Quantitative insights into cryptocurrency markets: portfolio optimization, risk management, and volatility patterns

Abstract

This thesis examines the risk characteristics, portfolio optimization, and volatility patterns in cryptocurrency markets, providing insights into the distinctive financial dynamics of digital assets. Across its analytical chapters, this research addresses both the challenges and opportunities posed by cryptocurrencies, which are characterized by high volatility and risk factors that are uncommon in traditional markets. Chapter 3 develops a portfolio optimization framework specifically for cryptocurrency investments, showing that strategic asset allocation can enhance risk-adjusted returns despite the pronounced kurtosis and semi-heavy tails typical of cryptocurrency returns. Findings from various portfolio construction techniques highlight the need for diversification strategies that consider the unique risk profile of digital assets. Chapter 4 assesses the effectiveness of Value-at-Risk (VaR) as a risk assessment tool in cryptocurrency markets, comparing VaR models under conditions of extreme market behavior. The results indicate that traditional VaR models may underestimate risk due to the volatility and heavy-tailed return distributions of these markets, underscoring the importance of alternative approaches such as extreme value theory to capture tail risk more accurately. Chapter 5 investigates calendar anomalies, specifically the day-of-the-week effect, analyzing patterns in cryptocurrency returns that may aid strategic decision-making. Observed day-of-the-week effects were found to be sensitive to market conditions and external events, such as the COVID-19 pandemic, reinforcing the need for continual assessment of temporal trends. The consolidated analysis in Chapter 6 synthesizes insights from each chapter, outlining the distinctive attributes of cryptocurrencies as an asset class and highlighting implications for both investors and policymakers. Comparisons with traditional assets, limitations of current models, and recommendations for future research are also provided. Overall, this thesis contributes to the academic literature and practical applications by enhancing our understanding of cryptocurrency risk management and guiding more resilient portfolio strategies for this evolving asset class.