« Steve Landsburg goes "meta" on me | Main | Putting the Econ into Econometrics »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

This strikes me as unfair, both questions get at the broader points.

1)If CEO pay was way off in 1981, why? Does this error that has presumably been fixed justify the astronomical rise in compensation?
2)Don't know why this question can't be read as being about the wisdom of options as a form of compensation. Assuming the bulk of the increase was from options ("a fair chunk"?) the question still stands, what did the CEOs do to earn that increase? If the increase is broad-based, and it usually is, the answer is: not much.

Hi Shiner,

1a) If CEO pay was way off in 1981, why?

I don't know that it was off, but I don't know that it wasn't either. That's the point.

1b) Does this error that has presumably been fixed justify the astronomical rise in compensation?

If there was an error and we've corrected it, how would that not be justified?

2a)Don't know why this question can't be read as being about the wisdom of options as a form of compensation.

It certainly can be. It's an interesting question. Roger Martin makes some interesting points against the use of stock options.

Eliminating stock options, however, does not necessarily mean that the Top 100 would necessarily earn less. (Presumably those stock options would be replaced with *something*) It would likely, however, mean that the compensation is less volatile.

2b) what did the CEOs do to earn that increase?

What did CEOs do to deserve the *decrease* in 2009? Need to consider the fluctuations in both directions.

Looks like you are guilty of the Grandfather Fallacy Fallacy, Mike.

You can't begin to start talking about this issue until you decide what it means for pay to be right. You have neglected to do this and so you don't really have anything to say yet.

That is not true of those whom you criticize. Their metric, implicit but not occult, is that the correct measure of reward is that which maximizes economic welfare. The claim is that welfare was higher in the 60's (or whatever) than now, therefore the usual explanation for rising income disparity, which is founded on "economic efficiency", is wrong. If you want to dispute that, then go ahead. But don't try to claim some sort of logical higher ground that you obviously do not possess.

"But one thing it does is make CEO pay highly correlated to overall stock market returns."

But not stock market losses. Gifted call options have the nice property (from the CEO's point of view) of zero risk of loss. Investors, on the other hand, can obviously lose money. Do you really want the person with no skin in the game placing your bets?


"You can't begin to start talking about this issue until you decide what it means for pay to be right."

Actually, that's my entire point, Phil.

"But not stock market losses. Gifted call options have the nice property (from the CEO's point of view) of zero risk of loss."

Only if you assume that if they had not got the options, they would not have received some other form of compensation.

Totally agree with your "grandfather fallacy" argument.

However, one argument from Moore's article that did make sense to me is the corporate board situation. Widely held companies often have boards that work for managements rather than shareholders, and this may be contributing to the executive pay increases. (This occurs because board members want to continue to be on the board, and its management that puts them there through recommendations. In some companies, you can be voted to the board even if a majority of shareholders votes against you, because votes "against" are counted as if there was no vote at all.)

If you looked at empirical data between the salaries of managers at private companies vs public companies (adjusted for size), you would probably find private company managers make less, as shareholder power is not divided in private companies as it is in public companies, making for a more accurate market price of executives. Private company salary data is hard to come by, however. Has this been done before? I'd be interested to read on the subject.

Take Yellow Media CEO. He earns millions and has run his company to the ground. Almost bankrupt. He got a lot of options at a much higher price than the current price but that's no problem for him...This year he will be entitled (yes, entitled) to another million or two of those same stock options.

On top, stock options are taxed like capital gains.

Take Manulife CEO...
Take Nortel CEO...

I have no problem with a large compensation for top executives but recent history proves that those guys are making a killing no matter how their company behave. Their salary should be based on a profit percentage spread over as many as 10 years. This would keep many of them to do short term moves to get their bonus and then the hell with next year and the stockholder.

I don't know that it was off, but I don't know that it wasn't either.

I do know that it was off. It was too much back then. It exceeded, back then, any resemblance to a just price for labour. From the tone of the question you quoted, John Moore is implying that a further deviance from a prior just price has been achieved.

This seems like a bit of a red herring, a little intellectual gymnastics in order to support an existing bias.

Clearly, executives should be paid as little as they are willing to accept. Given recent events, I find it hard to believe that the current crop of the executive class is fundamentally better at managing than their predecessors. So, if the executive class was content to have smaller shares of the profits a generation ago, then - to say nothing of moral and society at large - clearly shareholders are being ripped off.

