Quantitative Modeling - How to Avoid the Inherent Man Bias

In this episode I speak with Jim Masturzo, Head of Asset Allocation at Research Affiliates.

In his role, Jim oversees the research and publication of the firm’s capital market assumptions as well as the implementation of those views into a suite of tactical portfolios.

We begin our conversation discussing the foundational assumptions behind the capital market assumptions. Like most firms, Research Affiliates takes a long-term view on return and risk. In line with the firm’s guiding philosophy, they also introduce long-term mean reversionary effects.

Not surprisingly, these assumptions have been relatively bearish on U.S. equity returns for a large part of the last decade, and we discuss how to view the dispersion between these model forecasts and realized results.

We then shift our conversation to the application of tactical views. With capital market assumptions serving as the strategic backbone, Jim and his team develop a number of regime-based model portfolios that can be blended to express different tactical views.

But the team does not take a purely quantitative approach. Jim proactively acknowledges and seeks out model blindness. Rather than try to force idiosyncratic fixes into the models that might bias results, however, he and his team adopt qualitative trades to adapt the portfolios.

From strategic to tactical and quantitative to qualitative, this is a wide ranging conversation all about asset allocation. I hope you enjoy.