One of the most coveted titles for Wall Street analysts is that of the “All America Research Team” (AA), an all-star title given to those voted as top analysts by thousands of fund managers in an opinion poll organised by the Institutional Investor magazine. The result of this annual poll is prominently featured in the October edition each year and winners of the AA titles are celebrated by their employers and sought after by rival banks.
The opinion poll is largely subjective, asking institutional investors to evaluate analysts on a dozen or so dimensions. The top of the list includes industry knowledge, communication, responsiveness and written reports. Actual forecast accuracy appears near the bottom of the list.
While earnings forecast accuracy and recommendation impact are objective measures of performance for analysts, the future of their careers also depends on subjective evaluation, such as the opinions of institutional investors. Winners of the AA title earn around three times more than those without it.
This has big implications for analysts’ careers and interestingly, very different implications for male and female analysts. My work shows that men, for example, gain more from their connections with executives at the companies they cover than their female counterparts.
In our paper, Gender and Connections among Wall Street Analysts, Sterling Huang, Assistant Professor of Accounting at the Singapore Management University and I, found that men overall reap more benefits from connections than women both in terms of job performance and in terms of subjective evaluation by others. The connections we examined were alumni ties where a typical analyst, covering seven to eight stocks has on average, a contact with one or two companies’ senior officers or board members.
We evaluated Wall Street analyst year-end earnings per share (EPS) forecasts and buy/sell stock recommendations from 1993 to 2009. We measured their performance by the accuracy of their earnings forecasts and the price impact of their recommendations. We collected the analysts’ AA status from the Institutional Investor magazine over the same period.
For each analyst, we constructed their connections with senior officers and board members of the firms they cover, using biographic data of analysts and company executives. In terms of degrees of connectedness, we find that there is generally no gender gap in the connections the male and female analysts have. The female analysts are just as connected as their male colleagues, sharing a school tie with a senior officer or board member in about 25 percent of the firms they cover.
But we find a big difference in how much these connections help male and female analysts in their jobs. We find that while connections improve forecast accuracy for analysts across the board, the effect among men is significantly higher. For example, while connections lead to a 2 percent improvement in accuracy rankings in general, among men, there is a further improvement of about 1.8 percent. The effect of connections is even greater in their stock recommendation impact or how the market reacts to their buy and sell calls. Connections improve male analysts’ recommendation impact by about 1.2 percent, but not at all for female analysts.
It’s not just who you know…
Connections also directly contribute to male analysts’ odds of being elected an AA but not at all for female analysts, suggesting that investors subjectively value connections among male analysts but not among female analysts.
We also show that the very different impact of connections on job performance was particularly pronounced among young analysts. This vastly different ability to capitalise on connections at such an early point in their career paths could explain gender gaps that exist throughout long-term career trajectories. The cycle, it seems, starts at the entry level.
Our study echoes the earlier findings of my colleague Herminia Ibarra, in her 1992 paper, Homophily and Differential Returns, studying interactions of men and women in an advertising firm. She found that men were better able to use their network ties to improve their positions in the company they work for than women. In other words, they were also able to reap better returns from their connections.
Our own findings came as a surprise to us initially. We expected either no gender difference at all, or a simple gender difference in either the analysts’ connected, education, or ability to do their jobs. For example, one might expect that male analysts are just more connected; or that they are older and more experienced, and hence, maybe better at their jobs. Alternatively, one could hypothesise that female analysts are better at their jobs because first they are self-selected into a very tough work force, and second they are analytical and detail-oriented, important skills for being an analyst. But none of these simple gender stereotypes held true in our data. As mentioned before, men and women are equally connected. In addition, while 35 percent of women were Ivy Leaguers, only 25 percent of men were from the same prestigious universities.
We are thus left with the conclusion that the gender bias in business is much more subtle than we originally hypothesised. Upon leaving college, men and women analysts are equally skilled and qualified. But for whatever reason—perhaps social norms play a role—men are able to benefit from their social connections for career advancement more than their female counterparts. This could set off permanent differences in the career path and leading to the persistent gender gap at the top of the business world. It is telling that while 14 percent of Wall Street all-stars are women, but virtually none of the top bosses in any of the big firms are. It could be argued that even the most competent women remain in analytical roles rather than being promoted into general management because that kind of promotion entails subjective evaluations by others.
