News

Three WorldQuant University students created a practical new way to build investment portfolios as part of their capstone project for the Master of Science in Financial Engineering program. The 2026 graduates—Micah Ondiwa, Charles Etim, and Maurice Ogedegbe—published their paper at SSRN on April 9, 2026.

Their research helps address a common problem: Many investment approaches work reasonably well in calm times but can fail to protect savings when markets suddenly drop. The team developed an intelligent system that uses multiple AI models working together to forecast future returns and risks at the same time.

Unlike traditional methods that look only at past averages, their approach adapts to changing market conditions, such as calm periods, rising markets, falling markets, or turbulent times. The research pays special attention to protecting against rare but severe losses that can hurt investors the most.

Starting with 100 different assets from stocks, bonds, commodities, real estate, and cryptocurrencies during the 11-year span from 2015 to 2025, the students selected a practical mix of 22 assets. They then tested their system on real historical data.

The results stood out. Their method delivered stronger returns for the level of risk taken, suffered smaller losses during market declines, and offered better protection against extreme downturns than conventional approaches. It performed reliably across different market environments, especially during difficult periods.

Why this matters:

Everyday investors, retirement savers, and professional managers constantly face uncertainty from market swings, economic shocks, and global events. This research shows how smarter tools can help build portfolios that pursue growth while reducing painful losses. In today’s fast-moving world, innovations like this make professional-level risk management more effective and accessible. The project highlights the real-world value of the hands-on training at WorldQuant University.

Read more:

Ondiwa, Micah, Smart Portfolio: Ensemble Conditional Risk (ECR) Markowitz (December 31, 2025). Available at SSRN: https://ssrn.com/abstract=6261138 or http://dx.doi.org/10.2139/ssrn.6261138

Photo caption:

From left, Micah Ondiwa, Charles Etim, and Maurice Ogedegbe, MScFE ’26, who published research as part of their capstone project.