Revisiting the Leading Economic Indicators

Abstract

Leading economic indicators have long been a tool of American economists, particularly those working in the business sector, for anticipating turning points in the business cycle. Armed with knowledge of likely peaks and troughs in the pace of aggregate economic activity, business economists can advise corporate leaders as to the probable path of the macroeconomy, thereby influencing if not improving the quality of strategic decision making within organizations. This chain of events is predicated on the assumed reliability of leading indicators to forecast correctly the future, an assumption put to the test in this paper via a novel application of statistical process control (SPC) to a well-known set of leading indicators that have been studied for the better part of half a century.

To give context to the overall discussion, the paper begins with a quick review of the historical development of leading indicator forecasting as it evolved in the United States. This is followed with an explanation of statistical process control, the singular methodology used in this paper, but one seldom employed in general economic analysis save for the area of production economics and its emphasis on manufacturing. Once explained, the SPC process is applied to a representative set of eleven leading indicators that have been tracked quarterly or more frequently for anywhere from 38 to 71 years.

The results of the SPC analysis of this data pool of some 7,000+ observations suggest that collectively leading indicators reliably forecast business-cycle turning points, with the caveat that individually the effectiveness with which specific indicators within a set predict the future of the macroeconomy is subject to wide variation.