ontrol your destiny,” advised the legendary CEO Jack Welch, “or someone else will.” During Welch’s 1981 to 2001 reign at General Electric, the company’s value grew from $14 billion to more than $410 billion, revenues increased from $26.8 billion to $130 billion and stock prices rose by 2,200 percent. Fortune named Welch the “Manager of the Century” in 1999.
Yet Welch’s advice took an ironic twist two years later, when, at the age of 65, he was forced to leave GE. The company had a mandatory retirement policy for CEOs.
Such policies are common among large, publicly traded firms — some 47 percent require mandatory retirement for either their CEOs, board of directors or both. But, as one might expect, these policies are not without their critics. Do companies really benefit from forcing experienced leaders out the door?
Adam Yore, an assistant professor of finance at MU’s Trulaske College of Business, and Brandon Cline, an associate professor at Mississippi State University, are among the few researchers who have explored the question in depth. Their findings suggest that, well, it’s complicated.
“On the one hand, some studies have shown cognitive function declines with age, and we found that mandatory retirement policies can act as a mechanism to limit CEOs who serve beyond their effectiveness,” says Yore. “On the other hand, our study showed that experience is a significant positive factor of performance, which could counter assumptions about age-related performance.”
Yore became interested in the effects that CEO age has on firm value — its total worth in the marketplace — after stumbling across some CEOs in their 80s and 90s while researching an earlier project.
“We were missing some CEO ages in that sample, so I started hand-collecting the missing data,” he says. “My prior assumption would’ve been that, shortly after age 60 or 70, maybe they’re on the board or consulting or whatever. But they’re gone; they’re retiring. I thought: ‘That’s interesting. I wonder if that matters.’”
He’s not alone. At the Harvard Business Review, senior associate editor Walter Frick recalls that two years ago he found himself pondering a claim, published in a respected national magazine, that at age 40 to 45 scientific creativity begins a long, slow decline.
In his subsequent article for HBR, Frick noted that the youngest executive on the Review’s list of the “100 Best-Performing CEOs” was 46. Could it be, he questioned, that the most successful CEOs are actually past their creative peak? Frick says there was little guidance on the question from academe, just one paper from two economists at MIT and the University of Pennsylvania. That study, he says, suggested that, yes, there was likely a negative relationship between age and creativity.
“That is the only paper I know of specifically looking at age and CEOs before the Yore and Cline paper,” Frick says.
ore says he and Cline didn’t start out thinking about retirement policies. They at first focused solely on the relationship between a CEO's age and a firm’s value. Indeed, it never occurred to them to connect a firm’s value with mandatory retirement. Why not? Yore says that they simply assumed companies couldn’t just dump aging executives. “That’s age discrimination,” he says. “You cannot fire somebody because of their age.’”
Turns out that’s not always true. Yore soon learned that Equal Employment Opportunity law makes an exception for employees who are “bona fide executives or high policymaking employees.” Once these employees are 65 or older, companies can in fact force them out.
After realizing that “mandatory retirement policies” really means “mandatory” and that these policies exist for nearly half of the companies listed in Standard & Poor's Composite 1500, Yore and Cline broadened their scope. Along with exploring whether CEO age matters in valuations, they also wanted to learn what drives companies to adopt mandatory retirement policies in the first place. They asked, specifically, which firms are most likely to adopt them, what exactly do they hope to accomplish, and how are the interests of shareholders affected?
Yore pulled data from two databases indispensable to every corporate-governance scholar — Standard & Poor’s EXECUCOMP database, which contains executive information for all firms that are currently or have previously been included in the S&P 1500, and Standard & Poor’s COMPUSTAT, which provides fundamental financial and price data for both active and inactive publicly traded companies.
To examine their original question, Yore looked at CEO age with Tobin’s q, a common measure for a firm’s value expressed as the “market value” of its assets (what these assets would likely trade for at auction) divided by its “book value” (the actual cash worth or acquisition cost of those assets). In addition, he examined the relationship between CEO age and return on assets, a stand-in for firm profitability calculated by dividing a company’s annual earnings by its total assets.
