Is investing in foreign stocks still a good strategy for offsetting risk and boosting returns in your portfolio? How do social comparisons impact the different dimensions of trust that people can have for each other? How can companies use emotional cues to convey a particular identity to consumers? Wharton professors Karen Lewis,Maurice Schweitzer and Patti Williams, respectively, examined these issues — and what they mean for business and consumers — in recent research papers.
The Changing Value Proposition of Foreign Diversification
When news heard on side streets in Cairo or Panama City can reach New York, Paris or Beijing in the next instant thanks to the Internet and other technology, the world shrinks in a very real way. Distant markets that once operated subject only to local factors now form a global economic fabric. That may sound like good news to someone who applauds the efficient flow of capital across international borders. However, forces making the world smaller challenge investment strategies that depend on global diversification to offset risk and boost returns.
Wharton finance professor Karen Lewis and Sandy Lai, a professor at the School of Economics at Hong Kong University, have studied whether an international approach to investing has lost clout. In a recent paper, “Are the Gains from Foreign Diversification Diminishing? Assessing the Impact with Cross Listed Stocks,” they detail findings that could shift the ground rules for global investors.
During the 1980s, investment in foreign securities provided an effective hedge because markets moved subject to local economic climates. Beginning in the 1990s, however, as globalization took root, foreign equity markets started to move in closer step with markets in the U.S.
“There has been this sense in both academic and in more practitioner circles that the benefits of international diversification have been declining over time as things become more correlated,” says Lewis. Evidence from the paper, she notes, corresponds to existing clues that the world is growing less diverse for investors. “It’s a myth to think each equity market around the world is an island unto itself.”
Yet it also sounds strange to suppose that a cross listing, in and of itself, alters value. “Increasing [value] upon cross listing is one of those things that could cause you to scratch your head and ask why,” Lewis says. “It shouldn’t matter if you have a stock listed in one market and take it to another.”
But it does matter, as shown in the data that Lewis and Lai compiled. They measured effects for typical U.S. investors who want to reduce variability in their equity portfolios using cross-listed foreign equities. Their universe of cross-listed stocks comprised 380 companies from 39 foreign countries on the New York Stock Exchange and 196 foreign companies from 28 foreign countries on NASDAQ between 1970, or the earliest date available, and 2009. The two exchanges are typically the most accessible to ordinary investors.
The co-authors posed two overarching questions: Does increased correlation significantly reduce international diversification gains? And do foreign stocks add diversification in the first place because they are less correlated with the U.S. market or because they merely mimic the behavior of other foreign stocks already listed?
“We indeed found evidence of changes in the relationship between the U.S. market and foreign companies,” the authors write. Change occurs on two levels: foreign market indices narrowing the gap with the U.S. market, and also in the behavior of individual cross-listed companies.
As foreign market indices increasingly resemble their counterparts in the U.S., their market betas — the measure of an asset’s risk in relation to the market — approach 1.0. Foreign betas more than doubled over sample intervals, to 0.44 from 0.20. Cross-listed stocks march in even tighter step with the U.S. For a market-weighted portfolio of cross-listed stocks, the combined beta zooms to 0.8 from 0.35 in just one interval.
In the throes of a global financial crisis, the trend accelerated, the authors note. In fact, the only means to reverse it required risk outside the comfort zone for most investors. “The portfolio variance would decline only if the foreign stocks could be shorted,” the research shows.
Across all geographic areas, the pattern repeats itself — but that’s not to say that if you’ve seen one cross-listed stock, you’ve seen them all. “After cross listing,” says Lewis, “foreign stocks behave more like their counterparts in similar industries in the U.S. markets. They can still behave quite differently from each other.” In other words, she notes, a Peruvian mining cross-listed stock looks very different from a cross-listed Spanish banking stock.
Global investors who seek an edge will find more diversification by investing in foreign securities, provided that they can live with higher transaction fees and currency risks, Lewis notes. According to Lewis and Lai, “foreign companies that have not yet listed in the U.S. provide a greater risk reduction by as much as three times that of the cross-listed stocks over some periods.”
Lewis adds a note of caution on the limits to diversification. “Portfolio managers I know want to keep an eye on diversification,” she says. “But mostly, they’re after alpha, the return above and beyond what’s implied by the beta. They’re not going to make money if [they are] diversified. They make money by taking risk.”
Why Ranking Employees Could Harm Trust Among Co-workers
How much do you trust the colleague in the next cubicle? Would you trust him or her to put the finishing touches on a presentation you started before sending it to your boss? What about trusting him or her to keep a shocking secret about another co-worker?
Just because you might trust a colleague to do the former, doesn’t necessarily mean you would have faith in them when it comes to the latter, according to Wharton operations and information management professor Maurice Schweitzer. Schweitzer notes that trusting someone’s competence at work is different than feeling comfortable with that individual as a personal confidant. “Most people perceive trust in a uni-dimensional way — more trust or less trust,” Schweitzer points out. “But it’s much more nuanced than that.”