The question becomes: what has changed? There are probably far more companies today than there were before. Is there an extreme shortage of people capable of being CEOs that would justify the increase in wages?

What's the economic explanation for the rise? Mike, do you believe that we have better CEOs on average than we did thirty years ago?

From where I'm standing, "collusion" and a collective "race to the bottom" seem like the most credible answers.

Not sure I follow that. I'm comparing with investors. But I think you mean relative to themselves in some alternate universe? E.g.

Assume in all universes CEO has same base salary.

CEO in this universe: Free stock options, stock tanks, no other compensation.

CEO in universe A: No stock options, stock tanks, no alternate additional compensation.

CEO in universe B: No stock options, stock tanks, envelope stuffed with $1 million cash.

So I think you mean that I assume alternate universe A, and that there is no loss. But you are thinking relative to universe B where there is a loss of $1 million?

@ Mike Moffat:

Minor point about writing:

"There is no reason to believe that we hit the correct balance at any one point in time and deviations from that are unwarranted.

"This has nothing to do with the grandfather fallacy"

Just when I thought that I knew what you meant by the Grandfather Fallacy, you hit me with the second clause. I know you started a new paragraph, but still. . . . I went, "Huh?" ;)

I've been kind of bothered by conflating "top 100 public company CEO pay" with "CEO pay" in these debates. The fact is that the largest 100 companies are substantially larger than the largest 100 companies 30 years ago, so we can already point to one big thing that's changed, and could go a long way to explaining how both the current and historic compensation levels can be correct.

On top, stock options are taxed like capital gains.
This isn't true. The difference between the option price and market price when the option is exercised is considered employment income for tax purposes. After options are exercised, they stop being options and become shares owned, and when sold the difference between the market price at purchase and market price at sale is a capital gain/loss.

Executive stock option compensation reflects the value of the option when granted.

The recorded compensation for a given year has nothing to do with stock market performance.

Options exercised later on are included in personal income tax returns. That's where stock market performance is reflected. But that's future personal income, not recorded executive compensation for any year.

The closest thing to what you’re talking about is that annual corporate information forms include the current market value of vested and unvested options for named officers. The media occasionally reports those numbers. But that’s got nothing to do with recorded compensation numbers.

P.S.

"annual corporate information forms include the current market value of vested and unvested options for named officers"

i.e. that's for options not yet exercised by those officers

Neil,
"On top, stock options are taxed like capital gains.
This isn't true."

If you read french, the article below in LaPresse yesterday, say that options are just like a capital gain.

"Dans une étude parue le 3 janvier sur les 100 PDG les mieux rémunérés au Canada, le CCPA souligne que l'élite des dirigeants gagne en moins de trois jours l'équivalent du salaire moyen des Canadiens, évalué à 44 366$. Afin de réduire les inégalités dans la société canadienne, le groupe de réflexion demande que les options et unités d'actions soient dorénavant imposées à 100% comme tout autre revenu de travail plutôt que comme un gain en capital, qui profite d'une exonération d'impôt de 50%."

http://lapresseaffaires.cyberpresse.ca/economie/medias-et-telecoms/201201/05/01-4482975-yellow-media-les-options-de-marc-tellier-ne-valent-rien.php

I'm not an accountant. Who can give a final answer on this?

Stock option compensation is not taxed like capital gains - it's not taxed at all, unless the exercise price is set below the market price, which is typically not the case.

Stock option gains when stock options are exercised (referred to as stock option benefits) are taxed exactly like capital gains.

But stock option benefits are not classified as capital gains, and can't be offset against capital losses.

Normand,
My French is a little rusty, but I get the gist.

I have found plain English explanations of options taxes which do say there's a 50% deduction (like this one http://www.professionalreferrals.ca/2004/05/tax-rules-on-employee-stock-option-plans/ ), but this is inconsistent with how I was taught Tax (I am an accountant, but not a tax accountant, and it's been about 3 years since I last took a tax class), and I can't find any reference to such a deduction in the official CRA interpretation document (http://www.cra-arc.gc.ca/E/pub/tp/it113r4/it113r4-e.html).

Stock options in "Canadian Controlled Private Corporations" (CCPCs) are subject to a 25% deduction, which was 50% before 1987. Obviously this wouldn't be relevant to discussion about options in the publically traded companies included in the CCPA report.

JKH, please source your claims. Having just run through the interpretation bulletin, linked above, I can find nothing to support your claims.
You said: it's not taxed at all, unless the exercise price is set below the market price, which is typically not the case.

