The Economist Defends CEO Compensation, Part 2

This post picks up where the last one ended, addressing an article in the Economist’s survey on executive pay. The Economist says that CEO pay increases are for the most part warranted; we say they make the mistake of attributing corporate performance largely to CEOs.

This next article on the rationale for high CEO compensation (third in the series; #2 reports briefly on l’affaire Grasso) “start[s] with that moment in the 1980s when the long-established relationship between workers’ and managers’ pay began to break down,” the LBO era:

Private-equity outfits….by rising up against inefficiency and waste in the corporations of the 1980s….started the ideas that led to today’s pattern of pay.

The raiders’ antidote was ingenious. By loading up the company with debt that needed servicing, the buy-out firms in effect took the decision about what to do with free cashflow away from managers and restored it to the capital markets. By taking the company private, they returned control of the board to the owners. And—what matters here—by motivating and monitoring managers, they got them to take tough, commercially astute decisions.

This analysis was articulated by, among others, Michael Jensen, an American academic…Whereas executives in public companies earned about $3 for each extra $1,000 of profits, managers in the buy-out firms earned $64, according to Steven Kaplan, now of the University of Chicago….[I]t was time to pay managers more like entrepreneurs. The share option, until then chiefly a way of avoiding tax, began to be used as a device to make managers think like owners. Because an option confers the right to buy a share at a set price, it should motivate the manager to push the share price as far above that threshold as possible. A share, in theory, is the present value of all of a company’s future dividends, so the long-term value of the company will go up. As it turned out—and as this report will explain later—this elegant solution to short-termism created its own set of problems. But it certainly motivated managers to try to maximise profits.

It’s interesting to see how history is rewritten to serve the needs of the present. Let’s look at the LBO formula. They targeted companies with stable cash flow plus high levels of expense and/or undervalued assets. So the management team (note that then the entire top management cadre participated in the upside, not just the CEO) was under pressure to slash costs and sell assets to pay off debt. And in some cases, management did not play an important role in creating value’ sometime the raiders were able to earn their returns simply by breaking up the business (the sum of the parts was worth more than the whole). Oh, and the article has their source of funding wrong. It wasn’t the capital markets. It was banks. And many of the later deals went spectacularly sour (remember Campeau?), nearly precipitating a banking crisis (Greenspan’s finest moment was engineering a steep yield curve in the early 1990s, which was a massive subsidy to the banking system and enabled them to rebuild their balance sheets).

What the Economist’s discourse fails to acknowledge is that the activities of these raiders scared other companies into improving efficiencies, lest they become the next target. Many corporations cut staff, sold their headquarters buildings, disposed of private jets, and took a tougher look at costs. The raiders had done a service for all stockholders by forcing companies to improve their game. You didn’t need to give CEOs huge rewards to get them to behave more responsibly. All you needed was a more competitive market for corporate control.

What the Economist story does not say is that the line of thinking engendered by the Jensen article was very different than the practice of LBO firms. The management teams got quite low current income, and had everything riding on the success of the deal. The got modest salaries, any typically had to borrow to buy their equity interest (so they also personally had high debt service). If the company went bankrupt (a real possibilty), they lost.

Back to the article:

The new spirit of “shareholder value”, as it became known, has brought about a change in the job of chief executive. For a start it has steadily become more precarious. Between 1992 and 1997, according to Mr Kaplan and Bernadette Minton, of Ohio State University, chief executives could expect to last about eight years in the job. Between 1998 and 2005 that fell to six years, and executives from underperforming firms were more likely to have been booted out by boards. The National Association of Corporate Directors and Mercer Delta, a change consultancy, looked at the replacement of 163 chief executives between 2002 and the first half of 2005 and found that a quarter of them were “non-routine”—meaning that the chief executive had been forced out or poached.

The growing tendency to jettison the chief executive may be one reason why boards have increasingly been searching outside the company for a replacement. According to Kevin Murphy, of the University of Southern California, that now happens in about one-third of such searches, compared with only 10% in the 1970s.

