2014年-ECB欧洲央行_Capturing_the_Financial_Cycle_in_Euro_Area_Countries_9页_1mb

Summary of B: Capturing the Financial Cycle in Euro Area Countries

Core Content

This document explores the measurement of financial cycles in euro area countries for macro-prudential policy purposes. It highlights the importance of understanding financial cycles, which are more volatile and heterogeneous than business cycles, and discusses two primary methodologies used for their identification: spectral analysis and turning point analysis. The analysis is based on a set of macro-financial variables such as credit, house prices, equity prices, and interest rates, and provides insights into the characteristics of financial cycles across different euro area nations.

Main Views

1. Financial Cycles vs. Business Cycles

• Financial cycles are generally more volatile than business cycles.
• They exhibit strong cross-country heterogeneity, meaning that the length and intensity of financial cycles vary significantly between countries.
• Financial cycles tend to be longer than business cycles, with estimates suggesting they may last between 7 to 17 years compared to 3 to 8 years for business cycles.
• Financial cycle peaks are often associated with financial crises.

2. Importance of Financial Cycle Measurement

• Financial cycle measurement is essential for macro-prudential policy design, particularly in the euro area.
• There is no universally accepted definition of the financial cycle, and existing methods are preliminary and incomplete.
• A multivariate approach using several macro-financial indicators is considered more suitable than relying on a single variable.
• Robustness and policy relevance are key considerations in the selection of variables and methods.

3. Key Variables for Financial Cycle Analysis

• Credit growth, house prices, equity prices, and interest rates are the most commonly used variables to capture financial cycles.
• Real GDP and inflation are used to characterise business cycles.
• These variables are transformed into real terms and year-on-year changes to reflect their cyclical behavior.
• Credit is considered a key indicator due to its strong link with financial sector leverage and systemic risk.

4. Methodologies Used

a. Spectral Analysis

• A three-step procedure is used:

  1. Cohesion measure to identify the frequency range with the highest co-movement across variables.
  2. Band-pass filter applied using country-specific frequency bands to extract continuous financial cycles.
  3. Principal Component Analysis (PCA) to aggregate individual indicator cycles into a common financial cycle.

• Results show that financial cycles are longer than business cycles, with strong cross-country differences.

• The dominant frequency varies significantly, indicating that no single method can universally capture financial cycles.

• The first principal component represents a synthetic financial cycle as a linear combination of individual indicators.

b. Turning Point Analysis

• This method identifies local minima and maxima in the time series data.
• It has been applied to both business and financial cycles, providing insights into amplitude, duration, and slope of cycles.
• The analysis is based on standardised and smoothed variables, and uses common parameter settings as in Claessens et al. (2012).
• The method reveals that credit cycles have higher amplitude than business cycles, but similar duration.
• Equity price cycles tend to have higher volatility and more extreme changes.
• House price cycles are shorter in duration and less volatile than credit and equity cycles.

Key Information

• Financial cycles are more volatile and heterogeneous than business cycles, which is crucial for macro-prudential policy.
• The euro area has seen an increased need for financial cycle analysis due to the ECB’s macro-prudential oversight and the Single Supervisory Mechanism (SSM).
• The two methodologies used – spectral analysis and turning point analysis – provide complementary insights into financial cycle dynamics.
• The results show that financial cycles are longer (around 13 years) and more variable in length than business cycles (around 5 years).
• Country-specific characteristics play a significant role in the behavior of financial cycles, with credit and house prices being more dominant in some countries than others.
• Equity prices and interest rates contribute less to financial cycle determination, but their inclusion is still important for a comprehensive view.
• Amplitude and slope of cycles vary significantly across countries, indicating different levels of systemic risk and economic sensitivity.
• Synchronisation of financial and business cycles is not uniform across countries, with some showing strong alignment and others limited synchronisation.

Conclusion

• Financial cycle measurement remains nascent in the euro area, but two complementary methods have been developed to better understand its dynamics.
• These methods provide country-specific insights, which are vital for tailored macro-prudential policies.
• Robust and consistent financial cycle estimates are necessary for effective policy design and risk management in the euro area.