This code aims to replicate the autorgressive conditional beta algorithm presented in Blasques et al. (2024). The algorithm consists of a regression model with observation-driven time-varying paramaters. Both the regression parameters and the variance are defined by the score-driven updating equations. The model can be estimated with Quasi- Maximum Likelihood estimation. The model does not (yet) allow for extra regressors in the beta updating equations.
Blasques, F., Francq, C., & Laurent, S. (2024). Autoregressive conditional betas. Journal of Econometrics, 238(2), 105630. ISSN 0304-4076. Available at ScienceDirect. DOI: 10.1016/j.jeconom.2023.105630.