Analyzing the Effectiveness of the U.S. Yield Curve as a Leading Economic Indicator

Open Access
Author:
Kalgutkar, Aditya Anil
Area of Honors:
Economics
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
  • Russell Paul Chuderewicz, Thesis Supervisor
  • Russell Paul Chuderewicz, Honors Advisor
  • James R. Tybout, Faculty Reader
Keywords:
  • Yield Curve
  • Interest Rates
  • U.S. Treasury Yield Curve
  • U.S. Recessions
  • Predicting Recessions
  • Probit Model
  • Economic Indicator
Abstract:
This thesis aims to contribute to the growing body of literature investigating the United States yield curve’s predictive power of recessions. First, I discuss yield curve theory, explaining the structure of the yield curve and linking it to forward-looking interest rate expectations. I follow with a discussion of monetary policy, transmission lag, and potential pollutants of the yield curve’s signal to establish the importance of the slope of the yield curve and to highlight the need for its evaluation as a reliable economic indicator. I then conduct a literature review covering the various techniques and approaches used in this field over the past few decades. Drawing from previous conclusions, I create a framework for study largely based on the probit model. I look at a full sample, a pre-1995 sample, and a post-1995 sample and analyze R2 and log-likelihood values to assess the fits of various probit models. I conclude my analysis using vector autoregression (VAR) to measure the response between percent change in GDP and the slope of the yield curve. From my analysis, I conclude that the yield curve still holds its standing as an effective forward-looking indicator, especially when used in conjunction with other explanatory variables in probit models. The yield curve is fundamentally tied to the market’s expectations for future interest rates, which are determined by monetary policy. As long as central banks maintain credibility and markets continue to regard forward guidance, the yield curve should continue to be reliable.