Morten T. Hansen, Herminia Ibarra, and Urs Peyer

A lot of people have blamed short-term thinking for causing our current economic troubles, which has set off a debate about what time window we should use to assess a CEO’s performance. Today boards of directors, senior managers, and investors intensely want to know how CEOs handle the ups and downs of running businesses over an extended period. Many executive compensation plans define the “long term” as a three-year horizon, but the real test of a CEO’s leadership has to be how the company does over his or her full tenure.

ceo

This article contains the first ranking that shows which CEOs of large public companies performed best over their entire time in office—or, for those still in the job, up until September 30, 2009. To compile our results, we collected data on close to 2,000 CEOs worldwide.

It may come as no shock that Steve Jobs of Apple tops the list. However, our ranking does contain a few surprises. You’ll see some relatively unknown faces at the top. The inverse is also true: Some obvious candidates in terms of reputation don’t make the top 50, which we’re printing in this issue—or even the top 100 or top 200. (To view the top 100 and access a list of the top 200, go to hbr.org/top-ceos.) In fact, our list overlaps very little with lists of the most-admired or highest-paid CEOs. When we analyzed the data to see which factors increased the likelihood that an executive would rank high, we uncovered a few more surprises. Although one might expect context to have a big effect, we found a wide diversity of countries and industries represented among the top performers. The CEO’s background did matter, however, as did the situation left behind by his or her predecessor. Our data highlight the great extent to which CEOs account for variations in company performance, beyond those due to industry, country, and economic swings. That drives home how important it is to use objective, long-term measures to assess CEOs and to inform CEO searches and succession planning.

How Did We Judge Performance?

To create our ranking, we identified the CEOs of all publicly traded companies that had made Standard & Poor’s Global 1200 or BRIC 40 lists since 1997. Considering companies from Brazil, Russia, India, and China in our research was critical, given the growth in emerging economies. To be included, a CEO had to have assumed the job no earlier than January 1995 and no later than December 2007. (See the sidebar “How the Ranking Was Created.”) That’s one reason why you won’t find CEOs such as Jack Welch, Warren Buffett, Larry Ellison, and Bill Gates here. They all took the helm before 1995, though they probably would have done well if included.

All told, 1,999 CEOs made our first cut, and of those, 731 were still in office on the date we stopped measuring performance. The entire group represented 48 nationalities and came from companies based in 33 countries. The median age at which these executives had become CEO was 52, and those still in office had an average tenure of six years. Only 1.5% were women, and only 15% of the CEOs worked for companies based in a country that was not their country of origin. It is still not a global labor market for chief executives.

Then we looked at the list of 1,999 executives and asked, Who led firms that, on the basis of stock returns, outperformed other firms in the same country and industry? Our ranking combines three measures: country-adjusted return, industry-adjusted return, and change in market capitalization during tenure. Of course, shareholder return is not the only measure of performance, and it omits contributions companies make to a wide group of stakeholders. But it is the fundamental scorecard for CEOs of public companies. And it’s the same scorecard for everyone.

Just How Good Were the Top CEOs?

