With the U.S. economy roaring back to life in the third quarter of 2003, the revival of what economist John Maynard Keynes called the "animal spirits" of investors may be at hand. And that, to some observers, portends an uptick in merger-and-acquisition activity (see "The Start of Something Big?").
Inevitably, however, many companies will make deals that they will come to regret. Study after study shows that most mergers and acquisitions destroy value, at least in the short term. Why do so many acquisitions fail? Experts typically cite a number of reasons—synergies and cost savings that don't materialize, intractable integration problems, insuperable cultural differences, and so on. Acquisitions, they conclude, are inherently risky undertakings.
To those in the field of behavioral finance, however, the fundamental problem is psychological. CEOs who engage in M&A activity, they say, tend to be overconfident. And overconfident CEOs are prone to pursue risky deals, or pay too much (the "winner's curse"), because they overestimate the returns they can produce from an acquisition.
"Every CEO who goes into [an acquisition] thinks he is different—that he will be able to pull it off," says Hersh Shefrin, professor of finance at Santa Clara University's Leavey School of Business and a leading advocate of behavioral finance. As a group, says Shefrin, CEOs are afflicted with what he calls "the Lake Wobegon syndrome," named after Garrison Keillor's mythical community where "all of the children are above average."
Whether would-be dealmakers are now inclined to pay greater attention to psychological considerations is difficult to say. But the lessons provided by the collapse of the stock-market bubble in 2000 serve to advance the cause of behavioral finance, an approach that applies psychology and sociology to finance. Researchers in this growing field can cite numerous value-destroying deals that, they assert, were the result of CEO overconfidence (see "From the Annals of Hubris," at the end of this article).
Most evidence for what is sometimes called the "hubris hypothesis" is circumstantial. "You'd love to interview CEOs before [deals are made]," says Terrance Odean, associate professor at the Haas School of Business at the University of California, Berkeley. But the circumstantial case is strong, he adds.
"If you see that acquisitions lose money over and over and over again—at different times, and in different industries—then you have to start saying, 'It looks like a bit of a pattern,'" says Odean. "And it's not to say that every single one of them was a mistake, but on average, there seems to be a bias toward buying when you'd be better off not buying."
To behaviorists, the explanations for this bias boil down to two: "either people are making honest behavioral decision errors, or they're cynical," says Odean. "You can explain some of these deals with agency theory—maybe it was bad for the company, but it was good for the guy who made the decision. And sometimes that's true. But I prefer to think that rather than a lot of corporate leaders are selfish and cynical, at least much of the time a wrong decision was made because of a combination of factors like hubris."
According to behavioral-finance theorists, overconfidence is one of a host of well-documented cognitive biases—systematic errors in the way people process information (see "Troubling Thoughts," at the end of this article). Shefrin likes to warm his students to the subject of overconfidence by asking them to rate themselves as drivers. "Typically, 40 percent say they're above average, 50 percent average, and 10 percent below average," he says. "Few people are willing to admit to being below average."
Two other pioneers, Richard H. Thaler and Robert J. Shiller, advanced the integration of economics and psychology by solving puzzles of economic behavior (Thaler)—why people adjust so poorly to a decline in wages, for example, or are reluctant to take a high deductible in insurance—and revealing the excessive volatility of the stock market (Shiller). Indeed, although much research has been conducted in the lab, devoted to experiments and problems that illuminate the effects of cognitive biases, behavioral finance relies heavily on real-world data. Shiller and others, for example, have compared historical dividend values and stock prices; Odean and others have found evidence of irrational behavior in the actual account records of thousands of individual investors.
The emphasis of behavioral finance on studying how investors and markets actually behave—as opposed to positing how they should behave—has created serious problems for two core assumptions of prevailing economic dogma. One assumption is that markets are efficient, with prices fully reflecting available information. The other is that investors will make rational, well-informed decisions under uncertainty. Behavioral finance argues instead that markets are inefficient, prices are more or less irrational, and people—investors, CEOs, even CFOs—are prone to systematic errors of judgment.
