Graduation Year


Date of Submission


Document Type

Campus Only Senior Thesis

Degree Name

Bachelor of Arts



Reader 1

Benjamin Gillen


This paper takes an empirical simulation-based approach to investigate the performance of a relatively new and robust algorithmic portfolio optimization strategy. Using varied tuning parameters, we were able to home in on the optimal number of assets to incorporate into a portfolio for investors seeking superior returns with minimalized risk. The results from our study confirm that there is in fact a “goldilocks” level of securities to include in solving the portfolio optimization problem which is 15 securities in small asset universes and 25 in larger universes. We will provide background on the prior research conducted on portfolio optimization problems in the field of quantitative portfolio management strategies as well as how the subset strategy employed in this analysis generally performs versus other asset allocation strategies. The research allowed us to refine the implementation of the subset algorithm to asset allocation problems which makes for more efficient and performance maximizing use of this potable portfolio optimization strategy.

This thesis is restricted to the Claremont Colleges current faculty, students, and staff.