Bitcoin Price Risk—A Durations Perspective

An important aspect of liquidity is price risk, i.e., the risk that a small transaction leads to a large price change. This usually happens in a thin market, when trading opportunities are scarce and the time between subsequent trades is long. We rely on an autoregressive conditional duration model to extract the probability of a substantial price event in a particular time interval and, thus, an intraday risk profile. Our findings show that price risk is highest at times when European and U.S. investors do not trade. In a second step, we relate daily aggregates to characteristics of the Bitcoin blockchain and investigate whether investors account for features like confirmation time or fees when timing their orders.

A General Family of Autoregressive Conditional Duration Models Applied to High-Frequency Financial Data

In this paper, we propose a general family of Birnbaum–Saunders autoregressive conditional duration (BS-ACD) models based on generalized Birnbaum–Saunders (GBS) distributions, denoted by GBS-ACD. We further generalize these GBS-ACD models by using a Box-Cox transformation with a shape parameter λ to the conditional median dynamics and an asymmetric response to shocks; this is denoted by GBS-AACD. We then carry out a Monte Carlo simulation study to evaluate the performance of the GBS-ACD models. Finally, an illustration of the proposed models is made by using New York stock exchange (NYSE) transaction data.