Skip to main content

Financial Crises, Macroprudential Policy and the Reliability of Credit-to-GDP Gaps


The Basel III regulation explicitly prescribes the use of Hodrick–Prescott filters to estimate credit cycles and calibrate countercyclical capital buffers. However, the filter has been found to suffer from large ex-post revisions, raising concerns on its fitness for policy use. To investigate this problem, we study credit cycles in a panel of 26 countries between 1971 and 2018. We reach two conclusions. The bad news is that the limitations of the one-side HP filter are serious and pervasive. The good news is that they can be easily mitigated. The filtering errors are persistent and hence predictable. This can be exploited to construct real-time estimates of the cycle that are less subject to ex-post revisions, forecast financial crises more reliably, and stimulate the build-up of bank capital before a crisis.

This is a preview of subscription content, access via your institution.

We’re sorry, something doesn't seem to be working properly.

Please try refreshing the page. If that doesn't work, please contact support so we can address the problem.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5


  1. The data are available at

  2. “Table A.2 of the Appendix” reports the crisis periods country by country. Our key results are robust to using a specific chronology rather than merging the crisis episodes: see Table A.3.

  3. All calculations are based on the entire sample period, so the smoothing revisions would have not been available in real time. The question of how the patterns in the data can be exploited in real time will be tackled in Sects. 4 and 5.

  4. The title “Commonwealth economies” refers to Australia, Canada, New Zealand, South Africa, UK and USA. Northern European countries include Austria, Belgium, Denmark, Finland, France, Germany, Netherlands, Norway, Sweden and Switzerland. As common, the acronym GIIPS refers to Greece, Ireland, Italy, Portugal and Spain. Asian countries include India, Japan, Korea Singapore and Thailand.

  5. “Figure A.1 of the Appendix” shows the filtered and smoothed credit cycles for all countries in our sample. We discuss specific cases and regional differences in Sects. 5 and 6.

  6. We rely throughout the analysis on a smoothing parameter \(\lambda\)= 400,000, as prescribed by the Basel agreements. The reason is, once again, that we intend to focus on the endpoint problems of the filter selected by the regulator rather than its general performance.

  7. Notice that the error of interest is the discrepancy between (i) the real-time estimate of the correction delivered by the models, and (ii) the corresponding full-sample estimate calculated ex-post by the Hodrick–Prescott filter, i.e., \({\widehat{C}}_t-C_{t-h|T}\). The errors are model-, country- and horizon-specific. For each model and horizon, the table reports the RMSE obtained by averaging the errors over countries and quarters. We provide additional details on the estimated equations in “Table A.4 of the Appendix”.

  8. We refer the reader to Sect. 4 for formal definitions of these terms.

  9. This highlights a general challenge for our approach: insofar as forecasts are less volatile than their targets, a predictive model may be unable to generate corrections that have both the right sign and a plausible magnitude when the credit gap is underestimated to start with. However, in Sects. 5.2 and 6 we show that (i) the adjusted filter performs well on average across the crisis episodes included in our sample, and (ii) it can be further improved by introducing additional predictors in the regression model.

  10. The statistics are calculated over the full sample. The results are similar if the evaluation is carried out using a restricted sample 1981–2007 that excludes some initial observations (for which the estimates might be unstable) and the global financial crisis: see Table A.5.

  11. Canada, New Zealand, Singapore and South Africa are excluded from this part of the analysis because of lack of information about financial crises. A full description of the crisis episodes is provided in “Table A.2 of the Appendix”.

  12. See “Appendix 1.1” and BCBS (2010) for more details on the BCBS CCyB rule.

  13. See Sect. 2 for a description of the dataset.

  14. The Global financial cycle indicator is constructed applying a dynamic factor model to a large panel of asset prices, and, as such, it has no direct relation to the credit quantities used to calculate credit-to-GDP gaps.


  • Adrian, Tobias, Nina Boyarchenko, and Domenico Giannone. 2019. Vulnerable growth. American Economic Review 109 (4): 1263–89.

    Article  Google Scholar 

  • Afanasyeva, Elena. 2020. Can forecast errors predict financial crises? Exploring the properties of a new multivariate credit gap, FEDS Working Paper (45).

