Prediction of the Start of Next Recession

Ah-Hin Pooi and You-Beng Koh

The future value of the binary recession variable is modeled to be dependent on the present and past values of a set of m US economic variables selected from a pool of 14 variables via a conditional distribution which is derived from an -dimensional power-normal distribution. The mean together with the 2.5% and 97.5% points of the conditional distribution are used to predict the start of the next US recession. When and , some of the models can provide fairly good indicators for the start of the next US recession.