Bitcoin futures: trade it or ban it?

Abstract

This paper examines the impact of South Korea’s ban on Bitcoin futures on intraday spot volatility, liquidity and volatility–volume relationship. The results show that while reducing the permanent component of intraday spot volatility, the imposition of a ban on Bitcoin futures trading increases the transitory component. For intraday spot liquidity, different liquidity proxies indicate heterogeneous results. Moreover, we identify a positive and unidirectional effect of intraday spot volume on volatility. This effect appears to be stronger in the post-ban period. Overall, over the past few months, South Korea’s Bitcoin futures ban generally has had a significant impact on the intraday dynamics of the Bitcoin spot market.

KEYWORDS:

• South Korea’s ban on Bitcoin futures
• intraday dynamics
• component GARCH model

JEL Classifications:

• G12
• G13
• G14

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Based on the initiators of regulations, there are three types: (1) governments impose bans on cryptocurrency trading and opening new trading accounts, e.g. South Korea and China; (2) financial giants limit their clients to invest in cryptocurrencies. For instance, Merrill Lynch, JP Morgan, and Citi set certain prohibitions on their clients about Bitcoin-related investments (Cheng 2018); (3) High-technology firms ban all cryptocurrency-related advertisements, such as search engines (Google and Bing) and social networks (Twitter and Facebook). Moreover, based on the targets of regulations, there are three types: (1) Restrictions on Initial Coin Offer (ICO); (2) Limitations on cryptocurrency trading; (3) Bans on cryptocurrency derivatives trading.

2 The major difference between Corbet et al. (2018a) paper and our study is that we aim to examine the effect of a regulatory policy in the expanding Bitcoin futures market on its cash market rather than the impact of introducing new Bitcoin futures products.

3 We have at least two motivations to select 5-minute data frequency rather than other frequencies. First, 5-minute frequency data are available for the Bitcoin cash market. Second, 5-minute frequency data are commonly used in the previous high-frequency research, e.g., Darrat, Rahman, and Zhong (2003) and Bariviera, Zunino, and Rosso (2018). By using the 5-minute frequency data, we could compute 30-minute frequency liquidity measures.

4 On 01/12/2017, the Cboe and the CME announced to launch their own Bitcoin futures contracts in the near future (Corbet et al., 2018a). To control for the impact of this event on the Bitcoin spot variability, we create a futures announcement dummy variable which is equal to 1 on 01/12/2017 and 0 otherwise. Both futures trading and futures announcement dummy variables are added to Equations (3) and (4). Appendix B reports the estimation results. As shown, the impact of South Korea’s futures trading ban on the spot variations remains the same.