An investigation into the dynamic relationship between gold, silver and oil: an intra-day analysis

• Main Author: Stanford, Den (Author)
• Other Authors: Frijns, Bart (Contributor), Tourani-Rad, Alireza (Contributor)
• Format: Others
• Published: Auckland University of Technology, 2013-09-19T20:58:13Z.
• Subjects: Cointegration; Dynamic Cointegration; Gold; Silver; Oil; High-frequency Data; Intraday; Vector Error Correction Modelling; Vector Error Correction Model; VECM; Vector Auto Regression; VAR; Impulse Response Functions; Johansen-Juselius Technique; Long-run Relationship; Futures; S&P 500; Barclays Global Aggregate Bond Index; Price Leadership; Regression Analysis; Error Correction Model; ECM; Stock; Bond; Thesis
• Online Access: Get fulltext

 In this research, the long-run relationships between gold, silver and oil were studied using cointegration analysis. Their dynamic cointegration was also examined. Despite economic recession and crises, cointegration did not disappear, and the strength of dynamic cointegration rose and fell. Price leadership was also investigated by studying impulse response functions. It was found that in periods of weak or no cointegration, gold was led by silver (to a greater extent) and oil (to a lesser extent). It appears that in the periods when there is cointegration, gold is led by silver and oil, and oil is led by silver. The magnitude of responses may appear small, however, they are high in frequency, and low levels of impulse response functions may still be economically significant. Determinants of long-run relationships were also studied by analysing the impact of the stock and bond markets' returns and volatility on gold, silver and oil cointegration strength. It appears that both markets impact the commodities. Cointegration strength between gold and silver, gold and oil, and silver and oil falls during periods of crisis, recession and market turbulence in the stock and bond markets. Economic recovery was not found to have any impact on cointegration strength between the commodities. This could mean that commodities' price relationships are unaffected by the stock and bond markets during recovery periods, and are probably explained by other factors. It is also possible that although the economy seemed to be growing, this was a period of uncertainty with no real trends, in which case there are no meaningful results of the regression.