Volatility estimation for Bitcoin: Replication and robustness

Olivier Darné (1) , Amélie Charles (2)

1 LEMNA - Laboratoire d'économie et de management de Nantes Atlantique
2 Audencia Business School

Katsiampa [Volatility estimation for Bitcoin: A comparison of GARCH models. Economics Letters, 158, 3–6, 2017] compares several GARCH-type models to estimate volatility for Bitcoin returns. First, we propose a replication study (i) by verification, using the same sample and period (July 2010 to October 2016), and (ii) by reproduction, extending the sample until March 2018. We obtain only partially different results from those of Kasiampa (2017) on both samples. Second, we propose a robustness analysis (i) by reanalysis, using the robust QML estimator for computing the standard errors of the parameters, and (ii) by extension, taking into account the presence of jumps in the Bitcoin returns. The results show that the six GARCH-type models studied, namely GARCH-type models characterized by short memory, asymmetric effects, or long-run and short-run movements, seem not to be appropriate for modelling the Bitcoin returns.

Keywords: Bitcoin

Domains: Statistical Finance [q-fin.ST]; Quantitative Finance [q-fin]

Olivier Darné, Amélie Charles. Volatility estimation for Bitcoin: Replication and robustness. Économie Internationale, 2019, 157, pp.23-32. ⟨hal-01941102⟩