Newsroom

It’s no secret that incentives shape individual behaviors, but the challenge is to accurately predict the specific behaviors that may arise from a particular incentive. Researchers at the UCI Paul Merage School of Business have been exploring how incentives influence financial analysts. What they found raises a few red flags both for investors and for the industry at large.

UCI professors Devin Shanthikumar and Ben Lourie, along with Ph.D. alumna Tiana Lehmer, published their findings in the Review of Accounting Studies. Their article, “Brokerage Trading Volume and Analysts’ Earning Forecast: A Conflict of Interest?” explores the influence of incentives on the outcome of financial analysts’ research.

“The big picture is that incentives matter,” says Shanthikumar, “but it isn’t always obvious what effects incentives will have.”

Incentives Matter

For Shanthikumar, the goal of the research was to develop a better understanding of how incentives influence the brokerage industry. In the 2000s, she researched conflicts of interest in the investment banking industry during a time of intense regulatory interest in the topic.

The 2003 case of star analyst Henry Blodget was especially influential in bringing the issue of incentives to the forefront. Blodget, then working as a researcher for Merrill Lynch, was charged by the Securities and Exchange Commission with civil securities fraud after he was found to have publicly issue buy recommendations for certain tech stocks while expressing negative views of those same stocks in private emails. The lawsuit led to significant media attention and rule changes that broke the link between analyst research and investment banking.

The resulting separation of research from brokering removed the incentive structure which tied analyst compensation with investment banking outcomes. However, Shanthikumar says, the new rules only addressed one incentive problem. “Now the brokerage trading incentive will be that much more important to study,” she says.

Gathering Insight from New Data

Shanthikumar has been interested in studying how brokerage trading and incentives influence analyst behaviors for many years. The challenge has always been accessing the right data. When new data became available, the team recognized they had an opportunity to gain new insights into the link between incentives and analyst behaviors.

“All of this activity is happening inside the brokerage houses, and they don’t want to share all of that information,” she says. “We were able to get data that allows us, with a little creative research design, to figure out how to answer this question.”

Shanthikumar and her colleagues looked at data for each brokerage house and measured how much trading volume they executed in a particular stock. The team compared trade volumes at different brokerage houses while also looking at which analysts issued forecasts, and whether the forecasts were optimistic or not.

This new comparative approach was a huge first step, but the team still needed to work out whether the numbers firmly showed causation in the correlations between trading volumes and analyst reporting. To resolve the question of causation, they looked at how analysts changed over time, including when they moved from one firm to another. “Are analysts learning from what happened last year and updating their behaviors the next year? Using that data and using both correlations and exploiting data related to analyst changes we were able to better understand what was going on,” Shanthikumar says.

Linking analyst compensation to brokerage profits

Analysts working in-house at brokerages are paid to produce research and reports. While customers benefit from the insights reports provide, regardless of the analyst’s final recommendation, the concern is that the brokerage as a whole stands to benefit from recommendations that could drive trading volumes.

“It’s a little bit like being in customer service,” Shanthikumar says. “Analysts don’t directly bring in revenue to the company. They’re being paid to help the company by performing a service to investors. That’s one of the reasons why analysts’ incentives can be harder to understand.”

Simply put, analysts want to support the trading business of their brokerage house. The intended effect of incentives is to encourage analysts to provide helpful research data to the investors who are trading with the firm so that they make better trades and everyone wins. But is it really that simple?

“We do find evidence for the intended effects of incentives in our research,” Shanthikumar says. “They drive more volume for the brokerage house. So, volume is sensitive to the quality of the analyst. If they issue better earnings forecasts, or if they’re more timely, those things really do impact the trade volumes.” In short, higher quality research leads to more trading volume, providing an incentive for quality – the intended effect.

Uncovering unintended consequences

Shanthikumar and her colleagues were especially interested to see if analyst incentives had any unintended consequences. “We did find evidence for a few unintended effects,” Shanthikumar says. “The incentives lead analysts to issue more positive or optimistic forecasts. The main reason for that is that most of us can buy stock, but you can’t sell a stock unless you own it. If you issue a positive forecast, you generate more trades.”

Another unintended consequence was the analyst’s response to the optimism incentive. We found a stronger response to that optimism incentive than the accuracy incentive,” she says. “At least in the short term. Maybe over the course of their careers, that accuracy and quality incentive is going to help things, but on the one or two-year horizon what we see is that analysts are becoming more optimistic when they see that it works.”

Exploring the optimism effect

While the research shows that trading volume responds to analyst quality, volume also responds to optimism. “That is the thing we need to address,” she says. “It can drive the analysts to be too optimistic and it can lead us as investors to buy too much. This is a sign that the finance industry needs to reevaluate their incentives. Europe has already started to change practices by requiring direct payment for analyst research. For us as investors, this means we need to take that very positive research with a little grain of salt.”

One surprising insight from the team’s research came from looking at what happened when an analyst moved from one brokerage house to another. “What we found was that the trading volume followed them,” Shanthikumar says. “If a brokerage house got a high-volume analyst from another firm, they could increase their volume on those stocks that the analyst covers by as much as twenty percent. That’s a huge effect.”

The usefulness of analyst research

While they found strong evidence that analysts are strategically updating their behavior in response to these incentives and becoming overly optimistic, Shanthikumar wants to be careful not to paint with a broad brush. “I don’t want it to sound like we’re saying that analysts are just lying to us or that we can’t trust them,” she says. “Analysts are simply responding to the incentives they’re given. They’re also compensated for accuracy, so that can provide some balance to the equation as well.”

Shanthikumar has advice for current investors and for students who want to work in the finance industry. “For people who are going into wealth management or investment management, they need to understand how to use information in a sophisticated way,” she says. “Part of what that means is to adjust for the potential biases created by these types of incentives. We don’t want to ignore analysts. They have a lot of great insight and information that we can all learn from. But in light of our research, we need to make the appropriate adjustments.”

Take-aways for investors and the industry

Shanthikumar feels that their report has implications for pretty much everyone. “I teach MBAs and undergraduates to really be careful about incentives,” she says. “The incentives they design for their employees, the incentives their customers have, that their suppliers have, they all impact our behaviors. As parents, we design incentives for our kids to try to drive their behavior. We know how our incentives at work drive our own behaviors. But we need to be thoughtful about the intended and the unintended effects.”

The bottom line is that the financial industry needs to think about the role they want analysts to play. “What’s happening right now is a lot of experimentation,” Shanthikumar says. “In Europe, these kinds of incentives are no longer allowed. In the US we’re seeing the rise of these independent research firms that aren’t brokerage houses. They’re just doing the research. But how do they sell that research? Some are funded by the people using the research, others are funded by the companies that want coverage. Each of these is going to come with their own set of incentives and problems.”

Shanthikumar is optimistic about the future. “I don’t think we’re going to see legislation to step in and stop this. I don’t think the problem is that bad at this point,” she says. “We just need to be more sophisticated about how we respond to the research. We need to be aware of these conflicts of interest. What I’m hoping is that people might start to be even more sophisticated about other conflicts of interest that we may not have the research on yet.”