Skip to main content
Log in

The effect of IMF communication on government bond markets: insights from sentiment analysis

Review of World Economics Aims and scope Submit manuscript

Cite this article


This article examines whether the IMF sentiment conveyed by the Regional Economic Outlook (REO) provides new information capable of influencing government bond markets. To measure IMF sentiment, we use text mining techniques on an original dataset based on the qualitative content of the REO reports for 16 countries across three regions covered by the REO, Asia and Pacific, Europe, and Western Hemisphere, from 2005 to 2018. Our results suggest that the qualitative content of the REO reports has significant repercussions on bond yields, particularly in the Asia and Pacific region, and provides a positive signal in bond markets of countries participating in an IMF program in the Europe and Western Hemisphere regions. IMF sentiment towards the leading trade partner can also be an essential source of bond markets’ reactions. These findings are robust when controlling for IMF quantitative forecasts in the empirical procedure, accounting for an alternative sentiment measure and controlling for other potential determinants of bond yields. They thus shed new light on the importance of IMF communication for guiding and managing markets’ expectations.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others


  1. See the statement of G7 Finance Ministers and Central Bank Governors (, and the memorandum on the Work Program on Strengthening the Archi-tecture of the International Monetary System, issued by the IMF Executive Directors of the G7 countries, to the IMF Director and Executive Board, October 30, 1998 (

  2. The IMFC advises the IMF Board of Governors on the supervision and management of the international monetary and financial system, in particular during events that may disrupt the system.

  3. Western Hemisphere: Argentina, Brazil, Canada, Mexico and the United States. Europe: France, Germany, Poland, Russia, Turkey and the United Kingdom. Asia and Pacific: Australia, China, India, Japan and South Korea.

  4. Fig. 3 in Appendix 1 shows the geographical coverage of the REO reports.

  5. The publication of the REO report starts at different dates, depending on the regions: in October 2004 for Sub-Saharan Africa, September 2005 for the Asia and Pacific and Middle East and Central Asia, April 2006 for the Western Hemisphere, and November 2007 for Europe.

  6. Available at:

  7. GI words that are not relevant to IMF reports are excluded from the analysis by default.

  8. The standardization consists, for each sentiment measure, in subtracting the sample period’s mean and dividing by the sample period’s standard deviation of the sentiment measure.

  9. These changes might be explained by variations in the writing style due to turnover in the IMF’s writing teams.

  10. Positive (negative) Z-scores indicate that the degree of optimism (pessimism) in the text is above (below) average.

  11. We perfomed the Hausman test (Hausman, 1978) to check whether a fixed effects model is preferable to a random effects model. The test show a strong and significant heterogeneity amongst countries for most regressions. We thereby consider that a fixed-effect specification matches better the data generating process.

  12. \(\sum \limits _{i=0}^h \left( r_{i,t+i} - r_{i,t-1}\right) = \left( r_{i,t+h} - r_{i,t+h-1}\right) +\cdots +\left( r_{i,t} - r_{i,t-1}\right) = r_{i,t+h} - r_{i,t-1}\).

  13. Due to data availability, this analysis is not carried out for Turkey, Argentina, and Australia. Similarly, it is not possible to extend the analysis to the 20-year and 30-year government bond yields.

  14. For the sake of brevity, the results are not reported in the paper but are available from the authors upon request.

  15. See

  16. Standby Agreement from 05/11/2005 to 05/10/2008.

  17. 5 Flexible Credit Lines from 05/06/2009 to 11/02/2017.

  18. Standby Agreement approved on 20/06/2018.

  19. 8 Flexible Credit Lines from 17/04/2009 to 21/11/2021.

  20. Specifically, euro area countries appear to have responded to the 3% limit imposed by the Stability and Growth Pact by offering optimistic forecasts when they are most in danger of breaching the limit.

  21. The IMF is funded by a quota system where each country pays based on the size of its economy and its political importance in world trade and finance. More specifically, IMF quotas are distributed according to a four-pronged formula that considers a member country’s GDP, its economic openness, its economic variability and international reserves.

  22. The Fed major announcements made during the time windows considered correspond to the following dates: (i) October 12th 2012, (ii) April 28th 2010, April 27th 2011, and (iii) April 25th 2012, May 1st 2013, April 30th 2014, respectively.

