%0 Conference Paper
%A Shchekinova, E.
%A Puchkova, A.
%A Rovenskaya, E.
%A Dieckmann, U.
%B IIASA Institutional Evaluation 2017
%C IIASA, Laxenburg, Austria
%D 2017
%F iiasa:14425
%T Anticipating Recessions using Inclination Analysis
%U https://pure.iiasa.ac.at/id/eprint/14425/
%X Recessions are economic downturns that can be recognized from macro-indicators such as the Dow Jones Industrial Average (DJIA) and the Federal Reserve Interest Rate (FRIR). To provide early-warning signals of recessions and similar systemic transitions, here we propose a new approach based on pattern recognition, called inclination analysis [1, 2]. For this purpose, we develop a stochastic model based on time-series analysis to assess the probability of a recession to occur at a given moment in the past, present, or future. Calibrating our model to data proceeds in three steps, involving the coarse-graining of the available input time series, the identification of short series motifs that foreshadow recessions, and the optimization of key model parameters according to the model’s desired forecasting horizon.