Energy commodities are deeply linked with the global economy and in the last decade its financial derivatives are vastly traded. Thus the examination of the relationship among them is an important task with many implications. This dissertation aims to examine the crude oil and natural gas price relationship while also taking into consideration the non-linear structure of our series. The data used consist of Western Texas Intermediate and Henry Hub spot prices as well as weather and storage shocks for the U.S. The existence of cointegration is examined by implementing the Rahbek & Mosconi (1999) fix to our the underlying VAR model as our system of variables, is a mix of endogenous I(1) and exogenous I(0) variables. After the establishment of cointegration we examine linear cau-sality between crude oil and natural gas prices by applying the Granger causality test. However we don’t want to neglect the nonlinearities of the series. The obtained residuals will be tested for i.d.d with the BDS test (1996). We implement the Breitung & Candelon (2006) frequency domain causality test to the delinearized residuals by conditioning on weather and storage shocks. The purpose of this test is to reveal the true causal relation-ship between natural gas and crude oil prices and surpass the limitations of the standard Granger causality test. Finally, by performing this test we are able to establish whether causality stands in the short-run or in the long-run.
Collections
Show Collections