Published May 12, 2009

Should CEOs suffer like their shareholders?

By CHEW XIANG

	Email this article
	Print article
	Feedback

IS, SAY, $40,000 a year a good salary for an engineer? What about a teacher? A journalist?

Whatever the case, an answer is possible and can be defended with reference to fairly intuitive judgments, even if each profession is diverse and heavily stratified. That's why $600,000 a year for the boss of a charity is a headline in itself.

But our intuitions don't apply to the outliers: the Atlases holding the corporate world up on their shoulders. A million a year for a CEO used to be vaguely troubling. What's the benchmark for shock now? Ten million? A hundred million?

Is there a smell test for CEO pay? Given the astronomical sums involved, clearly not - because once you go over a certain sum, every CEO's pay just looks large.

That's why the appropriate mantra now is 'pay for performance', and companies peg executive pay to plausible standards such as economic value added, net profit, total shareholder returns and the like. Because performance can - however inexpertly - be measured, it can provide a reassuring context no matter how extravagant an individual's pay might seem at first glance.

The problem is that none of these measures, on its own, gives a good explanation of why the 2008 pay of some of Singapore's top executives went up or down from 2007, as a BT report today shows.

Keppel Corp's former chairman Lim Chee Onn (the highest paid in the sample of Straits Times Index companies) earned between $10 million and $10.25 million for 2008 - up at least 14 per cent from 2007, when he was paid between $8.75 million and $9 million. Yet the company's net profit fell about 3 per cent, and total shareholder return (what you get from buying the stock in January and holding it until December) was down over 60 per cent, underperforming the STI benchmark by 15 percentage points. The increase did, however, correlate almost one-to-one with the increase in the company's economic value added (EVA) - roughly, net operating profit minus cost of capital - which went up 14.6 per cent from $604 million in 2007 to $692 million in 2008.

EVA also explains CapitaLand CEO Liew Mun Leong's drastic salary drop in 2008 to $4.18 million from $21.7 million the year before. That's an 80 per cent fall; the company's EVA fell 71 per cent.

But EVA doesn't work for other companies, even those that use it to determine bonuses. ST Engineering boss Tan Pheng Hock saw his pay jump 30 per cent in 2008 to $3.94 million, from $2.88 million; EVA, however, fell over 5 per cent, as did net profit. Sembcorp Industries' Tang Kin Fei got a 30 per cent raise - EVA rose 22 per cent, so that's a good fit. But Tan Kwi Kin, in charge of Semb- corp unit Sembcorp Marine, was paid 12 per cent more in 2008, even though EVA more than doubled, including a gain from 'unusual items' in 2007. It also doesn't explain why Mr Liew, whose company achieved economic value-added of $660.6 million (just $32 million less than Keppel) got paid 40 per cent of what Mr Lim got. No CEO in the sample - save Mr Liew, it must be said - saw his pay go down by anything approaching the loss suffered by shareholders.

There are no doubt innocent, legal and completely moral explanations for these supposed aberrations. In some cases, reported pay is reward for performance over a period of years. Mostly, they arise because different companies use different measurements of performance, add that to peer group benchmarks, then leaven the results with a substantial dose of discretion. That includes factors such as: correcting incentive effects of the performance metrics; size of the company; perhaps even the likelihood of shareholder anger multiplied by the possibility of an embarrassing fuss in the media or the annual general meeting.

So how can we tell which CEO is a good deal for the company, and which one isn't earning his keep? Some critics focus on excess shareholder returns over a period of time; others use a combination of metrics. For still others, the key issue is that pay is still not sensitive enough to performance - whichever measure is used of it. And the use of peer benchmarks is routinely criticised for the Lake Wobegon effect, which is said to occur because no firm wants to have a CEO who is below average, and so no firm will allow its CEO's pay package to lag the average, causing an inexorable upward shift over time.

The amount of discretion involved, plus the multitude of possible performance benchmarks, makes it very difficult to tell the good from the duff - almost as if they didn't want you to know.