Modelling the price dynamics of cryptocurrencies with GARCH and SVCJ models
Cryptocurrencies (CCs) have gained more and more attention in recent years as a new asset class. Its decentralised feature brings enormous profits to buyers with a great amount of risk. As the CC market develops and takes in more and more institutional investors, it is essential to understand the pr...
Saved in:
| 主要作者: | |
|---|---|
| 其他作者: | |
| 格式: | Final Year Project |
| 語言: | English |
| 出版: |
Nanyang Technological University
2022
|
| 主題: | |
| 在線閱讀: | https://hdl.handle.net/10356/156795 |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 機構: | Nanyang Technological University |
| 語言: | English |
| 總結: | Cryptocurrencies (CCs) have gained more and more attention in recent years as a new asset class. Its decentralised feature brings enormous profits to buyers with a great amount of risk. As the CC market develops and takes in more and more institutional investors, it is essential to understand the price dynamics of cryptocurrencies for managing risks as well as pricing derivatives. In this paper, I examine the key characteristics of cryptocurrencies in their return and volatility process, especially time-varying volatility, leverage effect and jump discontinuities. Generalised Autoregressive Conditional Heteroskedasticity (GARCH)-type models and Stochastic Volatility with Correlated Jumps (SVCJ) model are used to capture the price dynamics of three popular CCs. The in-sample estimation shows SVCJ is superior in fitting the price dynamics and the out-of-sample backtesting results show that both models can produce accurate Value-at-Risk forecasts. The results imply that SVCJ is a better model in capturing the CC price dynamics especially when the tail risks need to be taken care of. |
|---|