  23. The ECB major announcements made during the time windows considered correspond to the following dates: (i) April 12th 2007, April 10th 2008, (ii) May 04th 2006, May 7th 2009 and (iii) May 2nd 2013.


  • Altavilla, C., Brugnolini, L., Gürkaynak, R. S., Motto, R., & Ragusa, G. (2019). Measuring euro area monetary policy. Journal of Monetary Economics, 108, 162–179.

    Article  Google Scholar 

  • Anderson, G., Galang, P., Gamba, A., Medina, L., & Zheng, T. (2021). How have IMF priorities evolved? A text mining approach. In IMF working papers 21/159, Washington DC: International Monetary Fund.

  • Batchelor, R. (2001). How useful are the forecasts of intergovernmental agencies? The IMF and OECD versus the consensus. Applied Economics, 33, 225–235.

    Article  Google Scholar 

  • Beck, N., & Katz, J. N. (1995). What to do (and not to do) with time-series cross-section data. American Political Science Review, 89, 634–647.

    Article  Google Scholar 

  • Boin, A., Stern, E., & Sundelius, B. et al. (2016). The politics of crisis management: Public leadership under pressure. Cambridge University Press.

  • Bollerslev, T. (1986). Generalized autoregressive conditional heteroskedasticity. Journal of Econometrics, 31, 307–327.

    Article  Google Scholar 

  • Brealey, R. A., & Kaplanis, E. (2004). The impact of IMF programs on asset values. Journal of International Money and Finance, 23, 253–270.

    Article  Google Scholar 

  • Breen, M., Hodson, D., & Moschella, M. (2019). Incoherence in regime complexes: A sentiment analysis of EU-IMF surveillance. Journal of Common Market Studies, 58, 419–437.

    Article  Google Scholar 

  • Dreher, A., Marchesi, S., & Vreeland, J. R. (2008). The political economy of IMF forecasts. Public Choice, 137, 145–171.

    Article  Google Scholar 

  • Evans, M. D., & Lyons, R. K. (2005). Do currency markets absorb news quickly? Journal of International Money and Finance, 24, 197–217.

    Article  Google Scholar 

  • Frankel, J., & Schreger, J. (2013). Over-optimistic official forecasts and fiscal rules in the Eurozone. Review of World Economics, 149, 247–272.

    Article  Google Scholar 

  • Fratzscher, M., & Reynaud, J. (2011). IMF surveillance and financial markets: A political economy analysis. European Journal of Political Economy, 27, 405–422.

    Article  Google Scholar 

  • Ganapolsky, E.J., & Schmukler, S. L. (1998). The impact of policy announcements and news on capital markets: Crisis management in Argentina during the tequila effect. In Working Paper, 5, Washington, DC: World Bank.

  • Gürkaynak, R. S., Sack, B., & Swanson, E. (2005). The sensitivity of long-term interest rates to economic news: Evidence and implications for macroeconomic models. American Economic Review, 95, 425–436.

    Article  Google Scholar 

  • Hallerberg, M., Strauch, R. R., & Von Hagen, J. (2009). Fiscal governance in Europe. Cambridge: Cambridge University Press.

    Book  Google Scholar 

  • Hausman, J. A. (1978). Specification tests in econometrics. Econometrica, 46, 1251–1271.

    Article  Google Scholar 

  • Hayo, B., & Kutan, A. M. (2005). The impact of news, oil prices, and global market developments on Russian financial markets. Economics of Transition, 13, 373–393.

    Article  Google Scholar 

  • Jaramillo, L., & Weber, A. (2013). Bond yields in emerging economies: It matters what state you are in. Emerging Markets Review, 17, 169–185.

    Article  Google Scholar 

  • Javed, F., & Mantalos, P. (2013). GARCH-type models and performance of information criteria. Communications in Statistics-Simulation and Computation, 42, 1917–1933.

    Google Scholar 

  • Jonung, L., & Larch, M. (2006). Improving fiscal policy in the EU: The case for independent forecasts. Economic Policy, 21, 492–534.

    Article  Google Scholar 

  • Kutan, A. M., Muradoglu, G., & Sudjana, B. G. (2012). IMF programs, financial and real sector performance, and the Asian crisis. Journal of Banking & Finance, 36, 164–182.

