e-Guest Lecture : Andras Fulop

Likelihood under Control and Estimating Dynamic Structural Macro-Finance Models (Joint with Jeremy Heng at ESSEC Singapore and Junye Li at Fudan)

Andras Fulop est professeur au Département de Finance à l'ESSEC Business School - France

Solved dynamic structural macro_finance models are often non-linear and/or non-Gaussian with high dimension and complex structure. We propose an annealed
controlled sequential Monte Carlo method that delivers an efficient estimate of likelihood function by constructing the globally optimal proposal distributions
relying on approximate dynamic programming and by following a tempering procedure for gradually absorbing information in observations. We further develop
a new adaptive SMC2 algorithm with annealed controlled sequential Monte carlo nested. Finally, we show that our proposed methodology performs well by estimating
two popular macro-finance models, a new Keynesian dynamic stochastic general equilibrium model and a consumption-based long-run risk model. The next steps in our research agenda will also be discussed.

Date : 20 mai 2021 de 12h30 à 14h00

La vidéo est publiée sur la chaîne YouTube de CY AS