从风险到缓冲：校准欧元区正中性CCyB利率

Luis Herrera, Mara Pirovano, Valerio Scalone AbstractThis paper proposes a novel yet intuitive method for the calibration of the CCyB throughthe cycle in the euro area, including the positive neutral CCyB rate. The paper implementsthe Risk-to-Buffer framework by Couaillier and Scalone (2024) in both a DSGE and macrotime series setting and proposes a calibration of the PN CCyB aimed to reduce the macroe-conomic amplification of shocks occurring in an environment where risks are neither subduednor elevated.The suggested positive neutral CCyB rates for the euro area are consistentacross methodologies and robust to alternative specifications, ranging between 1% and 1.5%.The results also highlight the role of different shocks and sources of cyclical systemic risk forthe calibration of the CCyB through the cycle. The flexibility of the method regarding themodeling tools, the selection of specific levels of risks as well as the choice of state variablesand of exogenous shocks make it particularly suitable to be tailored to national specificitiesKeywords:Financial stability, macroprudential policy, capital requirements, countercyclical cap- and policymakers’ preferences.ital buffer.JEL Codes: C32, E51, E58, G01.ECB Working Paper Series No 3075 1 Non-Technical SummaryIn the aftermath of the COVID-19 pandemic, an increasing number of jurisdictions adopted amore proactive approach to the use of the CCyB and set a positive rate for the buffer in theearly phases of the financial cycle, when cyclical systemic risk is not elevated. The implementedtarget positive neutral (PN) CCyB across euro area countries rates range from 0.5% and 2%,reflecting policymakers’ preferences, country-specific characteristics, but also the different cal-ibration methods used. While the experience and international guidance on the calibration ofthe CCyB to address cyclical systemic risk is well-established, methods to inform the calibrationof the target PN CCyB rate are relatively scarce.Against this background, this paper proposes a novel method to calibrate the PN CCyB ratefor the euro area based on the Risk-to-Buffer approach developed by (Couaillier and Scalone(2024)).The method we propose is grounded in state-of-the-art techniques and is technicallyrigorous, while also being intuitive and easy to implement. The main idea underlying the Risk-to-Buffer approach is that higher risk leads to a greater amplification of adverse shocks, leading tomore severe macroeconomic outcomes and higher banking sector losses. Hence, different levels ofcyclical systemic risks will correspond to different calibrations of the CCyB rate. The calibrationof the CCyB using the Risk-to-Buffer approach involves two steps.First, a macroeconomicmodel is used to generate risk-dependent scenarios, namely the impact on GDP of set of adverseshocks, obtained for different systemic risk intensities. While the same set of shocks is used ineach scenario, higher risk leads to a greater amplification of adverse shocks, leading to moresevere macroeconomic outcomes and greater losses for the banking sector. In the second step,the losses associated to the different scenarios are mapped to the capital requirements needed tocover them. Specifically, we specify a mapping rule such that the CCyB is calibrated to absorblosses occurring under adverse scenarios corresponding to different levels of risk. Consistent withthe use of the PN CCyB in an environment where cyclical systemic risk are neither subduednor elevated, the PN CCyB rate is calibrated to address median cyclical systemic risk, whilethe CCyB rate at the peak of the cycle is calibrated to tackle elevated cyclical systemic risk.A further advantage of the method is that it is sufficiently flexible to allow policymakers toselect the preferred reference risk level. We implement the Risk-to-Buffer approach and obtainsuggested calibrations for the PN CCyB rate in both a structural (DSGE) and an empirical(macro time series) modeling framework.ECB Working Paper Series No 3075 2 We find that, first, taking the median systemic risk level as the relevant reference, the calibratedPN CCyB rates are consistent across the two approaches. Specifically, both the structural andthe baseline time series approach (using the ECB’s domestic systemic risk indicator as statevariable) suggest PN CCyB rates of 1.25% and 1.3% respectively. Overall, considering a broadset of cyclical systemic risk variables to define the risk states, the suggested PN CCyB ratesrange from 1% to 1.5%. While for the calibration of the PN CCyB rate, we are agnostic aboutthe specific source of shocks and apply all at the same time, the results are robust also acrossdifferent shocks.A second interesting finding from the empirical approach relates to the relationship betweenthe degree of nonlinear amplification generated by different shocks or different risk variables indetermining the relative importance of the PN CCyB in the overall CCyB calibration. We findthat shocks a