Stress testing involves the use of simulation to assess the resilience of investment portfolios to changes in market regimes and extreme events. The quality of stress testing is a function of the realism of the market models employed, as well as the strategy used to determine the set of simulated scenarios. In this paper, we consider both of these parameters in the context of diversified portfolios, with a focus on the emerging class of cryptoasset-containing portfolios. Our analysis begins with univariate modelling of individual risk factors using ARMA and GJR–GARCH processes. Extreme Value Theory is applied to the tails of the standardised residuals distributions in order to account for extreme outcomes accurately. Next, we consider a family of copulas to represent the dependence structure of the individual risk factors. Finally, we combine the former approaches to generate a number of plausibility-constrained scenarios of interest, and simulate them to obtain a risk profile. We apply our methodology to the CoinShares Gold and Cryptoassets Index, a monthly-rebalanced index which comprises two baskets of risk-weighted assets: one containing gold and one containing cryptoassets. We demonstrate a superior risk-return profile as compared to investments in a traditional market-cap-weighted cryptoasset index.
CITATION STYLE
Koutsouri, A., Petch, M., & Knottenbelt, W. J. (2020). Stress Testing Diversified Portfolios: The Case of the CoinShares Gold and Cryptoassets Index. In Springer Proceedings in Business and Economics (pp. 43–64). Springer Science and Business Media B.V. https://doi.org/10.1007/978-3-030-53356-4_4
Mendeley helps you to discover research relevant for your work.