摘要 : Abstract We present a novel framework to predict recessions, the depth of recessions (mild or severe) and monetary policy pivots. The first phase introduces three ways to predict recessions. The first method evaluates the effectiv... 展开 Abstract We present a novel framework to predict recessions, the depth of recessions (mild or severe) and monetary policy pivots. The first phase introduces three ways to predict recessions. The first method evaluates the effectiveness of a few inverted U.S. Treasury yield curves. The second approach identifies a threshold between the federal funds rate and the 10-year Treasury yield. The final tool employs Probit modeling. The second phase of our analysis utilizes the 10-year/1-year Treasury yield spread to predict the duration of a recession. Historically, 12 consecutive months of a negative 10-year/1-year spread is associated with deeper recessions. When evaluating the tools using data going back to 1955, the 10-year/1-year spread has predicted all 10 recessions with an average lead time of 12 months. The 10-year/FFR threshold has predicted all recessions and monetary policy pivots with an average lead time of 18 months. At present, all three tools are signaling that a recession and/or Fed policy pivot is more likely than not within the next year. Through June 2023, the 10-year/1-year spread has remained negative for 12 consecutive months, which highlights the risk that an upcoming recession may not be mild. Given the historical accuracy of our framework, decision-makers may consider a possibility that the upcoming recession may be severe. 收起