Analysis of Cryptocurrency Jumps using Post-Double-Selection HAR-J models

Using a novel PDS-HAR-J approach, and nonparametrically-estimated jumps and volatility series are derived from financial diffusion processes, I find evidence that cryptocurrency markets exhibit price jumps which are followed by decreased volatility. The results are an important motivation for future option-pricing model development to include jumps and for future applications of honest inference for HAR-type models.