abstract:  "Striking a balance between model complexity and parsimony is of crucial importance for producing accurate forecasts and structural inference. In this paper, we propose a non-parametric VAR model which relies on splines to approximate the unknown function between the response vector and its lags. This spline-based VAR is highly flexible and potentially overfits the data. To circumvent issues related to overfitting we introduce a prior which shrinks the non-parametric model towards a simpler model."

This talk will describe the economic situation in Ukraine and give ananalysis of the current macroeconomic indicators and also a perspective onfuture developments

The seminar will be held both in presence and online at this LINK

Guest Lecture organizzata nell'ambito del corso 27352 di Diritto Pubblico e Amministrativo - Prof.ssa Monica Rosini 

We empirically examine the Capital Purchase Program (CPP) used by the US government to bailout distressed banks. We find strong evidence that a feature of the CPP – the government’s ability toappoint independent directors on the board of an assisted bank that missed six dividend payments tothe Treasury – had a significant effect on bank behavior. Banks were averse to these appointments – theempirical distribution of missed payments exhibits a sharp discontinuity at five. Director appointmentsby the Treasury led to improved bank performance and lower CEO pay. Political connections did notaffect appointment decisions.

The seminar will be held both in presence and online @ this link

The seminar takes place in presence and online via teams 

We showthat macroprudential policies dampen the impact of global financial conditionson local bank credit cycles. For identification, we exploit exogenous variationin the U.S. VIX and household and business credit registers in a small openeconomy, where banks depend on foreign funding and macroprudential measuresvary over a full boom-bust cycle. When the VIX is low, tighter macroprudentialpolicies (i) reduce household lending, notably for riskier (FX and high DSTI)loans and by banks dependent on foreign funding and (ii) increase localcurrency lending to real-estate firms, while leaving business lending to otherfirms unchanged. Furthermore, such periods are associated with (iii) less totallending (to households and firms) and with a lower share of FX loans at thelocal level, suggesting a compositional shift toward (less risky) localcurrency loans. As a result, when the VIX is low, tighter macroprudentialpolicies (iv) dampen house prices and economic activity in areas with higherFX-loans.

The seminar will be held both in presence and online at this LINK

The way filings of mergers and acquisitions (M&A) with the Security Exchange Commission (SEC) are written might depend on investor reactions to initial public announcements of the deal. The current research aims to test this investor feedback hypothesis in an empirical setting that is free of reverse causality. We therefore manually collect a large sample of SEC documents and quantify their readability. The results are unambiguous: Acquirers write documents with lower readability and file them faster when investors react negatively to the deal announcement. These results are in line with strategic obfuscation efforts and also are robust to alternative empirical choices, such as controlling for content complexity or the firm’s writing style and using an instrumental variables approach.

The seminar will be held both in presence and online at this LINK

The seminar will take place in presence and online via teams: click here 

We provide, even for high dimensions, algorithms to sample from exchangeable multivariate Bernoulli distributions and to determine the distributions and the bounds of a wide class of in[1]dices and measures of probability mass functions. The generalization to the class of multivariate Bernoulli distributions with mean p is also addressed.