For both firm value and firm profitability, Yore and Cline found that CEO age does have a statistically significant effect. However, the effect is far from constant. In fact, “for the vast majority of the CEO population, age doesn’t seem to matter,” Yore says. “The story is really in the tails.”
In other words, the age of the CEO seems to impact firm performance only among those with either very young or very old CEOs. On the young side, every year added to a CEO’s age brings a significant drop in firm value, up to about age 45. Then, for a couple of decades, things mostly level off, with firm value showing little variation across CEOs between the ages of 45 and 70. Past 70, the age effects kick in again, and every additional year of CEO age again brings a decrease in firm value. The story is similar when firm value (Tobin’s q) is replaced with firm profitability (return on assets).
These results alone, however, did not convince Yore that younger CEOs are a better bet than those past traditional retirement age. After all, instead of older CEOs leading to lower-value firms, it could be that lower-value firms attract older CEOs.
“For example,” Yore and Cline write in the study, “riskier high-tech firms with strong growth opportunities may prefer to hire younger, more technically savvy CEOs. Meanwhile, firms in more mature, debt-laden industries might instead prefer to select a CEO with extensive industry and financial experience.”
To find out whether such selection could explain the link between older CEOs and lagging firm performance, Yore and Cline reran the analysis, this time using what’s called an instrumental-variable model. The model, Yore explains, “allows for potential reverse causality in the relationship between CEO age and firm value.” Although the effect of age in the model was slightly weaker, it remained negative and significant: firm value fell 0.30 percent for each additional year of CEO age.
'Our evidence suggests that while retaining an aging CEO with accumulated firm-specific experience has positive effects on firm value, hiring an older, outside CEO with little institutional knowledge is detrimental to value.'
Having seen that firm value varies little across middle-aged CEOs, the researchers decided to set them as a reference group. So they grouped the CEOs into seven age groups with CEOs aged 53 to 57 as a midpoint. They then compared the other six groups to this middle-of-the-road group. This analysis revealed that the firms with the youngest CEOs, those younger than 43, were awarded an 11 percent valuation bump compared to firms in the middle-aged group. The group of firms with the oldest CEOs, those older than 68, suffered an 8 percent valuation discount — a 19 percentage point difference between the two extreme age groups. Replacing firm value with firm profitability, Yore and Cline found a 0.752 percent higher return on assets for the youngest CEO group and a 1.55 percent lower return for the oldest, for a 2.31 percent difference between two extreme age groups.
“Causal evidence is tough to come by in social science,” but these results are “certainly suggestive” that age is causing the decline in firm performance, Yore says.
The next question was why. Perhaps, Yore and Cline reasoned, older CEOs are less active, taking part in fewer mergers, joint ventures, divestitures, capital restructurings and the like. A check of the data proved their hunch right: Younger CEOs engage in about 10 percent more major corporate deals than the typical, middle-aged CEO, while older ones perform about 9 percent less.
But with age comes experience, and Yore says he wondered whether CEO experience might offset age’s negative effects. In particular, he wanted to see whether the amount of experience CEOs have as leaders at their current companies correlated with either firm value or profitability. Adding this “firm-specific human capital” into the analysis, Yore and Cline found that investors value a year of CEO experience at about 0.29 percent of firm value and experienced CEOs do, in fact, outperform their peers in terms of profitability.
Yore and Cline also examined CEO age and experience with long-run stock returns. Specifically, they formed stock portfolios on the basis of the age and experience of the CEOs on the day the portfolio was formed. Then they analyzed the stock returns for the following year while adjusting for differences in risk among the portfolios.
“The advantage of this methodology is that, while there may be systematic self-selection on the basis of valuation and operating performance for CEOs, this is all known to the market,” Yore explains. “Analyzing future stock returns should therefore be less susceptible to this self-selection problem in the data.”
