A nonparametric model of financial system-economic growth nexus

In this paper we examine a familiar topic in financial economics: the financial system-economic growth nexus. However, we depart from the extant literature in the sense that we empirically show that proposed models and variables are nonparametric, implying that the use of estimation techniques that assume a linear data generating process are questionable. We, thus, use a nonparametric panel data model to estimate the financial-economic growth relationship. We find that the banking sector shows a greater case of a statistically significant positive effect on GDP: for example, domestic credit and private credit both have a statistically significant positive effect on GDP in six of the seven panels. On the other hand, the evidence from the stock market suggests relatively less cases of a statistically significant positive effect on GDP: for only four panels in the case of market capitalization and two panels in the case of stocks traded.