Financial sector and output dynamics in the euro area: Non-linearities reconsidered

We analyze the feedback mechanisms between economic downturns and financial stress for several euro area countries. Our study employs newly constructed financial condition indices that incorporate banking variables extensively. We apply a non-linear Logistic Vector Smooth Transition Autoregressive (LVSTAR) model for investigating instabilities in the link between the financial sector and economic activity. The LVSTAR model allows for non-linear dynamics and regime changes between low and high stress regimes. It can also replicate the regime-specific amplification effects shown by our theoretical model. The amplification effects, however, change over time. Specifically after the Lehman collapse, we observe the presence of strong non-linearities and amplification mechanisms for some euro area countries. Thus, these strong amplification effects appear to be related to rare but large events, and to a low-frequency financial cycle. Prior to the financial crisis outbreak we find corridor stability even if the financial sector shock takes place in a high stress regime. More important seems to be the shock propagation over time in the economy. Only with the occurrence of rare but large events we find strong endogenous feedback loops and a loss of stability as described by the high stress regime of our theoretical model. The economy leaves the corridor of stability and is prone to adverse feedback loops.