The way out of recessions: A forecasting analysis for some Euro area countries

Predicting recessions with boosted regression trees

“…Against the background of the Great Recession, researchers have started to reassess major linear and nonlinear forecasting approaches (Bec et al, 2014;Ferrara and Mogliani, 2015, among others) and leading indicators widely used in applied business-cycle research (for example, see Drechsel and Scheufele, 2012). In this research, we use a machine-learning approach known as Boosted Regression Trees (BRT) to reexamine the predictive value of selected leading indicators for forecasting recessions in Germany (on boosting, see Freund and Schapire (1997);Friedman (2001Friedman ( , 2002; Friedman et al (2000), for a survey, see Bühlmann and Hothorn (2007)).…”

Les cycles économiques de la France : une datation de référence

“…Le modèle utilisé ici est un modèle autorégressif à seuil, de type Threshold AR (TAR), augmenté d'une fonction susceptible de capturer un effet rebond de forme assez générale (voir Bec et al, 2014). En notant dx t la série représentant les taux de croissance trimestriels, le modèle TAR avec effet rebond, noté BBF(p, m, ), s'écrit de la manière suivante : T3 1974T2 1974T4 1973T2 1974M8 1974Creux T3 1975T4 1975T2 1975T3 1975M5 1975Pic T3 1976T1 1976Creux T1 1978T4 1976Pic T1 1980T1 1980T4 1979T3 1980T1 1980M8 1979Creux T4 1980T3 1980T3 1981T1 1982M10 1980Pic T4 1982T2 1982M12 1981Creux T1 1985T4 1985T2 1984M10 1983Pic T1 1992T1 1991T1 1990T1 1991T2 1990M2 1990Creux T1 1993T3 1993T3 1993T4 1993T4 1993M11 1993Pic T3 1995T1 1996T3 1995Creux T2 1996T1 1997…”

Post-recession US Employment through the Lens of a Non-linear Okun's law

“…Obviously, as in any other empirical applications, parameters need to be estimated, which leads to an additional uncertainty in predicted values. To our knowledge, there is no available clear procedure when dealing with a STECM, in spite of papers dealing with other types of non-linear models (see for example Li, 2011, or Bec, Bouabdallah andFerrara, 2013, in the case of univariate threshold models). 11,56 11,58 11,6 11,62 11,64 11,66 11,68 Q4 2007Q1 2008Q2 2008Q3 2008Q4 2008Q1 2009Q2 2009Q3 2009Q4 2009Q1 2010Q2 2010Q3 2010Q4 2010Q1 2011Q2 2011Q3 2011Q4 2011Q1 2012Q2 2012Q3 2012 Observed Conditional Forecasts STECM As shown, the long-run elasticity is very close to the one estimated with private employment (0.6155).…”