The Global Financial Crisis (GFC) led to a deep and long-lasting economic depression worldwide. The financial meltdown that followed the subprime crisis in the United States spreads rapidly, with dramatic consequences for the well-functioning of the global financial system. Therefore, several countries experienced huge costs in terms of output deterioration and unemployment rates. This project aims at answering the following question: how should the financial system oversight be organized in order to minimize social costs? The research question embraces two of the main challenges of the last decades of research, at both a theoretical and an empirical level: one is to test the interplay between macroprudential policy and monetary policy when they pursue financial stability. The second one is to deal with endogenous non-linearities of the economy which affect the policy conduct. The theoretical part will consist of a small-scale macroeconomic model with a non-linear state-space representation, where the Central Bank sets an optimal policy rate facing a convex Phillips curve and an IS curve with non-linear risk-premia and credit feedback. However, the model allows the presence of a financial authority which flags the creation of new loans in the banking system when credit dynamics exceeds its fundamental. Through numerical simulations the model will show how the Central Bank re-adjust the economy under both different macro scenarios and with a different degree of macroprudential policy. Then, the empirical part will consist of an estimation of the effects of a monetary policy shock where dynamic multipliers are allowed to change smoothly with the macroprudential stance. The challenge here will be twofold: one is to isolate the endogenous responses of the Central Bank to both macro financial imbalances and the second is to find a valuable strategy to represent different macroprudential regimes by exploiting observed macro data.
The idea of considering credit as a key driver of financial crises and/or as an amplifying engine of exogenous shocks is not new in the literature. This view, particularly in the pioneering contributions by Fisher [1933], Minsky [1977] and Kindleberger [1978], looks at the credit dynamics as an endogenous force that is prone to generate instability. Once speculative opportunities are present in the system, prolonged increases in the demand for goods and financial assets transmit themselves to prices. Higher prices mean new profit's opportunities that attract further investors. This positive feedback is characterized by the "euphoria" of economic actors and the boom starts to develop. The so-called "mania" phase comes as people's judgment on speculation for profits does not reflect the economic fundamentals of the underlying goods or assets.
As booms are generally fed by expansions of banks' credit, the latter is taken as an indicator of the temperature of systemic risk. The project proposed will add to the debate on the subject matter by providing a theoretical model highlighting the relevance of credit dynamics for policy makers, as well as new empirical insights about the control of credit dynamics and the conduct of both monetary and macroprudential policies.
These measures are not a free lunch for policy makers. Though reducing systemic risk is a benefit for financial stability, a policy that damps credit growth may trigger side-effects, like credit crunch and fire-sales that, especially if the economy is in a bad shape. This points to the conclusion that policy targeting risk-taking can be relatively costly when the economy contracts. Intuitively this result is not surprising, as banks' risk-sensitivity tends to be lower in good times and higher in bad phases.
At the same time, a potential trade-off between tightening macroprudential stance and a monetary policy that leans against the credit cycle may emerge. Little is known from the literature at both theoretical and empirical level about the mutual influence of the two. Though in the near future MP and MPP are going to address financial instability as complementary policy tools, as pointed out by the former Vice President of the ECB, Vitor Constancio [2016], it is necessary to improve the theoretical understanding of how these two should coordinate, and the empirical relevance of their interaction to gauge the effects of policy actions. These goals motivate and support this project which aims at revealing how MPP stance affects the optimal response of the Central Bank to financial risks.
The advancements in the literature that the project offers are manifold and not just limited to the subject matter. Indeed, on the theoretical side, the model proposed follows a growing literature in the economic field that borrows from the mechanic engineering, called Nonlinear Model of Predictive Control (see Grune et. al [2015]). These class of models has not only the advantage of including nonlinearities in the state-space, but also of letting the Central Bank to solve its optimal problem in a finite-horizon setting. This latter is more in line with the reality of our economic systems where agents make decisions under limited information, although they form forward-looking expectations (see Sims [2006] and Woodford [2018]).
Also the empirical investigation aims at including some methodological advancement to the applied literature. Indeed, the structural MP shock in TVC-VAR will be identified by using a proxy that catches exogenous variations in MP shifts that are thought to focus more on the effects on financial variables (see Caldara and Herbst [2019]). However, at the time of this statement, the project proposes a first attempt to include such a proxy of MP shock in a TVC-VAR. There is one similar example in the literature that can be found in a draft by Mumtaz and Petrova (https://www.dropbox.com/s/yba024afg84o0gk/Paper_Draft_Nov2018.pdf?dl=0). The latter, however, estimate the effects of a structural fiscal shock.