Portfolio Optimization and Tail-Risk Analytics of Actively Managed ETFs
Year: 2026
Type: article
Abstract: This paper examines portfolio optimization and tail-risk analytics for a heterogeneous universe of 30 actively managed investment funds using daily Bloomberg data from 2020 to 2025. The study compares buy-and-hold, mean–variance, conditional value-at-risk (CVaR)-based, and tangency-type portfolio strategies under long-only and long–short constraints. The sample consists predominantly of actively managed exchange-traded funds, with PTTRX retained as an actively managed fixed-income mutual-fund comparator. The results show substantial heterogeneity across thematic equity, fixed-income, income-oriented, multi-asset, and alternative strategies, creating both diversification opportunities and meaningful differences in volatility, drawdown behavior, downside exposure, and tail risk. Historical results indicate that tangency-type portfolios are generally the strongest competitors to the buy-and-hold benchmark, while minimum-variance and CVaR-minimizing portfolios provide more conservative downside control. Dynamic allocation improves performance selectively: TVP records the strongest risk-adjusted performance, followed by TC95 and TC99, but these strategies are more sensitive to turnover, transaction costs, and implementation frictions, especially under long–short constraints. Tail-risk diagnostics based on empirical VaR, Expected Shortfall, maximum drawdown, left-tail Hill estimators, and peak-over-threshold generalized Pareto distribution (POT–GPD) methods show that downside tail exposure remains meaningful after portfolio aggregation. Overall, the findings suggest that actively managed ETFs are best evaluated as components of a joint investment opportunity set in which dependence structure, portfolio design, dynamic allocation, implementation frictions, and tail-risk exposure jointly shape performance.
DOI: 10.3390/jrfm19090665
Institutions: Texas Tech University, Kean University, Johns Hopkins University
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: hybrid
Estimated APC paid: $1,689
OpenAlex ID: W7167619795