This content is from: Portfolio

“I Sell My Winners Too Early and Hold Losers.”

How behavioral analytics can help active investors from repeatedly making the same mistakes.

Thirty years of research in behavioral science has taught us how human brains are wired and how emotions drive decisions. Still, active portfolio managers — even though they are in a vicious war with index funds — are frittering away their excess returns by making predictable mistakes.

“Information, and even computation, these days are commodities. Our value has to be in our critical thinking and human intuition,” said Brandon Snow, principal and CIO of $20 billion-in-assets Cambridge Global Asset Management, speaking at Essentia Analytics’ Behavioral Alpha conference in New York on Friday. 

Cambridge has created systems to track decisions and help its fund managers avoid well-known behavioral mistakes, such as holding on to losing stocks too long.

“I sell my winners too early and hold losers. That has important outcomes,” said Snow. “My group saw a similar pattern,” he added. Now that it has almost three years of robust data on its team, Cambridge is creating a buddy system and pairing fund managers who are lacking a skill — like profitable selling behavior — with somebody else with that specific talent. Cambridge is also beginning to share data it collects on behavior with consultants and other gatekeepers, demonstrating that they actually follow their investment process. “They’ll see all of our decisions,” he added.

[II Deep Dive: Active Equity Managers Actually Do Generate Alpha. Here’s How They Squander It.]

To capitalize on decades of behavioral research findings, some managers are collecting as much data as possible on their investors’ daily decisions. Some people bristle at the intrusions.

“We use that as a screening tool. If people aren’t willing to do this, we don’t want you,” said Tom Tully, portfolio Manager at Aperture Investors, at the conference. Tully added that an active manager needs to be rigorous in preserving the alpha it creates. That’s because Aperture, founded by Peter Kraus, former CEO of AllianceBernstein, charges management fees comparable to index funds. Clients then pay an incentive fee if the manager can generate alpha.

Aperture is just getting started designing systems so it can tag every piece of data about its portfolio managers’ decision-making. “Instead of just saying it, we can prove we are data driven,” said Tully.

Cambridge collects daily information from portfolio managers and analysts, including how much sleep they had, exercise, their personal life, and whether they were happy or sad. The firm is looking for patterns. On a bi-weekly basis, it reviews best and worst ideas and goes over questions such as what the manager would sell if he or she had to. 

Snow said that when he looked at his own trading activity, he found that 40 percent of his decisions on small positions and small stocks added no value. “That’s 40 percent of my brain power,” he stressed.

Related Content