Risk and Uncertainty in Cryptocurrency Markets

Alsamaani, Abdulrahman Abdulkarim S. (2023) Risk and Uncertainty in Cryptocurrency Markets. Doctoral thesis, University of East Anglia.

Preview

PDF Download (21MB) | Preview

Abstract

This dissertation encompasses three distinct studies, each with specific objectives. The first study explores the most effective approach for predicting the volatility of cryptocurrency returns using high-frequency data. It examines both prominent and lesser-known cryptocurrencies through various models, including GARCH, IGARCH, EGARCH, GJR-GARCH, HAR, and LRE, employing both univariate and comprehensive regression. The findings reveal that the HAR model is superior for one-day forecasts, while the EGARCH model excels for seven- and thirty-day forecasts. Additionally, the HAR + EGARCH combination outperforms other model pairs across all time frames. However, out-of-sample analysis presents mixed results, offering valuable insights for investors, portfolio managers, and financial professionals.

The second study investigates the relationship between cryptocurrency returns and uncertainty indices, particularly during the Covid-19 pandemic. It identifies which indices most significantly affect cryptocurrency market returns and the impact of index pairs on returns. Ten cryptocurrency returns and eight uncertainty indices are analyzed using Quantile Regression, Multivariate-Quantile Regression, and Granger Causality tests. Results indicate that the Cryptocurrency Policy Uncertainty index and the Cryptocurrency Price Uncertainty index considerably impact cryptocurrency returns, while the other indices have no influence on cryptocurrency returns. The Multivariate-Quantile Regression findings demonstrated that when the cryptocurrency market experiences a bull wave, the UCRY Policy Index + Central Bank Digital Currency Attention Index combination strongly impacts cryptocurrency returns. Nonetheless, when the cryptocurrency market experiences a bear wave, the UCRY Policy Index and the Cryptocurrency Environmental Attention (ICEA) index combination considerably impact cryptocurrency gains. During the crisis, most of the overall sample findings were verified. These insights will benefit investors, portfolio managers, and policymakers.

The third study aims to identify the best model to forecast the covariance matrix of cryptocurrency returns. It thoroughly evaluates five models: BEKK, Diagonal BEKK, DCC, Asymmetric DCC, and LRE, using three key criteria—Euclidean Distance (LE), Frobenius Distance (LF), and the Multivariate Quasi-Likelihood Loss Function (LQ). The LRE model emerges as the most accurate model to forecast the covariance matrix of cryptocurrency returns for daily and weekly predictions. The validation through Mean Squared Error (MSE) and Mean Absolute Error (MAE) loss functions supports these findings, except for LQ. These results have significant implications for investors and portfolio managers seeking to improve risk management strategies, enabling them to make better-informed decisions to mitigate portfolio risk.

Item Type:	Thesis (Doctoral)
Faculty \ School:	Faculty of Social Sciences > Norwich Business School
Depositing User:	Chris White
Date Deposited:	12 Aug 2024 07:29
Last Modified:	12 Aug 2024 07:29
URI:	https://ueaeprints.uea.ac.uk/id/eprint/96192
DOI:

Downloads

Actions (login required)

View Item