This paper outlines a framework for analysing the interaction between financial frictions at the household and firm level, liability dollarization and optimal monetary policy in a small, open economy subject to productivity and capital inflow shocks. It is found that, first, for the shocks under review, the extent of co-movement of financial variables pertaining to entrepreneurs and homeowners crucially depends on the degree of exchange rate flexibility. Second, for a central bank not concerned with financial stability, reacting to inflation and output is considered optimal. Third, including financial stability in the central bank's objectives results in an optimal monetary policy rule reacting to exchange rate depreciation, but not to credit growth, even in the case of large capital inflow shocks. In fact, reacting to credit growth reinforces the initial shock, increasing financial imbalances.
Given the indisputable cost of policy inaction in the run-up to banking crises as well as the negative side effects of unwarranted policy activation, policymakers would strongly benefit from earlywarning thresholds that more accurately predict crises and produce fewer false alarms. This paper presents a novel yet intuitive methodology to compute country-specific and state-dependent thresholds for early-warning indicators of banking crises. Our results for a selection of early-warning indicators for banking crises in 14 EU countries show that the benefits of applying the conditional moments approach can be substantial. The methodology provides more robust signals and improves the early-warning performance at the country-specific level, by accounting for country idiosyncrasies and state dependencies, which play an important role in national supervisory authorities' macroprudential surveillance.
This Occasional Paper presents a formal statistical evaluation of potential early warning indicators for real estate-related banking crises. Relying on data on real estate-related banking crises for 25 EU countries, a signalling approach is applied in both a non-parametric and a parametric (discrete choice) setting. Such an analysis evaluates the predictive power of potential early warning indicators on the basis of the trade-off between correctly predicting upcoming crisis events and issuing false alarms. The results in this paper provide an analytical underpinning for decision-making based on guided discretion with regard to the activation of macro-prudential instruments targeted to the real estate sector. After the publication of the ESRB Handbook and the Occasional Paper on the countercyclical capital buffer, it represents a next step in the ESRB's work on the operationalisation of macroprudential policy in the banking sector. This Occasional Paper highlights the important role of both real estate price variables and credit developments in predicting real estate-related banking crises. The results indicate that, in addition to cyclical developments in these variables, it is crucial to monitor the structural dimension of real estate prices and credit. In multivariate settings macroeconomic and market variables such as the inflation rate and short-term interest rates may add to the early warning performance of these variables. Overall, the findings indicate that combining multiple variables improves early warning signalling performance compared with assessing each indicator separately, both in the non-parametric and the parametric approach. Combinations of the abovementioned indicators lead to lower probabilities of missing crises while at the same time not issuing too many false alarms. In addition to EU level, they also perform relatively well at individual country level. Even though the best performing indicators have relatively good signalling abilities at the individual country level, national authorities are encouraged to perform their own complementary analyses in a broader framework of systemic risk detection, which augments potential early warning indicators and methods with other relevant inputs and expert judgement.
This paper describes the Macroprudential Database (MPDB) of the European Central Bank (ECB), which is an important component of the ECB's Statistical Data Warehouse. After explaining the rationale for creating the MPDB, the paper illustrates how it supports the macroprudential analysis conducted by the European System of Central Banks (ESCB), the European Systemic Risk Board (ESRB) and the national authorities of the Single Supervisory Mechanism (SSM) and the European Union. The structure of the database and a broad overview of available indicators are then presented, with a description of the relevant confidentiality issues. Examples illustrate how the MPDB is used for monitoring purposes and econometric modelling. Finally, the paper discusses remaining data gaps and expected future enhancements of the database.
