This paper constructs a composite-leading indicator (CLI) to forecast economic activity/business cycles in Mexico. This cycle is the result of deviations of the economy from its long-term trend. Here, a contractionary phase means a decline in the rate of growth of the economy, and not necessarily an absolute decline in economic activity. For this analysis it is necessary to select an indicator of economic activity (usually the Index of Industrial Production, IIP), as well as a group of variables that, when filtered/adjusted construct the CLI that forecasts the reference series. Selection of the components of the leading indicator is based on the forecast efficiency of the series and on their economic significance. Once selected, the relevant variables are aggregated into a single CLI that forecasts the detrended IIP. For detrending the series, the Hodrick-Prescott (HP) Filter method is applied.
This method is a smoothing technique that decomposes seasonal adjusted series into cyclical and trend components. One of the advantages of the HP Filter is that provides a reasonable estimate of the long-term trend of a series. Working Paper Number 01-08.