Financial contagion represents a very controversial concept in international finance being one of the most frequently referenced subjects and yet least understood. The literature highlights that shocks and crises can spillover from a country to others through various channels. Although it is hard to determine exactly the cause and channel that lead to the transmission of the initial shock, it is more than clear that these events are encouraged when economies are integrated or in the process of global integration. The core of this study is to capture evidence of financial contagion based on a sample of daily closing prices from 17 different market indices, for the period January 1st, 2007 – October 15th, 2021. We employed the shortfall method for estimating the risk and we built a model within the framework of Bayesian Neural Networks (BNN). Given the results, evidence of contagion was sensed between some of the 17 markets within the sample, however the causality between them differed from the full sample period to the sub-samples periods. Nevertheless, we found that for the sub-sample corresponding to January 1st, 2020 – October 15th, 2021, period that encompasses the global exogenous shock triggered by the COVID-19 pandemic that appeared in Romania at the beginning of 2020, the relationships between global markets were decoupled, contagion being sensed mostly regionally, at the level of the European countries.