To highlight these different dimensions and how they can move in opposite directions, Schweitzer and his colleagues — Michigan State University professor Jennifer Dunn and University of Washington professor Nicole Ruedy — examined trust in the context of social comparisons in the recent paper, “It Hurts Both Ways: How Social Comparisons Harm Affective and Cognitive Trust,” published in the journal Organizational Behavior and Human Decision Processes.
The paper defines two distinct types of trust. Affective trust is the “willingness to be vulnerable” to someone else based primarily on the emotional bond between the two people. Cognitive trust stems from an individual’s confidence in another person’s ability and integrity.
“Our findings challenge the assumption that higher levels of ability increase trust,” the researchers write. “High performance among peers can harm affective trust without benefiting cognitive trust. Conversely, underperforming peers instill low cognitive trust but fail to boost affective trust.”
The researchers tested this construct through a series of social comparison studies. The first one looked at how well 172 college students trusted close friends. Students were asked to describe and compare their own job prospects with that of a friend. They also answered several questions designed to measure how much they trusted the other person.
For the 85 participants who rated their friend’s job prospects as better than their own — leading to an upward comparison — affective trust was “significantly lowered,” the researchers write. On the other hand, cognitive trust clearly dropped in the cases of the 50 students who thought the friend’s prospects were worse than theirs. (Twenty-five participants expressed equal comparisons.)
The second study, which Schweitzer calls his favorite of the three, took a page from the ever-competitive world of LSAT scores. “Performance on the LSAT is self-relevant, so the scores are a very good measure for social comparisons,” he notes.
Under the guise of an “information sharing” study, 262 college students were asked to list the LSAT scores of up to 10 peers. Students then answered trust-related questions similar to those in the first study about two people on the list (who were chosen because they had the highest and lowest LSAT scores of the group). Finally, participants rated how likely either peer was to get into a top law school, as well as prognosticating their own chances.
Similar to the first study, cognitive trust was significantly lowered in the case of peers thought to be of lesser ability. But the researchers found something unexpected when looking at instances where participants rated peers as better than themselves — the presentation order mattered because it created what Dunn calls a “buffer effect.” Affective trust was only harmed when students were asked about the upward comparison situation (i.e., the peer expected to perform better) first. When the scenario was presented after the student was asked about the peer thought to have worse prospects, affective trust was maintained. “If we first see ourselves as better performers [in some cases], then the sting of being compared to someone even better is lessened,” Dunn says.
The final study examined situations where people were unfamiliar with the individual or individuals they were being asked to trust. The researchers told 341 undergraduate students that they would be paired via computer with another person to complete three independent tasks and one joint task. After completing the tasks, participants were given an evaluation of their own performance, as well as feedback about the work done by their counterpart, which was skewed so some students were told that their partners achieved better results, some worse and some the same. Participants then answered questions about trust, ability beliefs and emotions in relation to their partner.
Like the first study, affective trust was significantly lower when students were told that their partner was the better performer as compared to the other two scenarios. “Participants in the upward comparison condition perceived the trustee to have higher ability and felt stronger negative emotions toward [that person],” the authors write. Conversely, cognitive trust was lower in cases where the partner was said to have achieved worse results.
Schweitzer and Dunn note that the results outlined in the paper could have a direct link to how people operate in the workplace. For example, they point out that managers should think twice about ranking employees or publicly singling out an employee for his or her accomplishments due to the negative impact it might have on workers’ willingness to trust their peers and the resulting complications for team dynamics. Instead, they suggest offering team awards or rewards that could mitigate the effect of social comparisons.
Based on the “buffer effect” findings in the second study, the researchers also note that managers could preserve workers’ self-confidence by prefacing negative criticism with comments about ways that the employee is outperforming his or her peers. “Managers and peers who boost the self-evaluations of others before revealing outperformance information may suffer fewer harmful consequences,” the authors state.
Simply put, Schweitzer says, “Looking down on someone kind of helps you out, and you feel better about yourself.”
Forget Spokespeople — Emotions May Be Advertisers’ Best Marketing Tool
Advertisers often appeal to consumers by using celebrity endorsements or spokespeople who embody the identity that the company is trying to convey — for example, a famous football player selling athletic shoes or a mountain climber extolling the virtues of outdoor gear. The thinking is that if a customer shares or would like to project that identity, then he or she will be more likely to go out and buy the product promoted by that endorser.
But companies may not need to have a big-bucks talent budget to make such connections with consumers. A recent paper by Wharton marketing professor Patti Williams and University of Pittsburgh professor Nicole Verrochi Coleman suggests that marketers can actually achieve that goal in a more subtle way — through emotions. The paper, “Feeling Like My Self: Emotion Profiles and Social Identity,” was published in the Journal of Consumer Research.
Coleman says that the idea for this paper came about after she studied negative emotions and wondered if these feelings could be useful, particularly when connected to a social identity. Williams, who was Coleman’s advisor when she was a doctoral student at Wharton, notes: “Part of this is to answer the underlying questions of, ‘Do people want to feel emotions?’ and, ‘How can emotions, even negative emotions, help consumers and individuals fulfill their ultimate goals?’”
