Asking questions about the facts behind a recommendation can help sort out an advisor’s biases and conflicts of interest, a new study suggests.
The findings of the experimental
The two experiments that Bentley and three other researchers conducted for the paper involved YouTube videos with hidden monetary values based on the number of views rather than stocks or bonds, though. The researchers tested the influence of commissions on participating college students they designated as advisors suggesting the most valuable sets of YouTube videos for other participants.
The experiments “provide clean laboratory evidence” that face-to-face or written discussions after a recommendation “can help people distinguish the unbiased part of a report from the biased part,” Bentley and co-authors Robert Bloomfield of Cornell University, Shai Davidai of Columbia University and Melissa Ferguson of Yale University wrote in the introduction.
“Our results,” they continued, “have a straightforward practical implication: people who rely on advisors with expert and/or informed input can better identify bias when they engage in rich post-report interactions with those advisors, especially when they ask them fact-based questions.”
In designing the experiments, the researchers wanted to create a situation that had elements of objective facts and subjective interpretations of them, according to Bentley. The researchers tracked the difference between the advisors’ initial choices of video investments and the set they recommended to other participants when paid a commission based on their pick.
“In any sort of situation where someone is giving you advice, you have to figure out how real that advice is and how much they've drunk the Kool-Aid,” says Bentley.
Bentley describes the effect of the commissions given to the participants for particular recommendations as “self-deception.” The advisors convinced themselves that the video picks that paid them commissions were better than their initial selections, to the point that they recommended more choices that brought them kickbacks and later altered their own picks in favor of the ones that came with commissions, Bentley says. Questioning them about the facts revealed the “simplest tell” of bias: They deflected and didn’t answer, according to Bentley.
In the first experiment, the researchers arranged 66 pairs of designated advisors and advisees, with only half of them engaging in face-to-face discussion about the recommendations.
“We find strong evidence that advisors bias their recommendations toward their self-interest (i.e., the commission deck), and that post-report interaction significantly reduces the degree to which advisees rely on the biased part (relative to the unbiased part) of the recommendation,” the study states.
In the second experiment, the team gathered 144 recommendations and asked each advisor a “vague” question about their choice and a “fact-based” one seeking to know exactly how many videos they selected for investment from each set and why. Then, the researchers shared each of the picks with 18 advisees. One-third received both of the advisor’s answers, another third got only the response to the vague question and the last third didn’t get either of them.
“Advisees who receive both answers are better at distinguishing between the biased and unbiased parts of their advisor’s recommendation than advisees who receive only the vague answer or who don’t receive either answer,” the study says.
To explain the results, Bentley uses the example of a CEO who’s asked about a company’s earnings but responds instead with its adjusted earnings or politicians replying to the questions that they wanted to hear rather than the ones asked of them.
“The advisors who had conflicting incentives were much more likely to beat around the bush,” Bentley says. “The key is for wealth managers to be able to pick up on it. When someone doesn't answer the question, either they're trying to mislead me or perhaps they've misled themselves.”
The study is “very well done and a great contribution to the advice-giving literature,” according to Samantha Lamas, a behavioral researcher with Morningstar. Although it’s not “exactly indicative” of the financial advisor-client relationship due to several factors such as the planners’ greater expertise, higher accountability standards and motivation to help, advisors “are still subject to behavioral biases,” Lamas wrote in an email.
She recommends that clients bring a list of questions with them to meetings with advisors, such as one
“The paper had great examples of sense-making — we are prone to coming up with reasons that support what we see and/or our opinions,” Lamas says. “Advisors, as human beings, are prone to this bias as well and may not be aware of it.”
Questioning the basic facts of any recommendation could help advisors root out the bias when working with clients or fellow intermediaries in the industry. Part of the job of being an advisor is not falling for the kind of spin heard on company earnings calls, Bentley says.
“Very few people are willing to come out and say that something is white when it's black — instead they kind of talk around the issue or they deflect,” he says. “Wealth managers need to be asking the questions: ‘What are the facts here? What is Apple not telling us? How can I not fall for their spin or their self-deception?’”
