The most recent
Abstract
Colombia’s annual infation reached 13.3% in March of 2023, the highest rate since the start of the infation-targeting regime for monetary policy in 2000. However, some groups in the basket show signs of lower infation, while others show higher infation. The persistence of this trend is a matter of active debate that involves analyzing the trend component of both year-to-year and month-to-month changes in the price indices. This paper employs time series models to identify infation shift levels based on the 188 price indices in the basket. We categorize trend breaks as positive or negative and further classify them into tradable versus non-tradable, core versus regulated, and other CPI categories. Using trend models that incorporate these breaks, we forecast total and group infation. Our results show that including trend breaks enhances prediction accuracy for monthly annual infation across all time horizons