Analysis and Development of Leading Indicators Using a Bayesian Turning-Points Approach
提出一种贝叶斯决策方法，用于分析和改进区域先行指标模型，通过计算转折点概率来解读指标变化，帮助预测经济走势。
This article explores a Bayesian decision-theoretic approach for analysis and development of regional leading indicator models. The methods used here are derived from work by Zellner, Hong, and Gulati (1990) aimed at analyzing forecasts of turning points. Here, these methods are adapted so that they can be used to analyze existing regional leading-indicator series and to develop new improved versions of such series. The innovative aspect of this study is the use of the time series observations on the measure of economic activity that we wish to predict along with an explicit definition of a turning point, either a downturn or an upturn. We then establish a predictive relation between the composite indicator series and the variable measuring economic activity that allows a Bayesian computation of probabilities associated with the turning-point events. These probabilities are conditioned on the past data and the predictive pdf for future observations. Probabilities for the turning-point events make it quite clear how to interpret the information provided by period-to-period movements in the composite leading-indicator series. These methods can also be used to develop composite indicator series using the posterior probabilities from relations between individual indicator variables and the state of the economy as weights.