By William Kinlaw, Mark Kritzman, and David Turkington.
Published in the Journal of Investment Management, Q3 2021.
Recipient of the 2021 Roger F. Murray First Place Prize Award.
We introduce a new index that synthesizes economic data to forecast the relative likelihood of recession versus high growth.
The index uses the Mahalanobis distance to measure the statistical similarity of current economic conditions with past episodes of recession and robust growth. Our approach has a key advantage compared to approaches that simply aggregate data, such as the Conference Board indexes, or approaches that rely on regression models. It considers the distribution of recession data separately from the distribution of growth data. This feature, along with the construction of the index as a relative probability, has the consequence of shifting the weights that are placed on the index inputs based on their prevailing values. In addition, our framework makes it possible to measure how the relative importance of the economic variables from which the index is constructed varies through time, which yields valuable insights about the dynamics of the business cycle.
