Financial-cycle early warning · Advanced markets

Advanced-Market Overheating Index

References

Methodology

  • Chen, S. & Svirydzenka, K. (2021). "Financial Cycles — Early Warning Indicators of Banking Crises?" IMF Working Paper WP/21/116. The method replicated here (Advanced-Market track): indicator selection, frequency bands, thresholds, and index weights. ssrn.com/abstract=3970194
  • Christiano, L. J. & Fitzgerald, T. J. (2003). "The Band Pass Filter." International Economic Review 44(2), 435–465. The filter used for every gap on this page (see About & method). Free working-paper version: NBER WP 7257 (1999), nber.org/papers/w7257
  • Harding, D. & Pagan, A. (2002). "Dissecting the Cycle: A Methodological Investigation." Journal of Monetary Economics 49(2), 365–381. The turning-point dating algorithm the paper used to derive its cycle-length bands (see About & method).
  • Claessens, S., Kose, M. A. & Terrones, M. E. (2011, 2012). "Financial Cycles: What? How? When?" IMF WP/11/76; "How Do Business and Financial Cycles Interact?" Journal of International Economics 87(1), 178–190. The specific implementation of Harding–Pagan cycle dating that the paper follows to set its bands. imf.org (WP/11/76)
  • Kaminsky, G., Lizondo, S. & Reinhart, C. (1998). "Leading Indicators of Currency Crises." IMF Staff Papers 45(1). The signal-extraction threshold approach behind the paper's early-warning loss function.
  • Laeven, L. & Valencia, F. (2018). "Systemic Banking Crises Revisited." IMF Working Paper 18/206. crisis-era markers on each country's history strip use its published dates (Systemic Banking Crises Database II, IMF Economic Review 68 (2020), 307–361), verified against the latest update — Laeven & Valencia (2026), Systemic Banking Crises Database: 1970–2025, IMF WP/26/94, which extends coverage through 2025, adds no new advanced-market episode, and assesses the March-2023 US/Swiss bank failures as short of its systemic definition (durations capped at 5 years). imf.org (WP/26/94)
    • Crisis database with coverage through 2025

Software

  • statsmodels — cffilter (Christiano–Fitzgerald filter), Python. The exact filter implementation run for every gap, including the drift adjustment. statsmodels documentation

Data providers

  • FRED, Federal Reserve Bank of St. Louis. Share-price indices (OECD Main Economic Indicators) and real GDP (IMF/OECD) for most markets; US CPI (BLS). fred.stlouisfed.org
  • Eurostat. HICP and quarterly national accounts (real GDP) for the EU/EEA bloc and Switzerland. ec.europa.eu/eurostat
  • Office for National Statistics (UK). UK CPI all-items index (D7BT, MM23). ons.gov.uk
  • Australian Bureau of Statistics. Australian CPI, all groups (ABS Data API). abs.gov.au
  • Statistics Canada. Canadian CPI, all-items (vector v41690973, table 18-10-0004-01). statcan.gc.ca
  • International Monetary Fund — International Financial Statistics, via DBnomics. CPI (and real GDP where FRED lacks it) for Japan, Switzerland, Korea, Israel, and New Zealand; all three series — share prices (FPE_IX), real GDP, and CPI — for Hong Kong. db.nomics.world
  • Bank for International Settlements. Credit and residential-property series (supporting context; underlying provider of several FRED series). bis.org