Increased hiring of CEOs from the outside is a bad outcome. Staff recruited from the outside, including CEOs, are more likely to fail. Indeed, one might wonder if the shorter tenure and greater tendency to hire from outside are not causal, but rather, parallel and reinforcing developments.

Notice also how the discussion has shifted away from the management team, which was the group that the LBO artists rewarded, to the CEO. This is a keen bit of sleight of hand. It was controversial even in the LBO age that a few at the top benefited when many of the people at the LBO’d firm had to work harder merely to keep their jobs. Now we see the CEO being seen as the linchpin of corporate success.

This is absurd. While a truly awful CEO doubtless could destroy value in a hurry, major corporations have substantial annuities: established customer bases and brands, pools of management talent, trained employees, patents and other intellectual property, established systems and procedures. The CEO did not build these assets, particularly if he came from outside. Corporate performance is subject to a host of factors outside his control that can hurt or help: the prevailing level of economic activity, oil prices, interest rates, regulatory developments, and random events, like a disaster befalling a major customer or supplier. No one knows for certain how much a CEO contributes to a corporation’s results, but intuitively, it seems that the current model for CEO pay is giving them way too much credit.

…Chief executives are under pressure. They know that if the board does not like them they are out. In turn, executives have become more selfish and less loyal to their companies….Managers will need compensating for taking on the increased risk of losing their job and the shorter tenure that goes with it…Pay will also tend to be inflated by the growth in an outside market for executive talent. When your present employer is the only company likely to offer you a top job, it has a big say in your remuneration. The more other companies are bidding for your talents, the more you can demand…

Ahem, employment has become less stable up and down the food chain. That may certainly warrant a desire to be paid better, but that doesn’t justify it. CEOs, unlike the rank and file, have a unique ability to influence their compensation. Similarly, there is greater mobility at all levels, but what leads to high CEO pay is perceived scarcity of talent, which reflects conservatism in hiring more than the lack of managers who could fill the role well. If you read Jim Collins’ Good to Great, the CEOs who led the companies that substantially outperformed the stock market for 15 years were what he called “Level 5 leaders,” unassuming, ambitious for the company rather than themselves, and frequently on paper not qualified to act as CEO when they took the job. Collins noted:

One of the most damaging trends in recent history is the tendency (especially by boards of directors) to select dazzling, celebrity leaders and to deselect potential Level 5 leaders. I believe that potential Level 5 leaders exist all around us, if we just know what to look for, and that many people have the potential to evolve into Level 5.

Back to the Economist:

All the while another more general force was pushing up executive pay. As the average firm size increases, so each company must pay its top executives more. When managers control more assets, they can make more of a difference to absolute profits. Hence, in a competitive market full of bigger companies, boards will be prepared to spend more on talent. Using a schematic mathematical model, two American-based European economists, Xavier Gabaix and Augustin Landier, concluded that the sixfold increase in the size of American firms between 1980 and 2003 may account for much of the sixfold increase in managers’ pay during that period. Tiny differences between the abilities of top managers could explain large differences in pay. Chart 2 (above), from another study, shows how in historical terms the level of pay is still relatively low in relation to the size of companies.

This is garbage in, garbage out. Per above, this makes the mistake of attributing performance unduly to CEO action. A larger enterprise has more resources, including more managerial talent. The CEO’s job is not six times bigger because he has six times as many assets. When top managers are running divisions that are as large as public companies were 20 years ago, shouldn’t they be paid like a CEO too? Apparently the Economist doesn’t think so. Well, Jeffrey Immelt, who when we last checked, was running a company with a lot of assets (General Electric) and not paying himself very much ($3.2 million) begs to differ:

The key relationship is the one between the CEO and the top 25 managers in the company, because that is the key team. Should the CEO make five times, three times or twice what this group makes? That is debatable, but 20 times is lunacy,

The Economist also fails to consider that the increase in corporate size might be due precisely due to the fact that CEOs know that they will be paid better for running a larger enterprise, and play the game accordingly. Think I am making this up? There is ample evidence that CEOs make acquisitions that benefit them more than the company.

