{
  "title": "Comparing the Performance of Logit and Probit Early Warning Systems for Currency Crises in Emerging Market Economies",
  "publication": "IMF Working Papers, April 17, 2014",
  "sourceUrl": "https://www.imf.org/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489",
  "canonical": "https://www.imf.org/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489",
  "overlayPath": "/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489/index.md",
  "summary": "We compare how logit (fixed effects) and probit early warning systems (EWS) predict insample and out-of-sample currency crises in emerging markets (EMs). We look at episodes of currency crises that took place in 29 EMs between January 1995 and December 2012.",
  "sections": [
    {
      "heading": "Summary",
      "content": "- Compares how logit (fixed effects) and probit early warning systems (EWS) predict in-sample and out-of-sample currency crises in emerging markets (EMs).\n- Sample covers episodes of currency crises that took place in 29 EMs between January 1995 and December 2012.\n- Main behavioral findings:\n  - Stronger real GDP growth rates significantly reduce the probability of experiencing a currency crisis.\n  - Higher net foreign assets significantly reduce the probability of experiencing a currency crisis.\n  - High levels of credit to the private sector increase the probability of experiencing a currency crisis.\n- Overall finding on model performance: logit and probit EWS out-of-sample performances are broadly similar.\n- Noted sensitivity: EWS performance can be very sensitive to both the size of the estimation sample and to the crisis definition employed.\n- Policy conclusion: For macroeconomic policy purposes, a currency crisis definition identifying more rather than less crisis episodes should be used, even if this may lead to the risk of issuing false alarms."
    },
    {
      "heading": "Methodology and Scope",
      "content": "- Econometric approaches compared: logit (fixed effects) and probit EWS.\n- Geographic and temporal scope: 29 emerging market economies; January 1995–December 2012."
    },
    {
      "heading": "Key Results and Statistics",
      "content": "- Sample size: 29 EMs.\n- Time period: January 1995 to December 2012.\n- Pages: 26.\n- Volume: 2014.\n- Issue: 065.\n- Series: Working Paper No. 2014/065.\n- DOI: https://doi.org/10.5089/9781484355282.001\n- Stock No: WPIEA2014065\n- ISBN: 9781484355282\n- ISSN: 1018-5941"
    },
    {
      "heading": "Policy Implications and Recommendations",
      "content": "- Use a crisis definition that identifies more crisis episodes for macroeconomic policy guidance, accepting increased risk of false alarms to improve early detection.\n- Monitor and address:\n  - Weak real GDP growth as a risk factor for currency crises.\n  - Low net foreign assets as a risk factor for currency crises.\n  - High credit to the private sector as a build-up of crisis vulnerability.\n- Consider sensitivity of EWS outputs to estimation-sample size and crisis-definition choices when designing surveillance and early-warning frameworks.\n\nIMF Working Paper by Fabio Comelli, April 17, 2014.\n\n---\n\n Content in this bundle\n\n- wp1465 - References\n  - wp1465 - References (Markdown version){rel=\"alternate\" type=\"text/markdown\"}\n  - wp1465 - References (PDF){rel=\"external\" type=\"application/pdf\"}\n\n---\n\nSource: https://www.imf.org/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489"
    }
  ],
  "bullets": [
    "[Markdown version](/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489/index.md)",
    "[Structured JSON version](/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489/index.json)",
    "[Bundle manifest](/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489/bundle-manifest.json)",
    "Authors: Fabio Comelli",
    "Published: April 17, 2014",
    "Series: IMF Working Papers",
    "DOI: https://doi.org/10.5089/9781484355282.001",
    "Compares how logit (fixed effects) and probit early warning systems (EWS) predict in-sample and out-of-sample currency crises in emerging markets (EMs).",
    "Sample covers episodes of currency crises that took place in 29 EMs between January 1995 and December 2012.",
    "Main behavioral findings:",
    "Overall finding on model performance: logit and probit EWS out-of-sample performances are broadly similar.",
    "Noted sensitivity: EWS performance can be very sensitive to both the size of the estimation sample and to the crisis definition employed.",
    "Policy conclusion: For macroeconomic policy purposes, a currency crisis definition identifying more rather than less crisis episodes should be used, even if this may lead to the risk of issuing false alarms.",
    "Econometric approaches compared: logit (fixed effects) and probit EWS.",
    "Geographic and temporal scope: 29 emerging market economies; January 1995–December 2012.",
    "Sample size: 29 EMs.",
    "Time period: January 1995 to December 2012.",
    "Pages: 26.",
    "Volume: 2014.",
    "Issue: 065.",
    "Series: Working Paper No. 2014/065.",
    "DOI: https://doi.org/10.5089/9781484355282.001",
    "Stock No: WPIEA2014065",
    "ISBN: 9781484355282",
    "ISSN: 1018-5941",
    "Use a crisis definition that identifies more crisis episodes for macroeconomic policy guidance, accepting increased risk of false alarms to improve early detection.",
    "Monitor and address:",
    "Consider sensitivity of EWS outputs to estimation-sample size and crisis-definition choices when designing surveillance and early-warning frameworks.",
    "**_wp1465 - References**"
  ],
  "related": [
    {
      "title": "_wp1465 - References",
      "role": "document",
      "sourceUrl": "https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1465.pdf",
      "summary": {
        "path": "/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1465.pdf.md",
        "mime": "text/markdown"
      },
      "binary": {
        "path": "/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1465.pdf",
        "mime": "application/pdf"
      }
    }
  ],
  "alternates": {
    "markdown": "/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489/index.md",
    "json": "/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489/index.json",
    "bundleManifest": "/en/publications/wp/issues/2016/12/31/comparing-the-performance-of-logit-and-probit-early-warning-systems-for-currency-crises-in-41489/bundle-manifest.json"
  },
  "generatedAtUtc": "2026-09-17T10:59:30.729Z"
}