An assessment of contagion risks in the banking system using non-parametric and Copula approaches

Abstract

This study endeavours to shed some light on the Contagion risk in the Vietnamese banking system. In so doing, we analyse the contagion risk through stock returns on listed commercial banks by employing non-parametric and Copula approaches. A rich set of empirical approaches are employed, including non-parametric (Chi-plots, Kendall-plots) and parametric Copula estimations to define the dependence structure of pairs of daily returns, balanced by a total of 36 copulas with 17,456 observations over the period from July 2006 to September 2017. Our results show that the risk of each individual bank may transmit to other banks through stock returns, which are reflected in their price information. The results also suggest existence of contagion risk and strong dependency in the structure of stock returns of banks under analysis. As a consequence, to avoid negative returns for the portfolio, careful diversification is required while investing in the Vietnamese banking sector, when showing a Clayton relationship (left-tail dependency). Our findings have profound implications for investors, policymakers and authorities responsible for financial stability.