Heterogeneity in economic relationships: Scale dependence through the multivariate fractal regression

0561394 - ÚTIA 2023 RIV NL eng J - Journal Article
Tilfani, O. - Krištoufek, Ladislav - Ferreira, P. - El Boukfaoui, M. Y.
Heterogeneity in economic relationships: Scale dependence through the multivariate fractal regression.
Physica. A : Statistical Mechanics and its Applications. Roč. 588, č. 1 (2022), č. článku 126530. ISSN 0378-4371. E-ISSN 1873-2119
R&D Projects: GA ČR(CZ) GJ17-12386Y
Institutional support: RVO:67985556
Keywords : Detrended fluctuation analysis * Heterogeneity * Multivariate fractal regression * Scale-dependent effects
OECD category: Economic Theory
Impact factor: 3.3, year: 2022
Method of publishing: Limited access
http://library.utia.cas.cz/separaty/2022/E/kristoufek-0561394.pdf https://www.sciencedirect.com/science/article/pii/S0378437121008037?via%3Dihub

Heterogeneity of effects between economic variables has been a frequently discussed topic for many years now. However, the estimation of such scale-dependent effects has proved challenging. Here, we propose a multivariate multiscale regression approach based on the combination of detrended fluctuation analysis and detrended cross-correlation analysis, but the idea can be easily translated into other time and frequency domain frameworks. As illustrations, we pick two classic economic models – the Taylor’s rule and the money demand function for the USA and Japan – and we uncover evident scale-dependence in the individual effects not visible by the simple regression tools. Importantly, the proposed framework can be used in any discipline where studying the effects at various scales is of interest. Further applications are thus certainly at hand.
Permanent Link: https://hdl.handle.net/11104/0334058