  • Alessandri, Piergiorgio, Pierluigi Bologna, Roberta Fiori, and Enrico Sette. 2015. A note on the implementation of the countercyclical capital buffer in Italy. Bank of Italy Occasional Paper 25 (278): 38.

    Google Scholar 

  • Alessi, Lucia, and Carsten Detken. 2018. Identifying excessive credit growth and leverage. Journal of Financial Stability 35: 215–225.

    Article  Google Scholar 

  • Amiti, Mary, Patrick McGuire, and David E. Weinstein. 2019. International bank flows and the global financial cycle. IMF Economic Review 67 (1): 61–108.

    Article  Google Scholar 

  • Baba, Chikako, Salvatore Dell’Erba, Enrica Detragiache, Olamide Harrison, Aiko Mineshima, Anvar Musayev, and Asghar Shahmoradi. 2020. How should credit gaps be measured? An application to european countries, IMF Working Papers, (20/6).

  • Baron, Matthew, Emil Verner, and Wei Xiong. 2021. Banking crises without panics. The Quarterly Journal of Economics 136 (1): 51–113.

    Article  Google Scholar 

  • BCBS. 2010. Guidance for national authorities operating the countercyclical capital buffer,

  • BCBS. 2011. Basel III: A global regulatory framework for more resilient banks and banking systems, Revised version,

  • Cerutti, Eugenio, Stijn Claessens, and Andrew K. Rose. 2019. How important is the global financial cycle? Evidence from capital flows. IMF Economic Review 67 (1): 24–60.

    Article  Google Scholar 

  • Claessens, Stijn, Ayhan Kose, Luc Laeven, and Fabian Valencia. 2014. Financial Crises: Causes, Consequences, and Policy Responses. International Monetary Fund.

  • Darracq Pariès, Matthieu, Stephan Fahr, and Christoffer Kok. 2019. Macroprudential space and current policy trade-offs in the euro area, Published as part of the ECB Financial Stability Review,

  • DeLong, Elizabeth R., David M. DeLong, and Daniel L. Clarke-Pearson. 1988. Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach. Biometrics 44: 837–845.

    Article  Google Scholar 

  • Drehmann, Mathias, and James Yetman. 2018. Why you should use the Hodrick-Prescott filter-at least to generate credit gaps. BIS Working Paper 774: 24.

    Google Scholar 

  • Drehmann, Mathias and James Yetman. 2020. Which credit gap is better at predicting financial crises? A comparison of univariate filters. BIS Working Paper (878).

  • Drehmann, Mathias, and Mikael Juselius. 2014. Evaluating early warning indicators of banking crises: Satisfying policy requirements. International Journal of Forecasting 30 (3): 759–780.

    Article  Google Scholar 

  • Drehmann, Mathias, Claudio EV Borio, Leonardo Gambacorta, Gabriel Jimenez, and Carlos Trucharte. 2010. Countercyclical capital buffers: Exploring options, BIS working paper.

  • Edge, Rochelle M., and RalfR. Meisenzahl. 2011. The unreliability of credit-to-GDP ratio gaps in real-time: Implications for countercyclical capital buffers. International Journal of Central Banking 7 (4): 261–298.

    Google Scholar 

  • European Parliament. 2013. Directive 2013/36/EU of the European Parliament and the council,

  • Geršl, Adam, and Jakub Seidler. 2015. Countercyclical capital buffers and Credit-to-GDP gaps: Simulation for central, eastern, and southeastern Europe. Eastern European Economics 53 (6): 439–465.

    Article  Google Scholar 

  • Greenwood, Robin, Samuel G. Hanson, Andrei Shleifer, and Jakob Ahm Sørensen. 2020. Predictable financial crises, NBER Working Paper, (27396).

  • Hamilton, James D. 2018. Why you should never use the Hodrick-Prescott filter. Review of Economics and Statistics 100 (5): 831–843.

    Article  Google Scholar 

  • Jordà, Òscar., Björn. Richter, Moritz Schularick, and Alan M. Taylor. 2021. Bank capital redux: Solvency, liquidity, and crisis. The Review of Economic Studies 88 (1): 260–286.

    Article  Google Scholar 

  • Jordà, Òscar., Björn. Richter, Moritz Schularick, and Alan M. Taylor. 2011. Financial crises, credit booms, and external imbalances: 140 years of lessons. IMF Economic Review 59 (2): 340–378.