Optimistic to a Fault
The past two years have seen an explosion of interest in behavioral corporate finance. Shefrin, for example, is writing what he says will be the first textbook of behavioral corporate finance, to be published in 2005 by McGraw-Hill. When he proposed the book two years ago, he recalls, "the question I got from every publisher was, is there such a thing as behavioral corporate finance?"
Today, "many young academics are seriously at work on behavioral corporate finance," says Shefrin, studying the effects of managerial optimism and overconfidence, among other factors. For example, Ulrike Malmendier and Geoffrey Tate, assistant professors at Stanford University and The Wharton School of the University of Pennsylvania, respectively, conjecture that the longer a CEO waits to sell company stock or exercise stock options, the more likely he is to be overconfident.
"If this criterion is an accurate proxy for measuring which executives are overconfident," asks Shefrin, "then what do [overconfident executives] do? Are they more prone to engage in M&A activity? More prone to engage in investment projects within their firm—especially funded with cash? The answer is: yes to both."
J.B. Heaton, a trial attorney at Chicago firm Bartlit Beck Herman Palenchar & Scott who earned his Ph.D. at the University of Chicago with Thaler and Richard Vishny, has recently done much-cited work on optimism as a persuasive explanation of managerial decision-making. Heaton admits to being "agnostic" about whether psychological factors better account for asset prices than efficient-markets theory does; some arguments for market rationality—that arbitrage quickly erases pricing anomalies, for instance, and that investors learn from their mistakes—give him pause. But those arguments lack weight in the corporate-finance arena, he maintains.
"Managerial decisions, like the decision to embark on a certain restructuring program, to do a merger, to do a spin-off, to invest in certain projects, are decisions that are very hard to arbitrage," says Heaton. "If I disagree with one bank's decision to merge with another bank—if I think that's a value-destroying decision—there's very little I can do to arbitrage that decision [after the deal closes]. You can't undo what they've done, short of taking over the company or staging some kind of proxy contest that's very unlikely to succeed. From that it follows that irrational managerial decisions might survive longer than irrational decisions in very competitive asset markets, like the bond market."
As for learning from mistakes, this argument also falls short in the corporate-finance setting, says Heaton. "Psychologists have figured out—and we know this from everyday experience—that we tend to learn when feedback is very clear and very quick. For managers doing takeovers, or taking on new projects of any sort, feedback is often quite delayed. The feedback is also 'noisy'—not very clear."
What to Do
If overconfidence causes CEOs to make ill-advised acquisitions or launch poorly analyzed projects, what can companies do about it?
There are three things they can do, says Shefrin. The first is simply to raise awareness of the problem. "But that, typically, won't do it," he says. The second is to "teach people some psychological tricks—how to reframe business problems."
The third is to install effective group-oriented processes for evaluating mergers and capital investments. But Shefrin and others advise caution in putting groups together. Groups of like-minded people can lead to groupthink—amplifying rather than mitigating behavioral errors. Decision-making groups should be cross-functional, including people from finance, operations, marketing and sales, and perhaps human relations. All need to have accounting and financial literacy, says Shefrin, and "everybody has a responsibility to raise the alert about whether a particular decision leaves a group vulnerable to committing a particular type of error."
This sort of organized skepticism isn't new. When Alfred P. Sloan ran General Motors, once a group of GM managers had made a decision, Sloan would tell them "to go away and think about what was wrong with it," says Shefrin. "The next week, they would talk about that. That's a way of avoiding groupthink."
"Groups should far outperform individuals," insists Paul Schoemaker, chairman and CEO of Decision Strategies International Inc. and an adjunct professor at Wharton. But sometimes, when a powerful CEO is in charge, there is only so much that members in a decision-making group can do to challenge him, he notes. In such cases, the board of directors must be the check-and-balance of last resort.
