Publicação

Structural breaks and time-varying granger-causality: testing the spanning hypothesis

Ver documento

Detalhes bibliográficos
Resumo:This paper analyzes time-varying Granger-causality in the context of structural breaks and changes, employing recursive and recursive evolving window algorithms. A novel approach using the Flexible Fourier Form (Gallant, 1981) for modeling structural breaks is developed, and it is shown that there are no significant size distortions or power losses while accounting for structural breaks of unknown form and date in the deterministic component. Our methodology, applied to the Yield Curve Spanning Hy pothesis, reveals that the hypothesis cannot be universally upheld, as the yield curve only occasionally spans relevant information to predict yields.
Autores principais:Inverneiro, Miguel Teixeira Rodrigues
Assunto:Granger-causality Flexible fourier form Time-varying causality Structural breaks Yield curve Spanning hypothesis Recursive test
Ano:2024
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso embargado
Instituição associada:Universidade Nova de Lisboa
Idioma:inglês
Origem:Repositório Institucional da UNL
Descrição
Resumo:This paper analyzes time-varying Granger-causality in the context of structural breaks and changes, employing recursive and recursive evolving window algorithms. A novel approach using the Flexible Fourier Form (Gallant, 1981) for modeling structural breaks is developed, and it is shown that there are no significant size distortions or power losses while accounting for structural breaks of unknown form and date in the deterministic component. Our methodology, applied to the Yield Curve Spanning Hy pothesis, reveals that the hypothesis cannot be universally upheld, as the yield curve only occasionally spans relevant information to predict yields.