The CRA said: Benefits from a stock option are generally included in the individual's employment income in the year in which he or she sells or exercises the option. (The exceptions relate to CCPCs, where you're allowed to defer the taxes until the stock is sold, not the exercise price)

You said Stock option gains when stock options are exercised (referred to as stock option benefits) are taxed exactly like capital gains.

The CRA said: an employee who exercises a stock option and acquires shares is generally required to include in employment income, in the taxation year in which the shares are acquired, a benefit equal to:
•the fair market value of the shares at the time the shares are acquired by the employee
Minus:
•any amount paid or payable by the employee to the corporation for the shares, and
•any amount paid by the employee to acquire the right to acquire the shares

The only deduction is the 25% deduction allowed by for stock options in CCPCs

"The CRA said: Benefits from a stock option are generally included in the individual's employment income in the year in which he or she sells or exercises the option."

That's exactly what I said.

But it's not reported by the corporation as executive compensation.

It's a stock option benefit that's reported by the taxpayer as employment income years later, but it is not recorded as executive compensation.

The exec compensation is the value of the option when granted, and its not taxed.

The final part you quote excludes the tax treatment. It's not taxed as regular income. There's an adjustment equivalent to the capital gains adjustment first.

Here is what I found from CRA and titled "Capital Gains 2011"
http://www.cra-arc.gc.ca/E/pub/tg/t4037/t4037-11e.pdf

At page 15 under Employee Security Options, they say:
"If you qualify for a security option deduction on line 249 of
your income tax and benefit return, you can claim one-half
of the amount recognized (and reported as income) as an
employment benefit from the sale of eligible securities
in 2011."
It seems that JKH is right for public corporations unless I misunderstand something...aaaahhh I wish that a French blog like this one exist.

Yes, I'm right for public corporations.

Just to clarify, recipients of employee options issued by corporations or mutual fund trust (not just public corporations) are generally entitled to, effectively, capital gains treatment on the exercise of those options (provided they weren't in the money at the time they were issued, and certain other criteria are satisfied).

For corporations which are canadian controlled private corporations, pseudo-capital gains treatment is available even if the options were in the money when issued, provided that the employee holds the shares for a two year period after exercise (unlike employee stock options for most companies, CCPC employees don't have to realize an income inclusion until they dispose of the shares they acquired with their options). (see Sections 110(1)(d) and (d.1) of the Income Tax Act).

Incidentally, Neil isn't wrong when he says that the employee benefit is taxed as employee income, rather than a capital gain. It is, that's why they need the 50% deduction which to get, effectively, capital gains treatment. This can have perverse results. If the value of the shares ultimate go down and the employee realizes a capital loss, they can't carry-back the loss to reduce the income they received when they exercised the options (which you would be able to do had you realized a capital gain on the exercise of the options). This can lead to unfortunate situations where people exercise their options, triggering a hefty tax liability, then the share price plumets, and the employees can't come up with the cash to pay their tax liability (as you might imagine, this was a big issue back in the tech boom).

Finally, what is seldom mentioned in discussion of employee stock options is that, while they result in favourable tax treatment for employees, they have quite unfavourable treatment for employers. If a corporation pays its employee an extra $1,000,000, the employee will pay tax at rate of 46% (bad for the employee), but the corporation would get a deduction in computing its income, saving tax at rate of 26.5% (all rates are for Ontario). On the other hand, if I exercise employee stock options worth $1,000,000, I'll only be taxed at rate of 23%, but on the other hand the corporation can't deduct the "cost" of issuing new shares (and certainly, for existing shareholders, the newly issued shares are a cost).

The end result is that, provided a corporation is otherwise profitable (i.e., that a deduction is worth something to it), the fisc does slightly better by having the corporation compensate its employees with stock options rather than cash (shareholders, of course, get hosed, since the economic cost is the same whether the corporation pays cash or dilutes existing shareholders, but the tax treatment is worse with stock options).

Bob Smith,

"the employee benefit is taxed as employee income, rather than a capital gain"

Employment income, but at a capital gains rate, although stock option employment income can't be offset against capital losses

“Finally, what is seldom mentioned in discussion of employee stock options is that, while they result in favourable tax treatment for employees, they have quite unfavourable treatment for employers.”

I don’t think so, and not for reasons associated with your example.

