Granger Causality between Stock and Gold Returns – Evidence from Poland, Hungary and the Czech Republic

Abstract

Theoretical background: The safety of capital investments is one of the criteria in the decision-making process. Under conditions of uncertainty in financial market, investors’ interest in gold – as an alternative form of capital investment – is growing, especially in comparison with stocks perceived as more risky assets. Fluctuations in the prices of these assets make investors transfer their funds from gold markets to more profiable markets, or return to gold markets. From this point of view, it is important to analyse the causal relationships between returns on these assets. Purpose of the article: The aim of this paper is to investigate the causal relationship between the rates of return on investment in gold and stocks in selected countries of Central and Eastern Europe (Poland, Hungary, and the Czech Republic). We assume that the rates of return on the gold market constitute a Granger cause of the rates of return on the analysed stock markets. Research methods: To investigate the impact of the gold market on the stock market and vice versa , including the type and direction of Granger causality of the rates of returns, the VAR models were estimated and served as the basis for performing a Granger non causality linear test. The analysis of stationarity (ADF test) and cointegration (the Johansen test) of variables was carried out to select the final form of VAR/VECM model. The empirical data covered the period between December 1996 and December 2020. The analysis was carried out for the entire period and two sub-periods: the first one before the 2007 financial crisis and the second one, covering this crisis and the beginning of the ongoing COVID-19 pandemic. Stock indices included in this research are composed of the most liquid stocks, so-called blue chip shares traded on the Warsaw Stock Exchange – WIG20, the Budapest Stock Exchange – BUX index, the Prague Stock Exchange – PX index. The gold prices are expressed in domestic currencies (PLN, HUF, CZK, respectively). Monthly logarithmic returns on investment on the analysed markets constituted the variables of models reflecting the interrelations in the considered countries. Main findings: We show that, apart from two instances of unilateral causality running from gold to stock returns, no causal relationships were found in any of the directions (independence) for the entire sample. We prove that in the case of Hungary and the Czech Republic (significant at α = 5%), the rate of return on the gold market determined the rates of return on the stock market, whereas in Poland, no such causality was identified. To some extent, the research hypothesis was positively verified. The results also confim that changes in the stock market prices in all considered countries did not affect the change in the price of gold. After dividing the sample into two sub periods, no causal relationships were found between the analysed markets in the fist sub-period, whereas in the second sub-period, we indicate that the rates of return on gold determine the rates of return on stocks in two countries, namely Poland and Hungary. In this respect, the research hypothesis of one-way causality occurring in the analysed sub-periods is positively verified. Only in Hungary is the impact of gold on the stock market observed both in the entire period and in the second sub-period.