Risk-to-Buffer: Setting Cyclical and Structural Capital Buffers through Banks Stress Tests (with C. Couaillier)
In this work we present the Risk-to-Buffer: a new framework to jointly calibrate cyclical and structural capital buffers, based on the integration of a non-linear macroeconomic model with a Stress test model. The macroeconomic model generates scenarios whose severity depends on the level of cyclical risk. Risk-related scenarios feed into a banks' Stress test model. Banks' capital losses deriving from the reference-risk scenario are used to calibrate the structural buffer. Additional losses associated to the current-risk scenario are used to calibrate the cyclical buffer.
How does Financial Vulnerability amplify Housing and Credit Shocks? (with C. Couaillier)
In this paper, we study how households’ financial vulnerability affects the propagation of housing and credit shocks. We estimate a non-linear model generating impulse responses dependent on the evolution of households' debt burden. We use sign restrictions to identify a large set of financial and economic shocks. We find that a high debt burden: i) amplifies the response of the economy to housing shock, ii) makes the response to expansionary credit shocks less persistent and even negative after one year. Finally, recessionary shocks have larger effects with respect to expansionary ones of the same size.
Estimating Non-Linear DSGEs with the Approximate Bayesian Computation: an application to the Zero Lower Bound.
Estimation of non-linear DSGE models is still very limited due to high computational costs and identification issues arising from the non-linear solution of the models. Besides, the use of small sample amplifies those issues. This paper advocates for the use of Approximate Bayesian Computation (ABC), a set of Bayesian techniques based on moments matching. First, through Monte Carlo exercises, I assess the small sample performance of ABC estimators and run a comparison with the Limited Information Method (Kim, 2002), the state-of-the-art Bayesian method of moments used in DSGE literature. I find that ABC has a better small sample performance, due to the more efficient way through which the information provided by the moments is used to update the prior distribution. Second, ABC is tested on the estimation of a new-Keynesian model with a zero lower bound, a real life application where the occasionally binding constraint complicates the use of traditional method of moments.
Houses, Debt and Growth.
In this paper I study how debt and housing affect the trend of a closed economy. I build and estimate a DSGE model with endogenous growth and with heterogeneous agents, savers and borrowers. I find that in response to an increase in the nominal collateral value, savers decrease their investments in technology, thus reducing the productivity growth and the trend of the economy. The productivity growth is found to steadily increase during the Great Moderation, mainly explained by positive TFP shocks. In pre-crisis period, the housing boom and the debt overhang crowd out investments intechnology, lowering the trend. In addition, the slow recovery which followed the Great Recession is mostly explained by a strong negative investment shock. Counterfactual exercises show that the presence of debt in the economy creates a propagation mechanism affecting the evolution of the trend and depending on the variation of the nominal collateral value. In particular, the effect of the investment shock is postponed with respect to the counter-factual scenario in which the net debt in the economy is zero.