Nonperforming Loans and Competing Rules of Monetary Policy: a Statistical Identification Approach

We assess to what extent decisions taken by the Federal Reserve in setting interest rates can be interpreted in the light of monetary policy rules that are either built on standard objectives of output and price stabilization or based on alternative objectives of financial stability and regulation of the solvency conditions in the economic system. This goal is pursued through a comparison between the “Taylor rule” in its “original” and “augmented” versions, and an alternative “Solvency rule”. We use nonperforming loans as a proxy for the conditions of financial stability and solvency in the system. The empirical investigation is carried out following a structural vector autoregressive approach that exploits a statistical identification procedure. In this way, we are able to identify the causal structure among variables without imposing theoretical restrictions on the model. Our empirical findings provide very limited and incomplete support for the Taylor rule in its various forms while give comprehensive evidence in favor of the alternative Solvency rule.