What about the CFO? Ideally, the CFO should be a professional pessimist, offsetting the CEO's optimism with analytic rigor. But "being human beings, CFOs are just as vulnerable to the influence of very strong psychological forces as everybody else," remarks Shefrin. Cognitive biases, he says, can interfere with the rational use of financial tools. Discounted cash flow, for example, "is easy to teach as a technique, but much more difficult to put into practice. It involves a lot of judgment, and requires art as well as science."
In the future, Shefrin hopes, CFOs and CEOs will learn to understand the psychological issues associated with decision-making and craft business processes that mitigate the effects of those issues. Until that future comes, though, the sorry track record for mergers and acquisitions is unlikely to improve.
Edward Teach is articles editor of CFO.
From the Annals of Hubris
Many deals and projects succeed because confident business leaders take calculated risks. Overconfident leaders, however, tend to neglect the calculation. Here are two spectacular examples where hubris went before a fall.
Motorola launches Iridium.
In 1987, a team of engineers presented the case for a space-based cellular-phone system to three top Motorola executives, including then-chairman Robert Galvin. "The project was going to involve putting at least 50 satellites in space in order to do something that was extremely sophisticated and had never been done before," recalls Hersh Shefrin of Santa Clara University's Leavey School of Business. "It was going to require an outlay of several billion dollars over at least a decade." At the end of the two-hour presentation, what was the action item? "'Go for it,'" answers Shefrin. "No discounted cash flow, no net present value, no internal rate of return, no payback period—nothing." By the time Iridium finally debuted, in November 1998, Motorola had shelled out some $5 billion for the project. The next year, Iridium declared bankruptcy and was sold off to private investors.
America Online and Time Warner merge. There are two behavioral aspects to the catastrophic $163 billion uber-deal that was announced in January 2000, says Shefrin. One is inefficient markets. "Brealey and Myers [authors of the best-selling textbook on corporate finance] tell managers explicitly: Trust market prices," he says. "Well, if [AOL CEO] Steve Case trusted market prices, he would never have bought Time Warner with AOL stock. But he thought AOL was incredibly overvalued, and he was looking for a good deal."
The second aspect is overconfidence. Case and Time Warner CEO Gerald Levin had discussed a possible deal for months, but only a handful of Levin's colleagues knew about it. A merger of equals was the goal, but in December 1999, AOL stood firm: it wanted a majority stake in the new company. "What's Levin's action item?" asks Shefrin. "Does he get together with his board members and closest advisers and discuss, should we do it? No. He goes to his weekend retreat, he thinks about it—and he makes the decision." Levin settled for a 45 percent stake, and the deal was done. When it was announced, those who were shocked by the news included Time Warner's CFO, general counsel, and board. The company's stock has since dropped 70 percent, and some $200 billion in shareholder value has evaporated.
The literature on behavioral finance contains many examples of how cognitive biases lead people to make flawed decisions. Hersh Shefrin, professor of finance at Santa Clara University's Leavey School of Business, says that corporate managers should be particularly wary of the following.
Overconfidence and optimism. CEOs who overrate their abilities are prone to making M&A mistakes. Confidence is a sine qua non of CEOs; overconfidence is their Achilles' heel.
Loss aversion and aversion to a sure loss. Studies show that people are far more sensitive to a loss than to a gain of equal magnitude. Also, they have a tendency to gamble in order to avoid a sure loss. In corporate practice, this means that despite textbook advice to ignore sunk costs, managers will throw good money after bad in the attempt to salvage a failing investment.
Confirmation bias. Managers tend to overweight evidence that supports their views and underweight evidence to the contrary.
Frame dependence. How something is framed, or packaged, will affect how decisions are made. One example is the treatment of stock options in financial statements. "The traditional frame-independent, nonbehavioral perspective says it shouldn't matter whether you drag stock options to the income statement or just leave them in the footnotes," says Shefrin. "The underlying fundamentals are the same." Here, frame dependence goes hand-in-hand with inefficient markets, he adds. "If we had frame independence on the part of investors, and if markets were efficient, expensing stock options would be a nonissue. But because markets are inefficient, and investors respond differently depending on how the information is framed, it matters whether a particular treatment of stock options is made."