With this analysis, too, they found that CEO age, independent of experience, negatively impacts shareholder wealth: The old CEO stock portfolio underperformed compared to the young CEO stock portfolio. However, the negative impact of older age was significantly linked with older CEOs who have less firm-specific human capital. The firms led by older, experienced CEOs did not underperform.
“So mitigating this decline in age is this experience story,” Yore says. “Collectively our evidence suggests that while retaining an aging CEO with accumulated firm-specific experience has positive effects on firm value, hiring an older, outside CEO with little institutional knowledge is detrimental to value.”
For Frick, the Harvard Business Review editor, the inclusion of experience was one of the most valuable contributions of the analysis.
“In the real world, age and experience are tightly linked, so for this paper to disentangle them and to say, ‘Yeah, there may be this link with age, but that’s only when you’re taking the biggest advantage of age, experience, out of the question’ — I thought that was a really helpful, practical thing for managers who might have been reading some of this other research about age — even some of my own previous coverage of age.”
aving teased out the effect of age and experience on firm value and profitability, Yore and Cline were ready to examine mandatory retirement policies. Instead of assuming that firms adopt these policies out of a concern that CEOs older than 65 will face age-related declines, they first looked to see whether these policies might be a way to force entrenched CEOs out the door. “We have classic stories in finance,” Yore explains. “One of them is the possibility that the manager or CEO may run the company in his own interests rather than that of shareholders. And there’s a whole literature on CEOs trying to entrench themselves, making it very difficult for themselves to be fired. So we find governance mechanisms that evolve to try to prevent that.”
Board independence is one of the main mechanisms, but it doesn’t always work. There are CEOs who manage to co-opt their boards. “It’s possible that these kinds of policies might evolve as some sort of binding mechanism to expel long-tenured CEOs who are entrenched,” Yore says.
This because these policies are firm-level policies, implemented by shareholders, not boards, he adds. But, in fact, Yore and Cline didn’t find evidence to support the entrenchment theory. Instead, what they found was that the older a firm’s CEO, the more likely the firm was to have a mandatory retirement policy — which, they conclude, suggests shareholders do view increased age as the problem.
Once again, however, age was only part of the story. When Yore and Cline brought in the variable of CEO experience they found that the more firm-specific human capital a CEO has, the less likely his firm is to have a mandatory retirement policy.
“I think both of these results suggest that shareholders value each of these CEO characteristics — youth and experience,” Yore says.
Finally, the researchers sought to learn how shareholders react to mandatory retirements; that is, whether such policies helped or hurt firm value. If, they reasoned, shareholders see a problem with forcing out older CEOs, then they would be likely to discount firms that have these policies.
For this analysis Yore and Cline limited their sample to firms with CEOs at least 60 years old because, Yore explains, the presence or absence of a mandatory retirement policy becomes important to investors only when the firm’s CEO is approaching retirement age. They then split the resulting sample between those firms with a mandatory retirement policy and those without, and, for each group, examined the relationship between CEO age and firm value.
For the group without mandatory retirement policies, the researchers found significantly negative effects of CEO age on firm value. In fact, in this group, the effects of CEO age were three times as great as the effects on firm value for the full sample. But when they looked at the group with mandatory retirement policies, the negative effects disappeared. Doing the same analysis with “firm profitability” in the place of “firm value” brought similar results.
The researchers thus concluded that mandatory retirement polices are having the intended effect: They do, in fact, limit the negative consequences of having an older CEO. This doesn’t mean every firm with an older CEO could benefit from putting an expiration date on the boss, Yore says. After all, the study also showed that when a CEO has extensive firm-specific human capital — someone like Jack Welch — the negative consequences of age aren’t a threat anyway.
The key, says Yore, is balance.
“What we do know is that any ‘one-size-fits-all’ model of governance will create more problems than solutions,” he says. “Although we found that mandatory retirement policies can represent an effective form of governance, in order to mitigate issues regarding underperformance of older CEOs, a clear need also exists to account for all personal characteristics of executives — especially experience.”