期刊
EMPIRICAL ECONOMICS
卷 -, 期 -, 页码 -出版社
PHYSICA-VERLAG GMBH & CO
DOI: 10.1007/s00181-023-02523-6
关键词
Fuel prices co-movement; Artificial neural networks; Latent factor; C45; Q41
This paper proposes a novel approach to identify a latent factor in the co-movements of gasoline and diesel prices in three major European Union economies. The study demonstrates the presence of an unobservable component that influences fuel price co-movements and has implications for short-term fuel price forecasts.
This paper proposes a novel approach to identify the presence of a latent factor in the co-movements of gasoline and diesel prices in the three major European Union economies, (France, Germany, and Italy) using daily data from January 3, 2005, to June 28, 2021. More precisely, we advance an artificial neural networks algorithm estimated through a machine learning experiment through the backpropagation system to show that the neural signal is altered by an element that could coincide with a latent factor in the fuel price co-movements. We consider the role of the fuel tax systems and the connection between gasoline and diesel prices in these countries. The estimations indicate the presence of an unobservable component (the latent factor) in the fuel price co-movements, capable of influencing NN. This result validates the previous findings reported in the literature, indicating an excess co-movement in fuel prices. It also has implications in terms of fuel price forecasts in the short run.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据