Suppose the corporation issues option grants for stock exercisable at a current market price of $ 50. Suppose the options are valued by formula at $ 7.

As I said, that $ 7 is treated at employee compensation, but it is not taxable as income for the employee.

Although $ 7 IS the number that shows up in annual compensation surveys as stock option compensation, which was the original point, and what is incorrect about the post.

The reason it’s not taxable is that the employee would be subject to double taxation of option benefits, if it were. Suppose the employee exercises the option 10 years later at a market price of $ 100. Then the gross employment income from that is $ 50, which is taxable at the capital gains rate – i.e. it becomes taxable income of $ 25 at the regular income tax rate. So the employee ends up getting taxed once for the full realized benefit of the stock option, but at a capital gains rate. He'd be double taxed if taxed on his original compensation, so he's not.

Back to the employer - the economic effect of the original option grant and the ultimate employee stock option benefit on the company’s own balance sheet is as if the company had originally issued stock at $ 50 and the market price is now $ 100 – in fact that IS the realized effect of the full option transaction, start to finish, in this example.

So it’s not correct to compare the firm’s tax position by comparing it with the employee’s tax position.

The correct comparison is to that of an outright share issue at the time of the option grant, with the proviso that this doesn’t hold in the case where the options are never exercised at all.

And the reason for all this is that the option is a contingent capital transaction. If exercised, it's treated for tax purposes as an actual capital transaction would have been.

So the only tax question left is what is the tax treatment to the firm of the employee’s original compensation ($ 7), not the stock option benefit ($ 50), because the latter is of no more concern to the firm than would have been a shareholder’s capital gain of $ 50 from shares bought at new issue at $ 50. And the firm has no tax interest in such a shareholder gain - whether in stock issued outright or stock issued by options.

I'm not certain about the tax consequence to the firm of the original $ 7 dollar compensation expense. If it’s classified as an expense, it’s a non-cash expense. But I just don’t know what the treatment is.

There was a letter in the Financial Times on 5th Jan saying that the Swedish bank Svenska Handelsbanken has outperformed rivals despite having no bonus culture. See:

http://www.ft.com/cms/s/0/bc71c54a-36c6-11e1-b741-00144feabdc0.html#axzz1ikop8Vsh

Isn’t this all a classic example of the fact that you can tax away 100% of economic rent (or not pay it in the first place) and the economic effect is zero?

Outrageous CEO salaries are not the problem, they are a symptom of the problem; that being the system is now completely rigged in favour of a small percentage of the population claiming a growing share of the wealth we all help to create. These strawmen you keep creating only demonstrate that you just don't understand that.

Ralph Musgrave,

"Isn’t this all a classic example of the fact that you can tax away 100% of economic rent (or not pay it in the first place) and the economic effect is zero?"

I think there's a few hasty uniformity assumptions in that question, to put it mildly!

If it's the case that some firms can succeed in the market without a bonus culture, then that seems a good reason not to have third parties decide whether a bonus culture should or should not exist in a firm: firms are different, situations are different, employees are different, cultures are different... I do not believe that a uniform rule for all is going to be economically more efficient than letting competition reward whatever compensation scheme works best.

On the other hand, while ignorant third parties like us are not the right people to decide the structure of the compensation system, we are in a position to decide the tax system. The UK has a 50% top rate of income tax and our economy hasn't imploded (though I'd prefer a progressive consumption tax with higher rates/better enforcement and no taxation of investment, in order to make capital more plentiful relative to labour).

Has anyone read Roger Martin's new book on the subject, Fixing the Game? I just read it over the holidays.

What changed?

The role of CEO changed from being Manager-in-Chief to Stockholder's Agent. Formerly CEO's were expected to be good managers and expand the business by capital investment and market expansion. In the 1980's CEO's performance began to be measured by stock performance. This combined with the wave of hostile takeovers to radically change CEO behaviour incentives.

Before 1980 CEO's and labour had close or nearly matching desires: a stable business and growing corporate income.

After 1980 CEO's and labour parted ways. CEO's discovered that cost-cutting and stock performance were just as desirable and in fact necessary to survive in the new era. Labour was no longer an asset to be increase, it was a cost to be contained.

CEOs are no longer Institutional Managers but are Chief Investors with far shorter horizons.