Part of the difference lies in the fact that there are just so few female officers and directors. Hence, while men have male-male connections, for women, they have very few connections with other female. When we looked at our data again, we saw that female analysts with a connection to a female executive at a firm under their coverage had a 2.5 percent improvement in accuracy ranking, which is bigger than the overall effect of 2 percent. However, the male-male connection leads to an even bigger 4.7 percent improvement. Thus the value of the “old boys club” is hard to refute in our data. We believe our work reveals the bittersweet reality of the decades-long effort in pushing for gender equality. On the one hand, we should celebrate the fact that outright gender discrimination in education, hiring and promotions are on the decline. In our data, female analysts are not under-represented in the AA analyst pool. On the other hand, the evidence clearly points to a more subtle—yet perhaps more insidious—form of gender bias: men and women may be evaluated using different criteria in our subjective mind. This form of bias may take a very long time to overcome; it may be so ingrained in our culture and social norms that we are unconscious of this fact. If so, making this point loud and clear, as we do in our work, will be the very first step in helping to institute changes in the right direction.
This is, first of all, good research. I would love to see more research on yet another "insidious" weakness of humans: we tend to listen and take leadership for granted from tall, slender people. We tend not to take the leadership for granted of small overweight people. Now, if we would skip the man / woman thing, and would just evaluate career success on length and BMI, I am afraid you would find a strong correlation. Perhaps the best predictor if only two variables are to be chosen.
"Problem" is, men are taller than women, on average. Usually a bit slimmer too. Now, suppose research would show that this is the case; that there is a much simpler explanation for the differences in pay and career success, could we than please stop shouting "gender bias"?
Occam's razor should be applied at INSEAD too.
"On the other hand, the evidence clearly points to a more subtle—yet perhaps more insidious—form of gender bias: men and women may be evaluated using different criteria in our subjective mind. This form of bias may take a very long time to overcome; it may be so ingrained in our culture and social norms that we are unconscious of this fact. If so, making this point loud and clear, as we do in our work, will be the very first step in helping to institute changes in the right direction."
what if making this point loud and clear is not enough either?
The findings are very instructive, and I am all for gender equity being female. However, it does not seem to me that the percentage differences between male and female analysts are not significant. So, the research findings should be interpreted more cautiously instead of shouting obvious gender bias.
Thank you for a solid factual analysis that provides proof of bias.
Probably not unexpected - that gender and race/culture biases (both minor and major) remain even in multicultural, open societies (like where the professors are based). The Wall Street data analysis is good - in that it truly does provide the first step in helping to institute changes in the right direction. The great thing about finance - is that in the end - cash is king and biases are thus an inefficiency that will purge over time!
Leave a Comment
05/03/2015, 09.45 pm
The devil is in the nuance.
As with Big Data, the findings are only useful if we know how to interpret them. These numbers -- in a vacuum -- may lead some to draw particular conclusions (insidious bias is indeed very real) but it's important to suss out not just the number of relationships but also a) the quality of the relationships and b) how men and women differ in their understanding of how to deploy them. On the one hand, the findings are counterintuitive since women are the more relational gender but on the other, women still labor under the notion that work is a meritocracy. That just doing good work (and a lot of it) will earn recognition and build influence. In fact, men understand this quite differently and have historically been better at leveraging their connections for influence. It's critical for women to embrace the fact that the currency of business is influence and that influence doesn't come from sheer numbers on a rolodex. Instead, women should actually make the most of what comes naturally -- building relationships -- and then reframe the objective: it's the quality of relationships that counts and these relationships must be deliberately curated and cultivated. Further, if all women do is work, all they'll get is more work: it's critical that they take the time to build relationships up, down and laterally because the truth is that people trust, hire and promote those whom they like or those with whom they are comfortable. So, bottom line, it's not just about building the rolodex. It's about building influence.