Recognizing that one size does not fit all, investment managers have sought to refine their asset allocation approaches to incorporate a broader perspective. One critique of traditional optimization approaches is their sensitivity to input estimations, which can dramatically impact an investor’s portfolio. Some asset managers have sought to address this and other limitations by applying a strong understanding of investor needs within multi-objective optimization models.
Multi-objective optimization examines multiple potential portfolios along multiple dimensions, as represented by various quantitative success metrics. An optimized portfolio can collectively deliver the most exposure to desired success metrics, which, in turn, can maximize exposure to the various real-world objectives associated with those metrics. Compared to a suboptimal portfolio, an optimized portfolio could, for example, have higher expected portfolio yield, higher expected return and lower expected volatility.
We believe that by taking a broader view of an investor’s goals, multi-objective optimization can offer managers a more nuanced approach to portfolio construction that can better address diverse objectives. Our approach combines deep research into investor personas, simulation modeling and robust quantitative methods to inform portfolio design. This is one element of our overall asset allocation process.
In the paper linked below, we take a closer look at the multi-objective optimization approach.