The Investigation of Inter-linkages between Precious Metals Markets and Stock Markets

The current study is dedicated to the investigation of inter-linkages between stock market and precious metals market. The data considered are daily close prices of stock indices of Russia, the USA, the Great Britain, China and South Africa, fixing prices of gold, silver, palladium, platinum and Brent oil from 03.01.2017 to 13.02.2020. At first, we examined the whole timespan without taking the possible presence of structural breaks into account. Having identified no long-run equilibrium, we built VAR-model for the returns and analyzed short-run relationships. Then, by performing the structural change test for each variable, we found out that structural breaks truly took place. Thereby, we split the sample into three subsamples. For all of them we detected the presence of cointegration, which enabled us to employ VECM. The instruments used in our analysis included stationarity tests, Johansen test, Granger-causality test, impulse response functions. The important remark which we would like to make for VAR/VECM in our study: the consideration of a too large sample might be misleading both for short- and long-term interconnections interpretation as the relationships between the variables changed through time. RTS index was determined to be in long-run relationships with gold and palladium for all three subsamples, whereas long-run connections with silver and platinum were present only on some of them. The results on one of the subperiods also showed that RTS increments depended on the changes of gold, platinum and palladium prices. Some other interesting results achieved are the significance of the Chinese stock index for palladium in the short-run for the last two subperiods, which was presumably caused by the increasing demand of China for this white metal heavily used in the industrial sector. Gold prices had a positive long-run effect on SSE. We think it may be due to the leadership of China in the world gold production. S&P 500 was found to affect negatively gold prices and positively white metals prices in the long run, which was quite expected. However, these dependencies disappeared in the third subperiod. In addition, we introduce DCC-GARCH model to analyze how conditional correlations between the pairs «precious metal – stock index» varied in time. In this part of the paper we show that shocks both in the global markets and in the local ones influenced conditional correlations. That helped us conclude the following: the structure of investment portfolios containing these assets needed to be reconsidered quite often.