5. Conclusion
Recent literature has indicated the capacity of gold to act as a safe haven during times of financial turmoil. In this study we build on previous research, by investigating the capacity of gold, silver and platinum to mitigate extreme portfolio downside risk. We further investigate the cost or benefit of such risk reduction, by considering the impact on risk-adjusted returns. As the focus of the study is on rare events, we adopt a methodology appropriate to capturing infrequent but dangerous tail events. The Cornish-Fisher modified VaR adjusts the quantile of a distribution to account for higher-order moments of skewness and kurtosis. Empirical results are supportive of previous studies expounding the safe haven properties of gold. In particular, we conclude that gold provides strong downside risk reduction for equity investors across a range of short and medium horizons. In contrast, the equity risk reduction properties of silver and platinum, while strong at the shortest intervals considered, are abated at long intervals. For the longest horizons considered, both silver and platinum may be associated with increased downside risk for large allocations. We determine the contribution of various moments to downside risk reduction by examining modified VaR accounting, in turn, for two, three and four distributional moments. Evaluating the level of VaR reduction for each moment in turn, we find marginal risk reduction contributions from precious metals variance at all intervals studied. In contrast, the kurtosis properties of precious metals are found to reduce portfolio risk at short intervals but become a net contributor to risk at long intervals. A related issue, not examined here, is the role of predictability in returns in optimizing downside riskreduction (See Urquhart (2016), Urquhart, Batten, Lucey, McGroarty, and Peate (2015) and Charles, Darné, and Kim (2015) for evidence of predictability in precious metals markets).