    Article  Google Scholar 

  • Lombardi, D., & Woods, N. (2008). The politics of influence: An analysis of IMF surveillance. Review of International Political Economy, 15, 711–739.

    Article  Google Scholar 

  • Loughran, T., & McDonald, B. (2011). When is a liability not a liability? Textual analysis, dictionaries, and 10-Ks. The Journal of Finance, 66, 35–65.

    Article  Google Scholar 

  • Loughran, T., & McDonald, B. (2020). Textual analysis in finance. Annual Review of Financial Economics, 12, 357–375.

    Article  Google Scholar 

  • Marinheiro, C. F. (2011). Fiscal sustainability and the accuracy of macroeconomic forecasts: Do supranational forecasts rather than government forecasts make a difference? International Journal of Sustainable Economy, 3, 185–209.

    Article  Google Scholar 

  • Mihalyi, D., & Mate, A. (2019). Text-mining IMF country reports: An original dataset. Available at SSRN 3268934 .

  • Schmeling, M., & Wagner, C. (2019). Does Central Bank Tone move asset prices? CEPR Discussion Paper 13490. London: Centre for Economic Policy Research.

  • Svirydzenka, K. (2016). Introducing a new broad-based index of financial development. In IMF working papers, 16/5, Washington, DC: International Monetary Fund.

  • Twedt, B., & Rees, L. (2012). Reading between the lines: An empirical examination of qualitative attributes of financial analysts’ reports. Journal of Accounting and Public Policy, 31, 1–21.

    Article  Google Scholar 

Download references


We are grateful to Jean-Yves Gnabo, Gatien Bon, two anonymous referees as well as participants of the 2021 European Public Choice Society’s annual meeting and the 14th Annual Conference on The Political Economy of International Organization for constructive comments and suggestions.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Hamza Bennani.

Ethics declarations

Conflict of interest

All authors declare that they have no conflicts of interest.

Additional information

Publisher's Note

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


Appendix 1

See Tables 4, 5; Fig. 3.

Table 4 Number of reports published by the IMF (2000–2016)
Fig. 3
figure 3

Geographical coverage of the REO reports

Table 5 Summary statistics—number of times a country is cited in the REO reports

Appendix 2

See Tables 6, 7 and Figs. 4, 5, 6

Table 6 Share of sentiment words (in % of the total number of words)
Fig. 4
figure 4

IMF Sentiment towards each country, Asia and Pacific. Note: The figures show the evolution of Z-score for each sentiment index—\(t_{i,t}\) (solid line), \(t_{i,t}^{alt}\) (dotted line)—and for each selected country within the Asia and Pacific region

Fig. 5
figure 5

IMF Sentiment towards each country, Europe. Note: The figures show the evolution of Z-score for each sentiment index—\(t_{i,t}\) (solid line), \(t_{i,t}^{alt}\) (dotted line)—and for each selected country within the Europe region

Fig. 6
figure 6

IMF Sentiment towards each country, Western Hemisphere. Note: The figures show the evolution of Z-score for each sentiment index—\(t_{i,t}\) (solid line), \(t_{i,t}^{alt}\) (dotted line)—and for each selected country within the Western Hemisphere region

Table 7 Main trading partners

Appendix 3

See Tables 8, 9, 10, 11 and 12.

Table 8 Testing the effect of changes in the IMF sentiment on cumulative changes in 5-year bond yields, using an alternative sentiment measure
Table 9 Testing the effect of changes in the IMF sentiment on cumulative changes in 10-year bond yields, using an alternative sentiment measure
Table 10 Quota shares
Table 11 Testing the effect of changes in the IMF sentiment on cumulative changes in 5-year and 10-year bond yields, investigating the role of main IMF shareholders
Table 12 Testing the effect of changes in the IMF sentiment on cumulative changes in 5-year and 10-year bond yields, controlling for the Fed and the ECB monetary policy announcements

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bennani, H., Couharde, C. & Wallois, Y. The effect of IMF communication on government bond markets: insights from sentiment analysis. Rev World Econ (2023).

Download citation

  • Accepted:

  • Published:

  • DOI:


JEL Classification