BANKS‘ CREDIT LOSS FORECASTS: LESSONS FROM SUPERVISORY DATA

CENTRAL BANK MONITORING III/2023

Inflation continues to slowly fall in most of the countries under review, but remains high. After a pause in June, the US Fed raised the key interest rate range again in July, with other central banks also raising their rates. By contrast, in Poland the rates have already been lowered. Spotlight focuses on the communication of central bank forecasts and their related uncertainties. In our selected speech, BoE Deputy Governor Dave Ramsden describes the bank’s experience to date with quantitative tightening. The current issue is available here.

MONETARY POLICY DECISIONS

Inflation continues to decline but is still expected to remain too high for too long. The Governing Council is determined to ensure that inflation returns to its 2% medium-term target in a timely manner. In order to reinforce progress towards its target, the Governing Council today decided to raise the three key ECB interest rates by 25 basis points. Explore the full article here.

EBA PUBLISHES THE RESULTS OF ITS 2023 EU-WIDE STRESS TEST

The European Banking Authority (EBA) published the results of its 2023 EU-wide stress test, which involved 70 banks from 16 EU and EEA countries, covering 75% of the EU banking sector assets. This stress test allows supervisors to assess the resilience of EU banks over a three-year horizon under both a baseline and an adverse scenario. The adverse scenario is characterised by severe negative shocks to economic growth, higher unemployment combined with higher interest rates and credit spreads. In terms of GDP decline, the 2023 adverse scenario is the most severe used in the EU wide stress up to now. The individual bank results promote market discipline and are used as part of the EU supervisory decision-making process.

ECB PUBLISHES SUPERVISORY BANKING STATISTICS ON SIGNIFICANT INSTITUTIONS FOR THE FIRST QUARTER OF 2023

Aggregate Common Equity Tier 1 ratio up to 15.53% in first quarter of 2023. Aggregated annualised return on equity up to 9.56% in first quarter of 2023. Aggregate non-performing loans ratio (excluding cash balances) decreased to 2.24%. Share of loans showing significant increase in credit risk (stage 2 loans) decreased to 9.31%. Read the full article here.

PERCEPTIONS OF RISK AND POLICY OUTLOOK DRIVE MARKETS

Perceptions of the future path of monetary policy shaped markets as central banks continued their fight against inflation. A benign assessment of the risk landscape supported risky asset valuations, notwithstanding subdued earnings forecasts. EME asset performance was generally strong but was also sensitive to AE financial conditions and to the ebb and flow of the US dollar. Continue reading.

THE TWO-REGIME VIEW OF INFLATION

The global surge of inflation that started in 2021 took most observers by surprise. It threatened to thrust the world back into an inflationary environment that we thought we had finally escaped decades ago. This study – prepared by BIS staff – documents the stylized facts describing the two regimes – low and high inflation – and the transitions between them based on disaggregated price dynamics and the joint behaviour of wages and prices.

UNEXPECTED LOSS ESTIMATION – R SHINY

Simple application which demonstrates one of the possible variants of implementing mathematical models in practice, which does not require large financial costs. The mathematical model (in this case it is a basic model for calculating unexpected risk based on the CreditMetrics methodology) is programmed in the freely available programming language R and the user interface is implemented in the Shiny framework. These globally used software tools allow easy implementation of very complex mathematical models and graphically advanced user interfaces.

SCORING DEVELOPMENT APPLICATION – R SHINY

This application is a simple yet effective tool for creating scoring models from initial univariate data analysis to basic validation of the resulting model. The range of methods and tools implemented has been deliberately simplified for this demonstration. If desired, additional functionalities and mathematical methods can be added to ensure that the application meets the needs and methodological requirements of users. The application is programmed in the freely available programming language R and the user interface is implemented in the Shiny framework.

WE SUPPORT CONNECTION WITH UNIVERSITIES BY LECTURING

Credit Risk, Financial Derivatives, Quantitative Investments – University of Economics
Sampling Theory, Data Exploration, Derivatives – Faculty of Mathematics and Physics, Charles University