The purpose of this paper is to give an overview of variability of empirical results of several financial contagion studies, taking into account the role of financial markets, data sets and the applied definitions and methods that may explain the variability of empirical evidence.
The authors used qualitative analysis of published research materials about previous financial crises and analyzed the variability of empirical results of around 75 studies of financial contagion, taking into account the particularities of financial markets, data sets and tests methods.
The results of the analysis show that empirical studies provide heterogeneous results depending on applied definitions and methods, as well as chosen crises, destination countries and financial indices. Summing up all the relevant empirical findings the results supporting the contagion hypothesis are in clear dominance, but taking into account differences in definitions and testing methodologies the research did not reveal clear results as to which evidence dominates or should dominate.
The authors conclude that solely qualitative analysis of published research materials about previous financial crises does not give sufficient information to elaborate proper management measures to prevent serious consequences of financial crises. The authors propose that it is possible to obtain a more adequate picture of financial contagion by using a meta‐analysis, which the authors are planning to do in future studies.
The paper provides information about some reasons that explain the variability of the results that are presented in the empirical studies about financial contagion. This information can be used for elaborating policy proposals and regulations that can help alleviate possible negative consequences of financial contagion. The paper shows the way for future articles summarising financial contagion.
The study sums up previous findings on the field of financial contagion and shows the insufficiency of the traditional literature review to accomplish that task.