I'm puzzled. What's the difference between Landsburg's "Grandfather Fallacy" and Burkean conservatism? Burke argued that it's dangerous to make radical changes to social institutions which we've inherited from our, er, grandfathers, on the basis of currently fashionable theories. (In this case, the theory which Roger Martin points to is Jensen and Meckling's 1976 paper on the principal-agent problem, which advocated tying CEO compensation to the stock price.)

Why would CEO compensation in 1981 have been one-eighth of its fair level? Is there some reason that CEOs at that time would not have been able to negotiate up to a fair level?

--
My theory about rising CEO compensation is that nearly all CEOs will think that they're above average, so they'll feel that they deserve above-average compensation. It's like the fact that 80% of drivers think they're better than average. (It's a cognitive bias with its own Wikipedia article, "Illusory superiority"; Roger Martin calls it the "Lake Wobegon effect.")

Daniel Kahneman notes in passing in "Thinking, Fast and Slow" that the correlation between CEO quality and corporate success is not that high: there's a lot of factors outside the CEO's control.

"A very generous estimate of the correlation between the success of the firm and the quality of its CEO [based on research into CEO performance] might be as high as 0.30, indicating 30% overlap. To appreciate the significance of this number, consider the following question:

"Suppose you consider many pairs of firms. The two firms in each pair are generally similar, but the CEO of one of them is better than the other. How often will you find that the firm with the stronger CEO is the more successful of the two?

"In a well-ordered and predictable world, the correlation would be perfect (1), and the stronger CEO would be found to lead the more successful firm in 100% of the pairs. If the relative success of similar firms was determined entirely by factors that the CEO does not control (call them luck, if you wish), you would find the more successful firm led by the weaker CEO 50% of the time. A correlation of .30 implies that you would find the stronger CEO leading the stronger firm in about 60% of the pairs--an improvement of a mere 10 percentage point over random guessing."

Russil Wvong,

"My theory about rising CEO compensation is that nearly all CEOs will think that they're above average, so they'll feel that they deserve above-average compensation."

I'm not sure that this explains much in itself. I may THINK I deserve any number of things!

On the other hand, if CEOs are unusually willing to tolerate unpaid leisure (by the standards of high income groups) that could be a factor.

W. Peden: I would guess that a typical compensation committee is also likely to believe that their particular CEO is above-average, and thus deserves above-average compensation. Just like in Lake Wobegon.

"Just like in Lake Wobegon."

That's why the whole idea of making CEO compensation public (back in the early 90's in Canada) was questioned at the time. Could have a lot to do with expanding multiples over average worker compensation since then. Boards saw the averages publicized, and wanted their own place to be top tier.

Russil Wvong,

I think that that would be the causally important delusion.

JKH,

That raises an interesting question: are corporate salaries a bit like Giffen goods? If so, publicising them would have the opposite of the intended effect.

W. Peden,

I don't know enough about Giffen goods to answer than.

But I'd say this. I think economists may underestimate just how determined Boards may end being about getting the right candidate for a particular CEO job, and narrowing that candidate down from a relatively small group of contenders to begin with. I keep hearing the populist refrain that there's a bounty of people who can run large corporations as well as anybody else. Even if that's true, that's not the way the recruiting dynamics work, even before you get to the issue of compensation. But once the Board identifies the person it wants, it becomes the case that that's the person that it REALLY wants, and that dynamic puts incredible pressure on pricing, given that the desired candidate at that point is effectively in the position of almost being a pure monopolist in setting the price that HE or SHE wants. At least, that's the pricing dynamic at the margin, which ratchets compensation standards up systematically.

JKH,

I seem to remember reading something similar about "golden handshakes". When a board really need rid of someone, the premium of avoiding the losses from keeping them on as CEO is very high.

W. Peden,

Sounds right. The inverse of REALLY wanting a new guy.

The premium is for effective insurance against law suits and their associated costs - direct monetary, reputational, distraction cost, etc.

I'm not sure that the comparison point in the past needs to be right for there to be value in the observation that there has been a significant change. The existence of that change begs a significant explanation. Absent one, the change raises questions.

W.Peden: "When a board really need rid of someone, the premium of avoiding the losses from keeping them on as CEO is very high."
The why, when one of those high-pais CEO is goledn-handshaked, do most of them get immediately rehired somewhere else? e.g Bob Nardelli.
I can think of a prominent professionnal Quebec CEO, who famously said :" You don't need to know tomatoes to run a grocery chain" and promptly set it to the ground. He was hired after having botched his preceding appointment and then was promptly rehired somewhere else to the same results.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad