The causal relationship between financial development and investment in Botswana

Abstract

In this paper, we examine the causal relationship between financial development and investment in Botswana between 1976 and 2014, using the autoregressive distributed-lag (ARDL) bounds testing approach. Unlike some previous studies, our study divides financial sector development into two segments, namely bank-based and market-based financial development. We also employ a trivariate Granger-causality model in order to address the omission-of-variable bias associated with a bivariate causality model. In order to capture the breadth and depth of the financial sector in the study country, we employ both bank- and market-based financial development indices. These indices are constructed from an array of bank- and market-based financial development indicators. Our results show that there is a bidirectional Granger-causal relationship between both bank-based and market-based financial development and investment in the short run. However, in the long run, a distinct causal flow is found to prevail only from investment to bank-based financial development.