causality from econonomic activity to finance, and for the developed group bidirectional short-run causality (for two out of three proxies of finance) and unidirectional long-run causality from finance to econonomic activity. Finally, the estimation for the long run relationship indicates a positive relation between economic activity and all three financial development proxies for all groups.