Modeling volatility of the cryptocurrency market: BEKK and DCC GARCH models

With the emergence of cryptocurrencies, it has received an extreme attention from different associations such as investors, businessmen, media and academic people. However, the prominent feature of cryptocurrency is being a source of risk and high volatile investment. Thus, many professional investors found it an opportunity for diversification, hedging and exchange tool. Therefore, the cryptocurrency volatility analysis and its performance in the portfolio has obtained significant considerations. Analyzing the volatility of cryptocurrency independently may not lead to accurate results. For this reason, gold, which is one of the most precious reserve assets in the world, and the Standard and Poor's 500 index, which is the most effective stock market, have been included in the evaluation. Therefore, the aim of this study is to analyze the volatility of the cryptocurrency market and the volatility spillover between cryptocurrency market, onus gold and S&P 500 index. For this purpose, the Cryptocurrency Index (KPE) has been developed. The KPE includes Bitcoin, Ethereum, Tether, Binance, Cardano, Dogecoin, and Ripple, which represent 71% of the market value in the calculation. Then, the volatility and volatility effects of the KPE, ouns gold, and S&P 500 index have been analyzed using BEKK GARCH and DCC GARCH models. The study was conducted between January 2, 2018, and August 31, 2022. Finally, minimum variance frontier applications were used to analyze the impact of the KPE on portfolio risk and return. According to the study results, the KPE has the highest and gold has the lowest volatility. Additionally, it has been concluded that the multivariate volatility effect between markets is stronger than each market's own volatility effect, and there is a covariance that changes over time between time series. On the other hand, according to the predictive ability of the models, it has been concluded that the Student-t DCC GARCH model is the best model. Furthermore, according to the results of the minimum variance frontier technique, including cryptocurrency investments in the portfolio significantly reduces the risk level given a specific level of return or effectively optimizes return in a specific level of risk. Finally, various recommendations have been made for investors, portfolio managers, and for future studies.