An Application of Model and Non-Model Methods for Construction of a Composite Leading Indicators

Understanding how the business cycle is formed leads to make more favorable macroeconomic policies in line with economic stability and a reduction in the deviation of economic growth from the long-term path. Economic theories provide a comprehensive guide for economists to identify indicators that affect business cycles. A combination of leading indicators in a composite index can be useful in obtaining future signals from different sectors of the economy. The construction of a composite index requires several steps, and a non-model or model-based approaches should be used for this process. The purpose of this paper is to introduce modeling and non-model approaches for developing a leading composite index to identify the formation of business cycles. The modeling methods used in this paper are vector autoregressive, factor and Markov switching models. From the results of this paper, it can be argued that the model approach is more applicable because of the disadvantages of a non-model approach, such as the use of short and long-term periods of time. Failing to consider the sustainability characteristics of the target variable, ignoring the evolution of the leading indicator over a period of time and the absence of information on short-term deviations from long-run equilibrium are some reasons to use the model approach more than non-model one.