As a leader, you had to produce remarkable performance to make it into our top 50. On average, those CEOs delivered a total shareholder return of 997% (adjusted for exchangerate effects) during their time in office. That translates into a spectacular annual return of 32%. Subtracting industry effects, the annual return is 30%, and country effects, it’s 29%. On average the top 50 increased the wealth of their companies’ shareholders by $48.2 billion (adjusted for inflation, dividends, share repurchases, and share issues). Now compare that with the average performance of the 50 CEOs at the bottom of the full list of 1,999, who during their tenures produced a total shareholder return of -70%, which corresponds to an annual return of -20%. On average these poor performers presided over a loss of id=”mce_marker”8.3 billion in shareholder value. The #1 CEO on the list, Steve Jobs, delivered a whopping 3,188% industry-adjusted return (34% compounded annually) after he rejoined Apple as CEO in 1997, when the company was in dire shape. From that time until the end of September 2009, Apple’s market value increased by id=”mce_marker”50 billion. The #2 CEO, Yun Jong-Yong, ran South Korea’s Samsung Electronics from 1996 to 2008. Yun is an example of a leader who has stayed out of the limelight. During his tenure he capably transformed Samsung from a maker of memory chips and me-too products into an innovator selling digital products such as leadingedge cell phones. Under Yun shareholder wealth increased by id=”mce_marker”27 billion, and the total industry-adjusted return was 1,458%. Yun is also the #1 performer among the executives who have completed their tenure. For this group, the track record is in. For current CEOs, in contrast, we need to be cautious; their record could change a lot. Another top performer who has kept a strikingly low profile is John Martin (#6), who has led Gilead Sciences, a California-based biopharmaceutical company, since 1996. During his tenure he delivered an industry-adjusted return of 2,054%, or 26% on an annualized basis. Described as a “quiet leader,” Martin is figuring out how to make lifesaving drugs available in developing countries. He presided over the development of Gilead’s one-pill-per-day AIDS drug and also the antiviral drug Tamiflu.

Did Star CEOs Make the Cut?

When we compared this list with others rating CEOs, one of the most interesting things we noticed was who didn’t appear on our list. Take the Barron’s 2009 list of the 30 most respected chief executives in the world, which a group of editors had selected after speaking with investors, analysts, and executives. Five executives appear in both the Barron’s list and our top 30: Steve Jobs of Apple, John Chambers of Cisco, Jeff Bezos of Amazon, Hugh Grant of Monsanto, and Terry Leahy of Tesco. But several CEOs that were “most respected” according to Barron’s are nowhere near our top 50 (or even our top 200)—namely, Jamie Dimon of JPMorgan Chase, Satoru Iwata of Nintendo, Sam Palmisano of IBM, and Rex Tillerson of Exxon Mobil.

Many other celebrity CEOs also failed to make the cut, including Carlos Ghosn of Renault-Nissan, Sergio Marchionne of Fiat, John Mack of Morgan Stanley, Jeffrey Immelt of General Electric, Daniel Vasella of Novartis, and Robert Iger of Walt Disney. Some of these well-known CEOs have not necessarily done poorly; they’re just not among the top performers in the world according to the total shareholder return they’ve delivered so far.

When we looked at rankings of the highestpaid chief executives in America, we also found little overlap with our top 50. This might be due to different time frames; although we cover the CEOs’ long-term performance, some lists cover annual compensation only. Nevertheless, it’s interesting that none of the people on the Associated Press list of the 10 highest-paid S&P 500 CEOs for 2008, for example, are in our top 50. There is, however, a bit of overlap with rankings that look at longer compensation time frames. Five of the 50 highest-paid CEOs in Forbes magazine’s overview of 2003– 2008 CEO compensation appear in our top 50. One of the five is Steve Jobs, who was the thirdhighest-paid CEO on that list; another is John Chambers (the sixth-highest paid), who is our #4 top performer. Still, the relationship between making the lists of most-admired and highest-paid CEOs and placing in the top 50 (or even the top 200) on our long-term-performance ranking seems tenuous at best.

What Helps a CEO Perform Well?

If you’re a newly appointed CEO, how much does the situation that you inherited or your own background predict your placement in the ranking? Our analysis teased out some insights into the factors that matter. While far from exhaustive, these insights can inform to-day’s debates.

Country and industry. A quick glance at the list reveals how geographically widespread strong performance is; no one country dominates the list. CEOs from U.S.-based companies fill 19, or 38%, of the slots on our top 50 list, but that is not unexpected, since 42% of the 1,999 CEOs in the study were from U.S. companies. Sixteen countries are represented in the top 50, while 25 countries are represented in the top 200. CEO performance doesn’t cluster heavily in either free-marketoriented countries or emerging markets. In fact, our analysis shows that only 8% of the variance in CEO performance in the ranking can be attributed to country-by-country differences.

