Decoding Insider Information and Other Secrets of Old School Chums
Associate Professors Lauren H. Cohen and Christopher J. Malloy study how social connections affect important decisions and, ultimately, how those connections help shape the economy. Their research shows that it's possible to make better stock picks simply by knowing whether two industry players went to the same college or university. What's more, knowing whether two congressional members share an alma mater can help predict the outcome of pending legislation on the Senate floor. Key concepts include:
- The researchers found that analysts will outperform on their stock recommendations when they have an alma mater in common with any of the recommended firm's top executives or board members.
- Portfolio managers place larger bets on a firm when an alumni connection holds a top position at that firm. These bets tend to perform especially well.
- US senators are more likely to vote in favor of bills when other senators who graduated from the same university also voted in favor of those bills.
An old adage says that it's not what you know, it's whom you know. But outsiders can take heart: even for those who don't belong to a high-power social network, there's power in simply keeping track of who went to school with whom.
Associate Professors Lauren H. Cohen and Christopher J. Malloy have spent several years unlocking the power of school ties, studying how these social connections affect important decisions and, ultimately, help shape the economy. Their research shows that it's possible to make better stock picks simply by knowing whether two industry players went to college together. What's more, knowing whether two congressional members went to the same college can help predict the outcome of pending legislation on the Senate floor.
"And so we got to thinking, if a school connection makes information flow so easily in informal situations, this might be true in market situations as well."
The decision to study alumni connections was borne of the professors' personal school ties: both received their PhDs in finance from the University of Chicago. In 2007, each landed a position in the Finance Unit at HBS, where they teamed up to study issues related to empirical asset pricing. Hedge fund firms often hired the duo for speaking engagements, which led to informal post-speech chats with various fund managers, many of whom also sported PhDs. Cohen and Malloy found that these social exchanges went especially smoothly when it turned out that the hedge fund manager had gone to UChicago, too.
"We'd start talking about where to eat in Hyde Park, or about the school's PhD lab, and we'd go from there," Cohen says. "It seemed to grease the wheels for a good conversation and just made the whole thing more comfortable. And so we got to thinking, if a school connection makes information flow so easily in informal situations, this might be true in market situations as well."
Sell side school ties
In their paper Sell Side School Ties, cowritten with Andrea Frazzini, Cohen and Malloy look at school ties between sell-side analysts and the senior officers at the firms they cover, hypothesizing that analysts gain comparatively good information if they attended the same college as a firm's top executives or board members.
In short, the researchers found that the analysts outperformed on their stock recommendations when they had an alma mater in common with any of a firm's top brass. In fact, analysts' "buy" recommendations on a school-tied stock outperformed their "buy" recommendations on a non-school-tied stock by an average of 45 to 55 basis points per month or 5.4 to 6.6 percent per year.
"Literally, if an analyst who went to Princeton covers two stocks, a Harvard-tied stock and a Princeton-tied stock, and issues a 'strong buy' recommendation on both the Harvard stock and the Princeton stock, his Princeton stock will on average significantly outperform his Harvard stock," Cohen says.
This held true even when the researchers disregarded those analysts from Ivy League and other top-tier schools, meaning the results were not tied to an analyst being unusually smart or privileged.
"It's not about which school they attended, specifically," Malloy says. "It's just about utilizing the alumni network no matter where they went."
The researchers, however, did not find any significant school-tie effects when it came to sell recommendations.
"One explanation consistent with this finding is that managers are willing to reveal positive (but not negative) information about their firms," the paper states. "Alternatively, this would be consistent with analysts obtaining both good and bad news from their school-tied firms, but perhaps as a tacit agreement, acting only on the positive news."
Importantly, the research takes into account stock returns in the years before and after the October 2000 adoption Regulation FD (Fair Disclosure), in which the Security and Exchange Commission mandated that publicly traded companies had to disclose all material information to their investors at the same time. The researchers found that after the introduction of that rule, the return spread of school-tied and non-school-tied "buy" recommendations virtually disappeared. In the United Kingdom, however, where no such rule exists, the effect of school ties on buy recommendations stayed markedly positive.
"This provides more evidence that the alumni channel is really about information," Malloy says.
The small world of investing
In earlier research, published in the paper The Small World of Investing: Board Connections and Mutual Fund Returns, the three scholars found that portfolio managers placed larger bets on firms when there was an alumni connection, and that these bets performed especially well.
The paper examines portfolio choices according to various levels of alumni connections—from the vague connection of just having attended the same school to the strong connection of having attended the same school in the same degree program at the same time.
"You can use that information to predict how people are going to vote on an upcoming bill, based on their past favor-trading voting behavior."
Indeed, the performance of the stock increased according to the strength of the connection. In cases where the connection was simply a common alma mater, the researchers found that purchases of connected stocks outperformed purchases of non-connected stocks by 2.69 percent annually. But in cases where the portfolio manager actually attended school with the senior officer of a firm, working toward the same major, the connected stocks outperformed non-connected stocks by a whopping 9.04 percent annually.
Predicting congressional votes
Recently, Cohen and Malloy have turned their attention to the federal legislative process in the United States, researching how school ties influence logrolling—the quid pro quo process in which members of Congress trade votes to achieve mutual gain.
"A logging bill that may be really important to a senator in Vermont may not be that important to a senator in Kansas, where there are fewer forests," Cohen explains. "If they can do some kind of vote trading between them on that logging bill, then the Vermont senator can do the Kansas senator a favor later by voting on a bill about wheat subsidies. You vote for the bill that helps me but doesn't affect you, and then I'll vote for the bill that helps you but doesn't affect me."
In the paper Friends in High Places Cohen and Malloy studied the congressional voting record from 1989 to 2008 and found that senators were more likely to vote in favor of bills when other senators who shared the same alma mater also voted in favor of those bills.
"After party and state, this seems to be among the most reliable determinants of the way that senators vote," Cohen says.
Since most members of Congress list their alma maters on their official websites, and all congressional votes are a matter of public record, careful observers can create a virtual crystal ball of sorts by taking note of where politicians went to college, along with their respective voting records.
"You can use that information to predict how people are going to vote on an upcoming bill, based on their past favor-trading voting behavior," Malloy says. "So now you know where they stand better than others who aren't looking at these patterns."
The researchers plan to continue investigating the effects of social networking on congressional voting behavior, looking at other network connections such as former places of employment. Their forthcoming paper, "Trading Favors," due for release in the fall, should be of great interest to anyone who deals with the stock market.
"Legislation has a massive effect on firm value," Cohen says. "So if you're able to predict the outcome of a legislative vote about energy, which might affect the future of an oil company, for example, then you're in a great place as an investor."