Banking crises and business cycle: evidence for Italy(1861-2016)

Abstract

Purpose This paper aims to focus on the banking crises recorded in Italy in the period 1861-2016 and to propose a novel classification based upon the timing of the crisis with respect to the business cycle. Design/methodology/approach A simple and objective rule to distinguish between slowdown and inner-banking crises is introduced. The real impact of banking crises is evaluated by integrating the narrative approach with an empirical vector autoregression analysis. Findings First, banking crises are not always associated to economic downturns. Especially in Italy, (but this analysis can be easily extended to other countries), they have often limited their negative effects within the financial system (“inner” crises). Second, the simultaneity of macroeconomic effects (credit contraction and GDP recession) leave the causal link undetermined. Third, the empirical and narrative analyses performed testify that boom–bust mechanisms are an exception in the panorama of (Italian) banking crises; although when the economy experiences such episodes, the economic and social consequences are not only severe but also enduring. Research limitations/implications To classify historically recognized banking crisis episodes, the authors look at credit and GDP dynamics (and their ratio) around crisis years. Relying on a single definition of crisis is avoided. The classification provides an empirical rule to determine in what way banking crises differ. The classification is mostly based on the synchronization with the business cycle and, using the documented evolution of macroeconomic aggregates, it permits to highlight the fact that a variety of interactions occur between financial and real aggregates during and around banking crises. Originality/value As to the concept of systemic banking crisis, a qualitative judgment is often adopted to select relevant episodes, thus confirming the absence of a quantitative rule in classification criteria (Chaudron and de Haan, 2014). This paper proposes a simple and objective rule to distinguish between slowdown and inner-banking crises; the former occur close to a GDP contraction, whereas the latter appear to spread their effects with no substantial evidence of output loss.

Untitled

“…(2) When coming to the period of interest in the present analysis, namely 2010-2016, Bartoletto et al [2018] find that two different banking crisis episodes have occurred: 2011 and 2013-2016. While the former meets the requirements of a boom-bust crisis, the latter is classified as an inner-banking crisis.…”

“…As it is clear from Table 1, adapted from Bartoletto et al [2018], the boombust crisis did exert negative and permament effects either on the rate of growth of GDP and on that of credit. Conversely, the period of turbolence in the banking system occurring in 2013-2016, when included in the VAR model estimated by Bartoletto et al [2018], did not contribute to explain neither GDP nor credit dynamics. Estimated coefficient lagged dummy -0.05* 0.002 *,**,***: statistically significant at 10, 5 and 1% respectively.…”

“…To this scope we employ the classification of banking crises introduced by Bartoletto et al [2018], who classify a banking crisis episode occurring in a specific year Tj as a "slow-down" crisis if the following condition holds: | − | ≤ 1 (1) where Tpeak is the date of the upward turning point. In other words, a "slow-down" crisis spreads its effects in one year time window around a business cycle peak.…”

“…Source: adapted from Bartoletto et al [2018] THE DATA SET In this article we employ the data set of the Survey of Industrial and Service Firms (Indagine sulle Imprese Industriali e dei Servizi), available through the Bank of Italy's remote processing system. 2 The survey is run annually by the Bank of Italy, and collects specific information on individual Italian firms, including several measures of credit restrictions which are the focus of our analysis, jointly with firm size (see http://www.bancaditalia.it/statistiche/indcamp/indimpser).…”

“…The main result of the paper emphasizes that in Italy, over the period 2010-2016, the severity of credit rationing was affected by the type of turbolence in the financial markets. Following the definition of banking crises adopted by Bartoletto et al [2018], it is found that only the most severe episode, related to the sovereign debt crisis, has seriously increased credit constraints, with remarkable increase in credit rationing experienced by large firms earlier compared to SMEs. This evidence might have a twofold explanation.…”