    Article  Google Scholar 

  • Jordà, Òscar., Björn. Richter, Moritz Schularick, and Alan M. Taylor. 2017. Macrofinancial history and the new business cycle facts. NBER Macroeconomics Annual 31 (1): 213–263.

    Article  Google Scholar 

  • Laeven, Luc, and Fabian Valencia. 2020. Systemic banking crises database II. IMF Economic Review 68: 1–55.

    Article  Google Scholar 

  • Lo Duca, Marco, Anne Koban, Marisa Basten, Elias Bengtsson, Benjamin Klaus, Piotr Kusmierczyk, Jan Hannes Lang, Carsten Detken, and Tuomas A Peltonen. 2017. A new database for financial crises in European countries: ECB/ESRB EU crises database, ECB Occasional Paper Series, (194).

  • Miranda-Agrippino, Silvia, and Hélene. Rey. 2020. US monetary policy and the global financial cycle. The Review of Economic Studies 87 (6): 2754–2776.

    Article  Google Scholar 

  • Monnet, Eric, and Damien Puy. 2021. One ring to rule them all? New evidence on world cycles, CEPR Discussion Paper, (15958).

  • Orphanides, Athanasios, and Simon van Norden. 2002. The unreliability of output-gap estimates in real time. Review of Economics and Statistics 84 (4): 569.

    Article  Google Scholar 

  • Rey, Hélène. 2015. Dilemma Not Trilemma: The Global Financial Cycle and Monetary Policy Independence. National Bureau of Economic Research: Technical Report.

  • Schularick, Moritz, and Alan M. Taylor. 2012. Credit booms gone bust: Monetary policy, leverage cycles, and financial crises, 1870-2008. American Economic Review 102 (2): 1029–61.

    Article  Google Scholar 

Download references


We acknowledge comments and suggestions from Stijn Claessens, Luigi Federico Signorini and two anonymous referees of the IMF Economic Review, and retain all responsibility for any remaining errors. The views expressed in this paper are those of the authors and should not be attributed to the Bank of Italy or the Eurosystem.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Maddalena Galardo.

Ethics declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.



1.1 Basel Committee Recommendations on the Calculation of the Countercyclical Capital Buffer

According to BCBS (2011) the credit-to-GDP gap is defined as the difference between an economy’s aggregate credit-to-GDP ratio and its long-run trend. The long-term trend of credit-to-GDP ratio is computed with a one-side (recursive) Hodrick–Prescott filter with a smoothing parameter \(\lambda\) = 400,000. Credit denotes a broad measure of the stock of domestic credit to the private non-financial sector outstanding at the end of quarter t. The credit-to-GDP gap is then translated in a percentage of the bank risk-weighted assets by calculating a benchmark buffer rate based on the piece-wise linear rule:

  • \(CCyB_t=0\) if \(GAP_t<2\%\).

  • \(CCyB_t=0.3125*GAP_t-0.625\) if \(2\%<GAP_t<10\%\).

  • \(CCyB_t=2.5\%\) if \(GAP_t>10\%\).

The credit definition suggested by BCBS (2011) includes all private credit issued by banks and non-bank financial institutions. Authorities are, however, allowed to use (i) additional measures of the credit gap and/or (ii) alternative de-trending methods in order to better capture the specificities of their national economies.

See Tables A.1, A.2, A.3, A.4, A.5, A.6, A.7 and Fig. A.1.

Table A.1 Credit-to-GDP ratios, sample statistics
Table A.2 Number of quarters
Table A.3 Alternative crisis database
Table A.4 Model selection: ARDL
Table A.5 Relation between ex-ante and ex-post estimates of the credit gap, 1981q1–2007q4
Table A.6 Extended model selection: ARDL
Table A.7 Impact of the information set on the estimated credit gaps
Fig. A.1
figure 6figure 6figure 6figure 6

Credit-to-GDP gap: comparing filtering methods

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Alessandri, P., Bologna, P. & Galardo, M. Financial Crises, Macroprudential Policy and the Reliability of Credit-to-GDP Gaps. IMF Econ Rev 70, 625–667 (2022).

Download citation

  • Published:

  • Issue Date:

  • DOI:

JEL Classification

  • E32
  • G01
  • G21
  • G28