Assessing the cross-country interaction of financial cycles: evidence from a multivariate spectral analysis of the USA and the UK
Abstract:
In recent times, a large number of studies has investigated the empirical properties of financial cycles within countries, mainly based on band-pass filter techniques. The contribution of this paper to the literature is twofold. First, in contrast to most existing studies in the financial cycle literature, we perform a multivariate parametric frequency domain analysis which takes the complete (cross-) spectrum into account and not only certain frequencies. Second, we provide evidence on the cross-country interaction of financial cycles. We focus on the USA and UK and use frequency-wise Granger causality analysis as well as structural break tests to obtain three main results. The relation between cycles has recently intensified. There is a significant Granger causality from the US financial cycle to the UK financial cycle, but not the other way around. This relationship is most pronounced for cycles between 8 and 30 years.
Año de publicación:
2019
Keywords:
- Financial cycle
- Vector autoregressions
- Coherency
- Indirect spectrum estimation
- Granger causality
Fuente:

Tipo de documento:
Article
Estado:
Acceso restringido
Áreas de conocimiento:
- Finanzas
- Desarrollo económico
Áreas temáticas:
- Economía financiera
- Ciencia militar
- Probabilidades y matemática aplicada