A Leading Indicator Model for Ohio SMSA Employment

ABSTRACT The methodology used to build a leading indicator model for Ohio SMSA employment is described in this paper. A composite leading indicator series for each of the eight major metropolitan areas and the state of Ohio was constructed. These composite indicators appear to work well in a “real‐time” simulation of their actual use. The methodology employed here departs from traditional methods for constructing such leading indicators in the approach to weighting the individual indicator series in order to devise a composite index. An econometric approach to determining the weights was employed. The weighting method can be said to be “optimal” in that the weights were chosen to maximize the out‐of‐sample ability of the composite index to detect future changes in economic activity, proxied here by the level of employment. This approach to weighting the component series in devising a composite index is computationally expensive, since it requires that a number of models be estimated and simulated in an actual use environment. It does overcome the usual subjective nature of the weighting schemes employed, and has resulted in composite indexes for the eight metropolitan areas and Ohio that perform quite well.