The main objective of this paper is to investigate the interactions between the constructed Financial Condition Indices (FCI), assumed as a measure of financial stress; the World Pandemic Uncertainty Index (WPUI), as an external factor and the recession probability. The main research tools are the Principal Component Analysis and the random forest model machine-learning algorithm. In particular, we combine the FCI of 11 European economies - Bulgaria, Czech Republic, Croatia, Estonia, Hungary, Lithuania, Latvia, Poland, Romania, Germany and Turkey - with the WPUI, to explain the current global recession. We conclude that not only the traditional FCI, but in addition a non-economic factor, such as WPUI, are indispensable instruments for the prediction of turning points in the business cycle and recession probabilities, especially as concerns the 2020 lockdown. While this confirms the role of external shocks as an explanatory variable of economic dynamics, our interpretation is that these shocks affect the system not directly, but via monetary parameters, thus excluding money neutrality.