While we noticed some clustering of performance by industry, our analysis showed that 11% of the variance in performance could be attributed to the industries the CEOs came from. Some industries are overrepresented in the top 200—notably, energy, telecommunications, health-care equipment and providers, and retailing. Only 4% of the entire group we studied came from the energy industry, but 12% of the top 200 slots are filled by CEOs from energy companies. Though it is not surprising that the energy sector is overrepresented, it is revealing that some low-growth industries, such as retailing, are well represented, too. This shows that CEOs can attain exceptional performance even if they’re not in a booming industry.

Although six of the top 10 CEOs come from companies that are in the IT industry or are internet-based (Apple, Samsung, Cisco, Amazon, eBay, and Google), such high-tech companies are not overrepresented in the top 200. It’s no shock to find many CEOs from the automobile, automotive-components, and media industries in the bottom of the study group. A few CEOs from the auto sector did very well, however. For example, Chung Mong-Koo of Hyundai is #29 on our list. (Unfortunately, his reputation has been diminished by a 2007 conviction for embezzlement, for which he received a suspended sentence.)

Being an insider. Whether outsiders or insiders make better CEOs is a topic of much debate. One piece of conventional wisdom is that outsiders are more capable of instilling change and improving results, especially at underperforming companies, because they are more objective and beholden to fewer internal stake-holders and sacred cows. A good example of outsider success is #19 on our list, John Thompson. In 1999 he left his job at IBM to become the CEO of Symantec, a company experiencing lackluster performance, and during his 10year tenure transformed it into a standout.

The alternative view, espoused by Harvard Business School’s Joseph Bower and Rakesh Khurana and other management researchers, is that tapping insider talent for the CEO’s office is the better option. They argue that outsiders are expensive and that industry- and firm-specific knowledge is critical when it comes to generating long-term growth. Among the up-through-the-ranks leaders on our list are Yun Jong-Yong, who joined Samsung straight out of college and worked there 30 years before becoming CEO, and Mukesh Ambani (#5), who joined Reliance Industries in 1981, when it was still a textile company run by his father.

In our analysis of the 1,999 CEOs, however, we determined that insiders tend to do better. On average, they ranked 57 places higher than outsiders in the full list. Troubled companies were more likely than other firms to tap outsiders to be their chief executives. Thirtyseven percent of the companies whose returns were -46% or worse during the two years prior to a new CEO’s appointment chose an outsider for the top job, whereas 21% of all companies in the sample did. However, when we compared the results of CEOs who took over underperforming companies, the outsiders did no better than the insiders.

Having an MBA. In the aftermath of the financial crisis, pundits criticized MBAs, arguing that business schools had fostered destructive greedy behavior and taught executives the wrong models of management. So we decided to see whether CEOs with MBAs did any better or worse. When we looked at the CEOs from companies based in Germany, Britain, France, and the United States, where reliable information on degrees is available (1,109 CEOs in total), we found that the 32% of CEOs who had an MBA ranked, on average, 40 places better than the CEOs without an MBA. Even in the beleaguered financial sector, the MBAs tended to rank better than the nonMBAs. This finding suggests that MBA CEOs have not destroyed value, as some critics would have it.

A runway for performance. As a CEO, are you more likely to produce stellar performance if you inherit a struggling company from a mediocre predecessor? Or if you take over a strong company from a successful predecessor? Many would argue that a strong company is the best platform for generating superior results. We found that to be far from the case. The average rank of the CEOs who took over companies that performed poorly in the two years before they entered the job was 96 places better than the average rank of those who took over firms with great prior performance. This difference in rank becomes even wider when we look at the predecessors’ entire tenure. According to our analysis of the 790 companies for which we had data for both the predecessor and the successor, a stellar predecessor usually does not provide a good runway for his or her successor. On average, CEOs who took over from predecessors ranked in the top 50% of the overall group of 1,999 came in 583 places below those who didn’t. For example, John Bowmer (#143 on our list), the successful CEO of Adecco from 1996 to 2002, was succeeded by Jérôme Caille, who was reportedly asked to leave three years later amid disappointing results. Adecco’s performance under Caille placed him quite far down the ranking. Overall, we found few cases in which a highly ranked CEO passed the baton to a successor who was also highly ranked. Taking over from a low base presented a better opportunity for stellar CEO performance.