To explore this connection, Williams and Coleman started with the theory that “[if] an individual repeatedly experiences an emotion as relevant to serve the goals of a specific identity, [he or] she may come to acquire an association between the emotion and the identity, even absent goal pursuit.”
The researchers chose three identities for their study: athlete, environmentalist and volunteer. The identities were chosen because each correlated to a primary emotion — athletes with anger, environmentalists with disgust and volunteers with sadness. “Anger, disgust and sadness are basic emotions that you see even in the expressions of primates, and they trigger specific actions,” Coleman notes. Anger is associated with pushing back against others, which is relevant to athletes; disgust is associated with judgments of purity, which is relevant to environmentalists; and sadness is associated with empathy and a desire to help, which is relevant to volunteers. Even though these are negative emotions, they are potentially useful negative emotions to the goals each of these identities hold dear, the researchers say.
In all of their experiments, Coleman and Williams activated one of the three chosen social identities in study participants by having them write about a time when they performed well under that particular persona. Then they were asked to list things they could do to project that identity to others. (By contrast, a control group was asked to write about the previous day and list the things they were planning to do the next day.)
According to Williams, the researchers considered recruiting professionals in the three fields for the experiments, but ultimately decided that it was not practical. Still, in a pretest study, 91% of participants identified as volunteers, 86% as environmentalists and 84% as athletes. Coleman notes that the lack of a professional filter makes their research even more relevant. “Emotions can be tied to an identity even if someone is not a professional athlete or out there with Greenpeace,” she says. “And from a marketer’s perspective, the practical implications make a much stronger case” if they apply to the broader range of people who might consider themselves to have a particular identity as compared to the relatively few who might be professionally employed as such.
The first experiment looked at how the different identities influenced reactions to the emotions in advertisements, specifically a print ad to promote STD testing. After being primed with one of the three social identities, 295 participants watched ads that portrayed a certain emotion through text such as: “Finding out I had herpes made me feel so furious (or helpless or repulsive).” Participants were then asked to rate their attitude toward the advertisement on a scale of 1 to 10. As the researchers predicted, “by incorporating an emotion consistent with [consumers'] emotion profile, social identity-relevance was achieved and greater persuasion resulted.” Those with an athlete identity had higher attitudes toward the angry appeal, while those with a volunteer identity preferred the sad appeal.
Music was the medium of choice for the second experiment. The 183 pre-primed participants listened to clips of 10 songs (five sad and five angry). They were then asked to create a playlist of four songs and provide a price point at which they would be willing to pay for a concert ticket for the band. In line with the outcome of the first study, participants primed as athletes preferred angry music, and were willing to pay more to see angry music performed, and volunteers gravitated toward, and were willing to pay more for, the sad music.
The third experiment — involving movie watching and tea drinking — was the favorite of both Williams and Coleman. “Not only did I put my blood, sweat and tears into this experiment, but it really points out the mechanism that controlling and managing emotions is an intentional thing,” Coleman says.
Participants were primed for either the athlete, volunteer or environmentalist social identity or were made part of a control group, and they watched a 45-second movie clip that was pretested to provoke anger, sadness or disgust. Then they were instructed to taste a cup of tea that was named and described to indicate that it would either enhance someone’s emotions (IntensiTea), reduce them (TranquiliTea) or have no effect (HerbaliTea.)
In the end, participants who watched movie clips in line with their prescribed social identity sought to enhance the corresponding emotions (as exemplified by their choice of IntensiTea), while those who viewed a clip in opposition to their social identity looked to reduce the related feelings (by opting for TranquiliTea). For example, athletes who had watched an angry movie, or environmentalists who watched a disgusting film clip drank more of the IntensiTea, while athletes who had watched a sad movie or environmentalists who had watched an angry movie, drank more of the TranquiliTea.
The researchers did one last experiment via computer games to test emotions with identity relevant tasks. Participants were instructed to click on a moving button on a computer screen as many times as possible. But one game was presented as “Click for Charity” while the other was “Click for Fitness.” While playing, participants listened to either sad or angry music, depending on their priming. Overwhelmingly, those who were given clicking instructions and music in line with their social identity performed better than those who were given mismatched identities, music and tasks.
“These studies establish that emotions are implicated within social identities, begin to articulate how social identity and emotions interact to influence persuasion and consumption, and suggest that emotions can play a role in identity,” the researchers write.
Williams and Coleman were both encouraged by the robustness of their results, and also with how marketing professionals can apply them. As Williams notes, the paper shows how important emotions can be in helping companies appeal to consumers. “The buzz word out there now is consumer engagement,” she says. “One great takeaway from this is that emotions themselves are part of an authentic relationship and can lead to natural engagement.”
“Bringing identity and emotions together gives a fuller picture of the consumer and what it means to him or her to be engaged in a product,” Coleman adds. “Emotions are a critical part of the dialogue between brands and consumers and the experience that brands deliver to consumers. Hitting the right emotion can promote connection and engagement and a deeper sense that the brand is the right one if the consumer wants to express a particular identity.”