Disagreement in inflation expectations: empirical evidence for Colombia

ABSTRACT

The literature on expectation disagreements in emerging economies is scarce. This paper examines the disagreements in inflation expectations for the Colombian economy during the 2010–2017 period. We combine empirical tests with an analysis of a monthly survey of expectations of financial analysts in Colombia to obtain valuable evidence to formulate guidelines on the expectations modelling in developing economies. The findings indicate that disagreements present inertia and that inflation volatility increases disagreements. However, the central bank’s stance, as established through a press release, can reduce disagreement. Moreover, if central bank communication is clear and there is a credible inflation target, there tend to be fewer disagreements.

KEYWORDS:

• Disagreements
• inflation expectations
• communication
• clarity

JEL CLASSIFICATION:

• E37
• E58
• E44

Acknowledgments

We thank an anonymous referee for helpful comments on an earlier version of this paper. Any remaining errors are the sole responsibility of the authors.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ In the Colombian case, statistics are presented in an aggregate form and comprise measures of central tendency such as mean and median, and measures of statistical dispersion such as standard deviation, minimum and maximum values. There are no statistics available on the expectations of each respondent.

² Disagreements observed in February 2010 for the inflation expectation at the end of 2010 tend to be higher than the disagreements observed in October 2010. Divergence tends to increase again in February 2011, when forecast horizon changes again. See Dovern, Fritsche, and Slacalek (2012).

³ There is a methodology for computing weights optimally. In this regard, Knüppel and Vladu (2016) perform an optimal approximation for fixed-horizon forecasts using fixed-event forecasts by minimizing the mean-squared approximation error.

⁴ To analyse the readability of a text in Spanish, changes to the Flesch formula (1948) are required to capture typographical features of the Spanish language. See Barrio-Cantalejo et al. (2008).

⁵ The interpretation scale of the Flesch index is defined in the interval 0–40: very difficult; 40–55: somewhat difficult; 55–65: normal; 65–80: fairly easy; 80–100: very easy.

⁶ Appendix A.1 provides the details of the data sources and the time series used in this study. The descriptive statistics are presented in Appendix A.2.