Gold and silver as safe havens: a fractional integration and cointegration analysis

Data Availability: Concerning the data, we confirm that there are no legal or ethical restrictions on sharing the data publicly since they all come from Yahoo Finance, a public source (https://es.finance.yahoo.com/). In particular, https://es.finance.yahoo.com/quote/***, where *** refers to each of the indices. Copyright: © 2023 Caporale, Gil-Alana. This paper investigates whether gold and silver can be considered safe havens by examining their long-run linkages with 13 stock price indices. More specifically, the stochastic properties of the differential between gold/silver prices and 13 stock indices are analysed applying fractional integration/cointegration methods to daily data, first for a sample from January 2010 until December 2019, then for one from January 2020 until June 2022 which includes the Covid-19 pandemic. The results can be summarised as follows. In the case of the pre-Covid-19 sample ending in December 2019, mean reversion is found for the gold price differential only vis-à-vis a single stock index (SP500). whilst in seven other cases, although the estimated value of d is below 1, the value 1 is inside the confidence interval and thus the unit root null hypothesis cannot be rejected. In the remaining cases the estimated values of d are significantly higher than 1. As for the silver differential, the upper bound is 1 only in two cases, whilst in the others mean reversion does not occur. Thus, the evidence is mixed on whether these precious metals can be seen as safe havens, though it appears that this property characterises gold in a slightly higher number of cases. By contrast, when using the sample starting in January 2020, the evidence in favour of gold and silver as possible safe havens is pretty conclusive since mean reversion is only found in a single case, namely that of the gold differential vis-à-vis the New Zealand stock index. Luis A. Gil-Alana gratefully acknowledges financial support from the Grant PID2020-113691RB-I00 funded by MCIN/AEI/ 10.13039/501100011033, and from an internal Project from the Universidad Francisco de Vitoria.