A0695
Title: Estimation of multivariate mixed causal and noncausal models: A review
Authors: Francesco Giancaterini - Maastricht University (Netherlands) [presenting]
Alain Hecq - Maastricht University (Netherlands)
Gianluca Cubadda - University of Rome Tor Vergata (Italy)
Abstract: Several strategies to estimate multivariate mixed causal and noncausal models have been proposed in recent years. The performance of the two most common estimators of these models are investigated, both when population parameters are known and unknown. The first estimator aims to maximize the approximate log-likelihood function, requiring a parametric specification of the error distribution. The second is a semi-parametric estimator that minimizes a specific objective function. The two existing estimation methods are compared using a bivariate process of Bitcoin-USD and Ethereum-USD exchange rates.