比特币市场风险度量——基于GARCH-t模型与GEV模型的比较研究

A Comparative Study of the RiskMeasurement of the Bitcoin Market Based on the GARCH-T Model and the GEV Model

选取2012年1月1日至2016年12月30日以美元标价的比特币市场数据,对其市场风险进行实证研究,发现收益率序列具有尖峰厚尾、波动集聚等特征。并利用GARCH-t模型与GEV模型计算VaR,发现GEV模型相比GARCH模型能更好的捕捉比特币的尾部特征,用GEV模型拟合收益率尾部分布可以更精确地度量比特币市场风险。

This paper makes an empirical study on the simple return rate sequence of the special currency with the market data from January 1, 2012 to December 30, 2016, and finds that the sequence has the characteristics of peak, thick tail and fluctuation agglomeration. Using GARCH-t model and EVT model to calculate VaR, it is found that EVT model can better capture the tail characteristics of bitcoin compared with GARCH model, and EVT model can be used to fit the tail distribution of yield, which can more accurately measure bitcoin market risk.