CEU Electronic Theses and Dissertations, 2025
| Author | Nassylkhan, Adiletkhan |
|---|---|
| Title | Simple vs. Sophisticated Portfolio Optimization: U.S. Retail Level Portfolio Study Case |
| Summary | Over the past decade, exchange-traded funds (ETFs) have revolutionized retail investing by providing low-cost access to diversified portfolios that typically track an index or sector rather than a single stock. This shift raises the question of how portfolio optimization methods perform when applied solely to ETF-based, multi-asset class portfolios. This research evaluates the out-of-sample performance of five advanced portfolio optimization methods relative to the naïve equal-weighted (1/N) rule from the perspective of a retail investor. The analysis is based on a portfolio comprising 30 U.S.-listed ETFs that represent both traditional and alternative asset classes. A 252-day rolling-window simulation is used to generate daily out-of-sample performance data from March 30, 2009, to May 7, 2025, with daily rebalancing. I test the optimization methods using two approaches. First, I apply optimization methods at the whole portfolio level without any constraints on asset class exposure. Second, I impose asset-class-specific weight constraints and perform optimization within each asset class. The results show that in the first case, complex optimizers often produce portfolios heavily concentrated in short-term Treasuries, allowing the naïve strategy to outperform substantially. In the second case, performance improves as optimizers operate within asset classes that share similar characteristics; however, this improvement still does not substantially outperform the naïve approach. Overall, the equal-weighted strategy consistently matches or outperforms the complex methods in terms of Sharpe ratio, turnover, and computation time. These findings suggest that for retail investors, the theoretical benefits of complex optimization are often outweighed by estimation error, high turnover, and computational cost. |
| Supervisor | Lee, Tomy |
| Department | Economics MA |
| Full text | https://www.etd.ceu.edu/2025/nassylkhanadiletkhan.pdf |
Visit the CEU Library.
© 2007-2025, Central European University