Understanding temporal dynamics of jumps in cryptocurrency markets: evidence from tick-by-tick data

Authors:

Abstract
Cryptocurrency markets have recently attracted significant attention due to their potential for high returns; however, their underlying dynamics, especially those concerning price jumps, continue to be explored. Building on previous research, this study examines the presence and clustering of jumps in an extensive tick data set covering six major cryptocurrencies traded against Tether on seven leading exchanges worldwide over nearly 2.5 years. Our analysis reveals that jumps occur on up to 58% of trading days, with negative jumps predominating in both frequency and size. Notably, we observe systematic clustering of jumps over time, especially in Bitcoin and Ethereum, indicating interconnected market dynamics and potential predictive power for market movements. By employing high-frequency econometric tools, we identify temporal patterns in jump occurrence, highlighting heightened activity during specific trading hours and days. We also find evidence of jumps influencing intraday returns, underscoring their significance in short-term price dynamics. Our findings enhance understanding of the cryptocurrency market microstructure and offer insights for risk management and predictive modeling strategies. Nevertheless, further research is needed to develop robust methodologies for detecting and analyzing co-jumps across multiple assets.
Keywords
Funding Information
  • Deutsche Forschungsgemeinschaft (IRTG 1792, IRTG 1792)
  • Horizon 2020 Framework Programme (825215, Topic ICT-35-2018, Type of actions: CSA)
  • European Cooperation in Science and Technology (CA19130)
  • Grantová Agentura České Republiky (19-28231X, CAS: XDA 23020303)
  • Humboldt-Universität zu Berlin