Let’s look at the banking industry. Even a casual reader of the business press knows that the industry has been consolidating for the last 25 years, and the biggest banks hold an ever-larger share of total industry assets. That’s good, right? We have all been told that big banks are more efficient.

It’s pure and utter bunk. Every study ever done of cost efficiency of banks has found that the industry has an increasing cost curve, meaning bigger banks are LESS efficient than small banks. The only thing the studies disagree on is where the cost inefficiency starts (the range is $100 million in assets to $5 billion, which is still not a very large bank).

So these bank deals are not lowering banking industry costs. Yes, the banks may cut costs post acquisition, but they could have made those cuts in the absence of a deal.

It would therefore appear that CEO egos, as well as perhaps CEO desire for greater pay is a major motivator. Bank CEO pay is strongly correlated with the size of bank assets. And the CEO of the acquired bank always gets a handsome golden parachute as his consolation prize.

This happens in other industries as well. Carly Fiorina, for example, received a major increase in pay after the Compaq acquisition, a deal she lobbied for heavily.

The Economist again:

Had executives exploited lax governance to cash in, you would expect them to have done far better than high-flyers in other professions. Going by Social Security data, executives of non-financial companies represented 20% of the 30,000 or so Americans with the biggest incomes in 1994. But in spite of the huge increase in their pay over the following ten years, executives made up only slightly more of that group in 2004, according to Mr Kaplan and Josh Rauh, also from the University of Chicago. The results from tax-return data indicated a larger increase in the number of executives among the very richest Americans. But even then, the authors conclude, the growing prosperity of the very rich cannot be explained by the growing prosperity of executives alone: other high-flying groups were rapidly getting richer, too. There are fewer rich lawyers than rich executives, but in the decade to 2004 the lawyers joined the richest group at a faster rate. So, too, with financiers, who are both more numerous than they were and more numerous than high-earning executives.

Let’s not miss the little finesse in the paragraph above, that while CEOs didn’t represent a much bigger share of the top 30,000 households, they made bigger gains in the (undefined) richest group. The Economist ignores a reason why CEOs haven’t increased their representation at the top: trickle-sideways. As CEO compensation has risen dramatically, it is acceptable for service professionals to be paid accordingly. A CEO who makes $10 million or more isn’t going to object to a lawyer or investment banker who makes over $2 million. In fact, someone who made less than that would be suspect. And traders are a separate case. Their current high levels of pay reflect high liquidity in the markets and the low cost of leverage. But they too may benefit indirectly from high CEO pay. In the 1980s, there was a huge hue and cry when it was revealed that the head of Salomon Brothers’ bond arbitrage group, Lawrence Hillebrand, made $24 million, because it was so much in excess of what anyone earned in those days. Rising CEO compensation means that an 8 figure pay package is no longer controversial.

The Economist again:

…look at a second study by Messrs Kaplan and Rauh. They assembled several groups of companies with similar-sized assets (to allow for the fact that pay depends partly on a company’s size) and, within each size-group, sorted the chief executives into ten pay classes. They then took all the executives from the top class in each group and compared their companies’ share-price performance with that of their competitors; ditto for the second-to-top class, and so on. Chart 3, above, shows a clear relationship between pay and performance.

I’d like to read the methodology for this study, because most other studies have found no relationship between CEO pay and performance. And on the surface, this one sounds dubious, for it seems to show a relationship between CURRENT pay and CURRENT performance. That ain’t how it works. It is rare that actions taken in one accounting period produce results in the same period. That’s the logic behind options and other long term incentives, that actions in one period produce results in the future, but the study did not appear to have a way to reflect that.

The study similarly does not prove causality (if all CEOs got options, then duh, the ones whose stock went up the most would have the highest pay). We have no inkling whether the CEO actions resulted in better stock price performance, or whether they got lucky, or were benefiting from the actions of past managers.

Look at the pharmaceutical industry. It takes 10 years to get the average drug approved. Patents last for 17 years, and it similarly has taken over a decade to perfect the sales techniques used by drug detailmen. How can you possibly attribute much of any drug company’s performance to its current CEO? Yet most parties today, including the Economist, accept that as orthodoxy.