Publication Date:
2025
Short description:
(2025). Risk Management for Cryptocurrency Portfolios . Retrieved from https://hdl.handle.net/10446/316068
abstract:
Cryptocurrencies have transformed finance by opening new avenues for investment and innovation, while exposing portfolios to extreme volatility, fat tails, liquidity shocks, and shifting regulation. Risk Management for Cryptocurrency Portfolios provides a rigorous, practice-oriented toolkit for this landscape. The book blends postmodern portfolio theory, heavy-tailed statistics, and empirically tested optimization methods into a coherent framework tailored to digital assets.
Starting from the data, the authors assemble a consistent set of 40 major tokens and examine hourly performance, stylized facts, and benchmarks. They study stationarity, the non-normal nature of returns, and tail risk using Hill estimators and generalized Pareto modeling and quantify distances between return series to guide diversification. The portfolio core begins with mean-variance analysis, the capital market line, and coherent risk measures. Building on this foundation, the book develops mean-CVaR optimization and equivalent formulations, with MATLAB implementations and step-by-step case studies.
Strategy chapters compare long-only and long-short constructions, including Jacobs et al. and Lo-Patel approaches, momentum variants, and portfolios under turnover constraints. Performance is evaluated with maximum drawdown and widely used ratios such as Sharpe, Sortino-Satchell, and the Rachev ratio.
The dynamic optimization introduces ARMA(1,1)-GARCH(1,1) models with Student's t-innovations, multivariate t-distributions and t-copulas, and the simulation of return scenarios. Robust optimization addresses model misspecification by treating observed return distributions as uncertain; readers learn box and ellipsoidal uncertainty sets, Kantorovich distances between discrete distributions, and robust CVaR portfolios on historical data.
Validation is integral. A backtesting suite consisting of value-at-risk tests, including binomial and traffic-light procedures, plus Kupiec, Christoffersen, and Haas tests, assesses model quality and contrasts historical, dynamic, and robust allocations. Written for practitioners, analysts, researchers, and graduate students, the text is selfcontained and comprehensive. Clear exposition, empirical examples, and ready to run MATLAB code make advanced methods usable in day-to-day portfolio construction. Risk Management for Cryptocurrency Portfolios equips readers with insight and tested techniques needed to build, stress-test and refine crypto portfolios with confidence.
Starting from the data, the authors assemble a consistent set of 40 major tokens and examine hourly performance, stylized facts, and benchmarks. They study stationarity, the non-normal nature of returns, and tail risk using Hill estimators and generalized Pareto modeling and quantify distances between return series to guide diversification. The portfolio core begins with mean-variance analysis, the capital market line, and coherent risk measures. Building on this foundation, the book develops mean-CVaR optimization and equivalent formulations, with MATLAB implementations and step-by-step case studies.
Strategy chapters compare long-only and long-short constructions, including Jacobs et al. and Lo-Patel approaches, momentum variants, and portfolios under turnover constraints. Performance is evaluated with maximum drawdown and widely used ratios such as Sharpe, Sortino-Satchell, and the Rachev ratio.
The dynamic optimization introduces ARMA(1,1)-GARCH(1,1) models with Student's t-innovations, multivariate t-distributions and t-copulas, and the simulation of return scenarios. Robust optimization addresses model misspecification by treating observed return distributions as uncertain; readers learn box and ellipsoidal uncertainty sets, Kantorovich distances between discrete distributions, and robust CVaR portfolios on historical data.
Validation is integral. A backtesting suite consisting of value-at-risk tests, including binomial and traffic-light procedures, plus Kupiec, Christoffersen, and Haas tests, assesses model quality and contrasts historical, dynamic, and robust allocations. Written for practitioners, analysts, researchers, and graduate students, the text is selfcontained and comprehensive. Clear exposition, empirical examples, and ready to run MATLAB code make advanced methods usable in day-to-day portfolio construction. Risk Management for Cryptocurrency Portfolios equips readers with insight and tested techniques needed to build, stress-test and refine crypto portfolios with confidence.
Iris type:
1.3.01 Monografie o trattati scientifici - Books
List of contributors: