Asymmetric volatility in cryptocurrency markets: New evidence from smooth transition GARCH models

This paper investigates the presence of asymmetric volatility dynamics in Bitcoin, Ethereum, Ripple, and Litecoin. Asymmetric effects between good and bad news are traditionally modeled using threshold GARCH models that allow only for two possible variance regimes. We experiment a slightly flexible specification for the conditional variance by using a Smooth Transition GARCH (ST-GARCH) model, where a continuum of intermediate states is allowed between the two extreme volatility regimes. We feature an inverted asymmetric reaction for the majority of cryptocurrencies. The presence of positive return-volatility relationship, which is different from other traditional assets, supports the safe-haven hypothesis in cryptocurrencies.