Logo
Munich Personal RePEc Archive

Is the Internet Search Driving Oil Market? A Revisit through Time-Frequency approaches

Bouoiyour, Jamal and Selmi, Refk (2015): Is the Internet Search Driving Oil Market? A Revisit through Time-Frequency approaches.

[thumbnail of MPRA_paper_66214.pdf]
Preview
PDF
MPRA_paper_66214.pdf

Download (1MB) | Preview

Abstract

The present research seeks to address whether internet search drives oil market. For this purpose, we perform two analyses to empirically gauge the relevance of Google search Index as a measure of investors’ attention. Firstly, we test if extracting public moods oriented to crude oil using web contents, can help to predict crude oil. Secondly, we analyze the informational content of three oil events (OPEC cuts, 2008 global financial crisis and Libya war) in terms of their effects on the behavior of the crude oil. To achieve this goal, we intend to decompose the causality between attention and oil price into different time scales and frequencies using frequency domain causality test and nonlinear causality test-based wavelet. To ascertain the robustness of our results, we replicate the same testing procedure using another attention proxy which is the number of tweets. The paper decisively confirms that there is a short-run relationship between attention and crude oil. In addition, we show that world crude oil responding to oil events display sharp differentiation. If OPEC cuts had short- and medium-run causality and Libya war exhibits a short-term causality, the attention to global financial collapse had a longer time interval and a wider scale of influence. The first finding implies that internet search is a very practical way to compute investors’ attention that can help in predicting short-run fluctuations in the oil market. For the second outcome, different shock origins and distinct properties of oil events may be advanced as possible element of explanation that may exhibit different effects on crude oil.

Atom RSS 1.0 RSS 2.0

Contact us: mpra@ub.uni-muenchen.de

This repository has been built using EPrints software.

MPRA is a RePEc service hosted by Logo of the University Library LMU Munich.