Systematic vs. Discretionary Decision-Making

Corey Hoffstein shared some of his thoughts on investment philosophy this week. He explains his preference for systematic over discretionary approaches to investing. Because “any disciplined approach can be systematized,” he believes, there’s little reason rely on a discretionary manager who could break from discipline. Systematic approaches can be back-tested and applied to hypothetical scenarios, he argues, so there is a more robust track record to analyze than with a discretionary manager.

I sympathize with this idea generally, but as always, it depends. Some type of disciplined, systematic process is crucial, but the question is the best place to draw the line.

Of course, any systematic approach or model is first created by a discretionary process. Even if the model is built with a sophisticated piece of software that can self-optimize, it was originally programmed by a human, and in most cases, humans continue to troubleshoot and monitor it. It is up to humans to decide when to modify a systematic strategy, and when to abandon one rules-based approach for another.

In the same sense, most discretionary managers incorporate systematic techniques into their process. This could be a simple stock screen or an advanced model – but if the manager makes idiosyncratic, overriding decisions, it’s a discretionary approach.

In many ways, these debates between systematic and discretionary investors are a matter of cognitive style. Different people think differently, and there are successful investors in both camps. My view is that the key is knowing yourself, your strengths, and more importantly, your weaknesses. The first section of this piece by Cliff Asness describes how smart people sometimes speak in “different languages” when it comes to investing.

Personally, I prefer a discretionary approach, using rules-based techniques as decision tools. I’m skeptical of many back-tested or stress-tested systematic approaches, but not because I think they can’t work. It’s because: (1) from my perspective, the “arms race” between quantitative strategies means any back-tested approach may be ineffective in the future. And (2) because systematic approaches can miss or misinterpret the extremes of human behavior. There are great quantitative investors, but I recognize my lack of skill in that domain.

As with many other choices, the “best” answer is probably different for different investors. As Asness writes, quantitative and qualitative investors agree on a great deal, they sometimes just speak in different languages.