In this episode we discuss various approaches to measuring investment risk and outline some of the pitfalls of using the traditional standard deviation of returns metric. We also examine a journal article which introduces a novel methodology for measuring portfolio tail risk by integrating multivariate extreme value theory with orthogonalized returns. The author, Miloš Božović, addresses the computational complexity of traditional risk models by using principal component analysis and GARCH filtering to transform correlated assets into independent series. These individual components are then analyzed using the generalized Pareto distribution to provide precise, closed-form estimates for Value at Risk and Expected Shortfall. Empirical testing on U.S. stock data and currency portfolios demonstrates that this approach identifies extreme market co-movements more accurately than standard parametric methods. Ultimately, the research offers a robust framework for financial institutions to manage risk during periods of significant market stress and volatility.

Sources

Kim, M., & Zhou, A. (2024). "The Measurement of Investment Risk." World Scholars Review. (Mentored by Dr. Gerard Dericks, Hawaii Pacific University.)

Božović, M. (2020). "Portfolio Tail Risk: A Multivariate Extreme Value Theory Approach." Entropy (Basel), 22(12), 1425. doi:10.3390/e22121425.

Episode Note

This episode draws on the sources listed above and incorporates AI-assisted research synthesis. All content has been reviewed and curated by the host. It is intended for educational purposes only and does not constitute investment or financial advice.