The Link Between Decision-Making and Investment Returns

It exists — we just proved it.

By Clare Flynn Levy

It makes intuitive sense: an investor who consistently makes good investment decisions should outperform one who doesn’t. But a statistically significant link between decision-making quality and investment returns has never been established — until now.

Essentia’s research team — which, over the last decade, has produced some of the investment industry’s pioneering work on decision-making — has long been on the lookout for a formal link between decision quality and portfolio returns. Their latest research not only identifies a statistically significant relationship, but shows that an active equity fund manager who has made skilled decisions over the last year is 1.5x more likely to outperform their benchmark over the next 12 months than a manager who hasn’t.

You read that right: there is a way to tell whether a manager is more or less likely than other managers to deliver on their intentions of outperformance in the foreseeable future.

A full paper is in the works, but here is the gist.

As we have done with most of our research, we used Essentia’s proprietary database of anonymized portfolio manager decision metrics. In this case, we studied the daily holdings data of 123 long-only equity portfolios, spanning 2014 through 2023.

Applying our peer-reviewed Behavioral Alpha Benchmark methodology, we investigated seven different decision types (stock picking, sizing, entry timing, scaling in, size adjusting, scaling out and exit timing) and the value added or destroyed (as measured by the impact on returns relative to the portfolio’s benchmark) by each of these decisions.

For each portfolio, we computed a Behavioral Alpha^® Score — our proprietary measure of demonstrated investment decision-making skill — for each calendar year, each of which had to contain greater than 80 decisions to qualify. Any BA Score greater than 50 means the portfolio management team is adding value through their overall decision-making; any score less than 50 means their decision-making is, in aggregate, destroying value.

Before looking at the decision quality score’s impact on returns, we sought to validate the model by testing what we already knew: that there’s no relationship between the average fund manager’s performance vs. their benchmark in one year and their performance vs. that benchmark in the next.

And that was confirmed when we crunched the numbers: as the chart below shows, we found no relationship between a portfolio’s relative returns from one time period (t-1) to the next (t).

We next tested whether the average portfolio’s Behavioral Alpha Score — i.e., the quality of the portfolio manager’s decision-making — over a given 12-month period has a significant direct relationship to the portfolio’s relative returns over the subsequent 12 months. The answer was no: we found no relationship between a portfolio’s BA Score at time t-1 and its relative return at time t. The score itself, in other words, does not predict future returns.

That was good news from our perspective; we would have struggled to believe the opposite finding. But then things got interesting: our next step was to look at the odds of subsequent outperformance by managers with Behavioral Alpha Scores over 50 — i.e., evidence of decision-making skill — and that of managers with scores under 50.

Here, we found a powerful signal: for portfolios with a BA Score greater than 50, the odds of outperforming (their stated benchmarks) over the subsequent 12 months are 1.51 times higher than they are for portfolios with a BA Score less than 50.

Transposing this conclusion to our widely-used Behavioral Alpha Frontier diagram tells the story in very clear, simple terms: portfolios above the dashed line — which indicates where decisions are neither adding nor destroying value, compared with what would have been achieved by chance — are 1.51 times more likely to have positive relative returns than those below it.

I appreciate the magnitude of that statement, and don’t make it lightly. And as I said at the top of this piece, it is intuitive: if we measure skill based on relative P&L added or destroyed through decision-making, then it stands to reason that the skilled manager should perform better — particularly given our recent finding that investment skill itself tends to persist into the future.

This is but one early and relatively narrow finding in what’s sure to be a huge area of exploration by academics and industry practitioners alike; much more research, refinement, peer review and discussion will follow.

In the meantime, we’ve developed a new, free tool that makes it easy to identify a manager with decision-making skill in comparison with peers. The number of portfolios involved is growing daily, making it rich for exploration.

The “past performance disclaimer” is fundamental to investment industry communication: we all know that past performance does not guarantee future results (and variations thereof). This new research actually confirms that statement. But it also uncovers a valuable new way to quantitatively compare and contrast equity managers and their relative probability of outperformance.

Watch this space — we’ll have much more to share on this topic in the coming weeks and months, including a white paper with expanded analysis and commentary.

And in the meantime, if you’re an allocator who wants to use this technology to assess and compare portfolio managers, or a manager who wants to understand your own decision-making so you can improve it — and tell a richer story to allocators in the process — don’t hesitate to reach out.