ASEM 2018 INTERNATIONAL ANNUAL CONFERENCE PAPER BUSINESS CYCLES: A STATISTICAL PROCESS CONTROL PERSPECTIVE.

Since the late 1920s economic recessions have been considered part of an overall business cycle analysis with unknown, and variable, amplitude and frequency. The National Bureau of Economic Research formed a Business Cycle Dating Committee in 1978, and since then there has been a formal process of announcing the NBER determination of a peak or trough in economic activity. Over the year's researchers have focused on predicting the onset of recessions as well as future values of leading indicators of economic performance. A lack of definition of the business cycle has led to variable interpretations of observed economic performance and verbal gymnastics when economists describe the current state of the economy. A number of authors have used Markov Switching, CUSUM charts, Logistic Regression, and Support Vector Machine algorithms to try and predict the onset of a recession. Another extensive area of research has focused on using models to predict economic indicators. These prediction models include: Auto-regression, Factor (Principle Components) Augmented models, and Factor Augmented Auto-regression models. The authors envision a future where economic analyses start with a statement as to whether the economy is operating under common or special cause variation. Once this is determined analyses that are appropriate for the current status can be applied. This paper reviews the state of the art on Business Cycles, how they are quantified currently, define opportunities for partitioning economic indicators into special and common cause variation, and explore the impact of this change and its effects on the research into this area. [ABSTRACT FROM AUTHOR]