The gold standard. Because success seems so hard to sustain, perhaps the best measure of a CEO’s performance should go beyond his or her time in office. But most measures of performance, ours included, don’t look at whether a CEO leaves behind a strong or a weak company. The ultimate gold-plated list, then, would com-prise CEOs whose companies performed well not only during their tenure but after it. To construct such a ranking, we extracted from our database the executives who retired from their post three or more years ago, arriving at a list of 803 CEOs, whom we then ranked according to their company’s performance during their tenure and for the three years after their departure. (See the list “Whose Companies Performed Well After They Left?”)

While CEO rankings have exploded into a cottage industry, they haven’t really improved our understanding of what drives CEO success, because of the continued fixation on shortterm performance and the paucity of data on CEOs outside the U.S. In compiling this list, we’ve tried to overcome those two obstacles. We believe that examining CEOs’ contributions through a longer-term lens will give us a clearer view and better insights.

The top 50 list shows that no country or industry has a corner on performance. But taking a longer perspective did bring to light a number of “hidden gems”—quiet CEOs who delivered outstanding results year in and year out, away from the glare of the cover stories nd business school case studies. Their success makes a persuasive argument for a new approach to evaluating CEOs. Only by analyzing performance over their tenure and beyond can we begin to understand the nature of great leadership.

The Top 10 CEOs

Steve Jobs – Apple – Information Technology

Yun Jong-Yong – Samsung Electronics – Information Technology  

Alexey B. Miller – Gazprom – Energysider

John T. Chambers – Cisco Systems – Information Technology – MBA

Mukesh D. Ambani – Reliance Industries – Energy – MBA

John C. Martin – Gilead Sciences – Health Care – MBA

Jeffrey P. Bezos – Amazon.com – Retail

Margaret C. Whitman – eBay – Information Technology – MBA

Eric E. Schmidt – Google – Information Technology

Hugh Grant – Monsanto – Materials – MBA

How the Ranking Was Created

We selected the CEOs to track from the S&P Global 1200 and the S&P BRIC 40; we included CEOs from companies that had been in one of those indexes at least four consecutive years. To make sure we had reliable, and sufficient, data, we excluded CEOs who had assumed their role before 1995 or after 2007. We measured their financial performance through the last day of their tenure or September 30, 2009, if they were still in office. All told, we ended up with 1,999 CEOs from 1,205 companies.

Metrics. We pulled financial data from Datastream and Worldscope and calculated daily company returns for the entire length of each CEO’s tenure (or until September 30, 2009, if the CEO was still in office). We also calculated daily returns for the three years following a CEO’s tenure (if applicable). We used three metrics:

Country-adjusted company returns (TSR). We computed a company’s total share-holder return (including dividends reinvested) for the CEO’s tenure. We then computed the average return that firms from the same country had over the same period and subtracted that figure from the company’s return. This excluded any increase in stock return that was merely attributable to an increase in the general stock market.

Industry-adjusted company returns (TSR). We also deducted the average return for the industry, to exclude any increases in returns that were the result of rising fortunes for the overall industry.

Market capitalization change. We measured the change in equity market capitalization over the CEO’s entire tenure. We adjusted this measure for inflation in each country and translated values into U.S. dollars, using 2006 exchange rates. We added to this figure the inflation- and currency-adjusted value of the dividends and shares repurchased, and subtracted the adjusted value of shares issued.

We then ranked all CEOs for each metric— from 1 (best) to 1,999—and took the average of these three rankings to arrive at the final ranking. Using three metrics is a balanced approach; while the first two metrics risk being skewed toward smaller companies (it is easier to get large returns if you start from a small base), the third metric is skewed toward larger companies.

Analysis. We performed regression analy-ses on the data set of 1,999 CEOs. This allowed us to “control” for some factors and isolate the effect that one factor (such as having an MBA) had on a CEO’s standing in the ranking. The effects are reported in this article.

Harvard Business Review