Building a coincident recession index: an application to the Peruvian economy

Does a decline in production signal a temporary slowdown or mark the beginning of a recession? To answer this question, this paper constructs a monthly coincident probabilistic index to detect recessions in the Peruvian economy using a nonlinear Markov-switching model. In constructing this index, emphasis is placed on the information content of consumer and business surveys, as well as real and international financial variables. The final index promptly and reliably detects the recessionary period associated with the 2008–2009 international financial crisis, even in a real-time analysis. However, because this index was developed using limited information due to the scarcity of data, its ability to detect future recessions has yet to stand the test of time. (Audio: Department of Economic Publications)