The Research Behavior of Professional Investors

How do professional investors discover new ideas, investigate them (and the larger body of old ideas), and make decisions?

Tracking (lay) investor behavior

The title of this posting was mostly cribbed from a paper by Toomas Laarits and Jeffrey Wurgler, “The Research Behavior of Individual Investors.”  The authors tracked the online actions of lay investors, “including how much time they spend on stock research, which stocks they research, what categories of information they seek, and when they gather information relative to events and trades.”

To do so, they accessed clickstream data from 2007 — before the financial crisis and well in advance of the meme-stock frenzy — so it’s appropriate to consider how relevant the findings are to today’s world.  But the conclusions of the paper ring true:  in general, not much time was spent researching ideas before trades were made and decisions didn’t appear to follow classic models of financial behavior.  Surprise, surprise.

As noted by the authors, analyzing the data is not easy:

As others have pointed out, one challenge for this literature is going beyond associations and observing, at an investor level, the full line from attention to action.  The literature is also piecemeal in terms of focusing on particular determinants of attention.

There are a variety of impediments to the few studies of this type that have been conducted.  Without eye-tracking software, you can’t tell where on a page a user is looking.  And, in any case, you “cannot observe an investor’s accumulated ‘stock’ of knowledge as opposed to the flow of information exhibited by the clickstream.”

When it comes to the level of activity of investors, the authors find “a strong differentiator of research behavior.”  The takeaway highlighted most reviews of the paper:  the median time spent on research before trading was six minutes, although the average was a half hour.

Also evident was investor segmentation:

[The analysis] contrasts investors who focus on earnings, dividends, and other slow-moving fundamentals versus those who focus on news, message boards, brief summary statistics, and price charts.  The latter investors also concentrate their research in speculative stocks.

Overall, “there is a large amount of unexplained heterogeneity in research behavior.”

Tracking (professional) investor behavior

Reading that paper, it’s impossible not to think of how a similar review of the behavior of professional investors would look.  For example, since the paper tracked individual investors buying and selling stocks, imagine these questions as you might reword them to apply to the portfolio manager of an equity fund:

How much time do individual investors spend on stock research?  Which sites do they use?  Which stocks do they focus on?  When do they do their research relative to their trades or corporate events?  And, perhaps most importantly, what types of information do individual investors care about — and what do they ignore?

You can add a host of other questions in that vein.

For someone outside an organization, trying to divine the answers to any of them is very difficult.  The stylized diagrams of investment process don’t address even the basics of behavior — and most due diligence efforts are spent on consuming general descriptions of activity rather than discovering the nitty-gritty of how things are done and why.

But within an organization, the possibilities open up wide.  At one level, technologies now exist that can paint a much more finely detailed picture of “the research behavior” of the people involved.  As a 2022 posting on this site described, there are complicated issues of privacy and transparency to consider when deciding whether and how to track the behavior of employees.  The benefits from improvements gained through those observations must be weighed against the cultural impact and nth-order effects involved.

Many of the challenges mentioned in the individual investor study apply here, including the difficulty of seeing “the full line from attention to action” and understanding the interplay of what is already known with the diet of new information.  Also, the sources of ideas and data are magnitudes greater than those used by individual investors.

The imitation game

While the example provided above was of an equity portfolio manager, the concept floated here isn’t limited by investment role or type of organization.  Understanding how decisions are made (not how they are said to be made) should be a top priority, while realizing that the “research behavior of professional investors” is hard to pin down.

There is a new urgency for evaluating current behavior in detail:  the prospect of reworking investment processes using artificial intelligence tools and agents.

Alan Turing conceived of “the imitation game,” now more commonly known as the Turing test, to judge how well a machine can imitate human capabilities.  At one level, analyzing existing investment processes means identifying the elements of them that might be replicated by computing power (so those steps can be done faster and at a lower cost, as long as accuracy is not sacrificed).

That imitation — the swapping of human for machine — is a limited form of improvement, although it may be a realistic first move.  The greater mission is coming up with new ways of working that exceed the old, whether they include an extensive use of artificial intelligence or not.

That requires research, reason, and creativity.  And risk taking.

Is your organization up to the task of the day?

 

If you would like to discuss the possibilities, please get in touch.