| Description |
Recessions have been a concern for government and business leaders and the populace in general for hundreds of years. From the Tulipmania in 1634 to the Great Depression in the 1930's to the Dot Com bubble to the Great Recession in 2007 they have all instilled fear and anxiety in the populace because of financial insecurity, but also because of the relative uncertainty of knowing when a recession would begin. This has changed, to a degree, with the discovery of the relationship between the term spread in U.S. Treasury bonds and recessions. Arturo Estrella1 states that this relationship exists because of "monetary policy" and "market expectations" (Estrella, 2005). This relationship works both ways, so when a term spread is positive it is indicative of a growing economy and when the term spread is negative it is indicative of a recession. Professor Campbell Harvey2, Duke University, has successfully predicted recessions proving the validity of the relationship between term spreads and economic activity. A difference between these two researchers is how they calculate the term spread, either using a ten-year Treasury bond or a five-year Treasury bond minus a three-month Treasury note. The lack of uniformity in calculating the term spread with U.S. Treasuries is what led me to my research topic, determining the ideal calculation for the term spread to predict recessions by using probit models and linear regressions. |
