Γραμμικές και μη γραμμικές σχέσεις μεταξύ των διαταραχών των τιμών πετρελαίου και των χρηματιστηριακών αγορών
View/ Open
Keywords
Χώρες εισαγωγείς πετρελαίου ; Απόδοση πετρελαίου (Brent) ; Απόδοση ; Ευρωπαϊκοί χρηματιστηριακοί δείκτες ; Δείκτης βιομηχανικής παραγωγής ; Επιτόκια ; VAR ; Impulse responses ; Διάσπαση διακύμανσης ; Έλεγχος αιτιότητας Granger ; VECM ; NARDLAbstract
In this dissertation, is investigated how changes in Brent crude oil prices influence stock
market returns in eight European oil-importing countries (Belgium, Germany, France,
Italy, Spain, Netherlands, Poland, Sweden). The identification of relationships between
the two variables is initially conducted, considering two additional control variables: the
industrial production index and short-term interest rates on government securities. Then
Vector Autoregressive (VAR) and Vector Error Correction Models (VECM) frameworks
were used to apply the methodologies of impulse responses, variance decomposition, and
Granger causality tests for the period from September 19, 2003, to May 19, 2023. The
main finding indicates a positive correlation between stock market returns and oil market
returns during this period. The variance decomposition also revealed that the volatility in
the oil and stock markets is primarily explained by their own changes after six months,
while endogeneity was confirmed in several cases across all countries. In the subsequent
stage, through the Nonlinear Autoregressive Distributed Lag (NARDL) framework, was
examined whether there is cointegration between variables and short-run, long-run
asymmetry in the behavior of stock market returns when positive or negative changes
occur in oil prices. The results were mixed, and asymmetry was not confirmed in most
countries, while in all cases, the cointegration test indicated the presence of long-term
relationships between oil and financial markets.