Arnold Kling  

Stock Options as Tax Deferral

PRINT
U.S. as Debtor... Online Trade Barriers...

Kevin Hassett thinks that stock options are used because they are a form of tax-deferred compensation.


It is sensible that the tax code encourages equity compensation. Shareholders benefit when management has a significant stake in the appreciation of a firm's stock. But the tax code wrongly encourages reliance on options and only options. While options may be the most efficient form of compensation in some circumstances, simply granting shares of stock to executives is desirable as well. If policy makers want to fix a genuine problem, stock compensation should be treated in the same way that options are. And the rule change involved is quite simple. An executive who receives unrestricted and marketable stock shouldn't be required to pay tax until he decides to sell it.

I continue to agree with Hall and Murphy, who argue that firms use stock options to understate employee costs. If we are correct, then changing the tax laws would have no effect on the use of stock options. Indeed the fact that firms must treat stock grants as an expense even though they do not have to treat stock option grants--even though Hassett sees stock grants and stock options as substitutable forms of compensation--demonstrates the absurdity of the accounting treatment of stock options.
For Discussion. How might stock grants differ from stock options in the incentives that they create for corporate management?



COMMENTS (23 to date)
Don Lloyd writes:

Arnold,

"...Indeed the fact that firms must treat stock grants as an expense even though they do not have to treat stock option grants--even though Hassett sees stock grants and stock options as substitutable forms of compensation--demonstrates the absurdity of the accounting treatment of stock options...."

The 'absurdity' is the treating of stock and option grants DIFFERENTLY. The expense of BOTH stock and option grants fall fully on the shareholders directly. This has two implications. First, ANY artificial expense charged to the company will inherently overstate the degree of injury to the shareholders. Secondly, since both the stock and option grants are real expenses to the shareholders, they must effectively generate compensation tax deductions. Since it is inconceivable as a practical matter that every shareholder individually be given a tax deduction every time any employee exercises an option or sells stock, this tax deduction must be processed through the company's tax return.

Regards, Don

Kyle Markley writes:

You're right, the different accounting treatment for grants and options _is_ absurd. Grants shouldn't be expensed either. :)

Reading the linked PDF gives me a good insight into why we disagree about this. The authors say that "when a company grants an option to an employee, it bears an economic cost equal to what an outside investor would pay for the option."

That's what I disagree with. That "economic cost" is not an actual, tangible, accounting-worthy cost. I don't believe opportunity costs are, well, _costs_. Maybe that makes me old-fashioned (?) but I don't think that makes me wrong.

Sean writes:

Given that the potential relative return of options is much greater than that of actual stock, executives have a much greater incentive when they are granted options to drive up the price of their stock in the short-run while ignoring the long-run value of the company. Removing biases in the tax code towards the issuance of stock options might help to remove incentives to artificially inflate stock prices.

Joe Kristan writes:

I think the popularity of options a matter of both tax and financial statement considerations(not to mention the glassy-eyed look some employees get at the mention of the word "options"). I do find fewer options in partnerships, where the tax treatment of compensatory options is unsettled; the tax treatment is clearly a factor.

I still fail to agree with those who say options are not an expense. The issuance of options (or stock) to an employee forfeits an opportunity to issue the instrument for cash. If it is considered income by the receipient, logic dictates that the grantor is foregoing the same value. The expenditure of a corporate resource seems to me the definition of an expense.

Don Lloyd writes:

Joe,

"...The issuance of options (or stock) to an employee forfeits an opportunity to issue the instrument for cash. ..."

If you take out your bank checkbook and write check number 1004 to Arnold, it is true that you can no longer write check number 1004 to me, but I will be perfectly satisfied with check number 1005. If you write check number 1025, you may need a new checkbook, but you still will be able to write check number 1026.

When a company issues new shares, it does not preclude the company from issuing additional fully equivalent new shares.

Regards, Don

Jim Glass writes:

"...who argue that firms use stock options to understate employee costs."

I don't get it. Who is supposed to be being fooled? Shareholders know about all stock options -- certainly *these* days they do -- and they are the only ones who incur any cost, so who is supposed to be being fooled?

While I'm aware that the plural of "anecdote" is not "data", I'll still relate from personal professional experience that the reasons why companies I know use options are:

1) Options give an "up side" to executives with no downside. They like that.

2) While stock grants may be preferred by company owners, because they do provide a "down side", the Tax Code discriminates against them.

This is a "non-trivial" factor, as they say -- *especially* as the Tax Code also gives special tax breaks to certain kinds of options. (Which itself is *added* to the "no down side" attraction of options to executives.)

3) In some cases businesses I've known have also used options to save the cash cost of compensation, when conserving cash was important, especially in a start-up stage.

But no business that I've ever worked with has ever granted options to fool anybody about the cost of compensation. Fool whom? Like the shareholders who control the company and grant the options don't know the cost to them of the options they're granting? They're out to deceive themselves?

Now admittedly these aren't Fortune 500 companies I'm talking about. But the markets know about the options issued by major publicly traded corporations, and I find it hard to believe that the markets don't value their shares using that knowledge -- certainly *these* days. The studies I've seen indicate they do.

So I am definitely on the "tax rules count" side of this, from personal experience.

Bernard Yomtov writes:

Kyle,

Not only are opportunity costs real, but in fact ALL costs are opportunity costs. When you spend money on something what you are doing is giving up the chance to spend it on something else, now or later.

Jim,

I agree that it seems silly to think that options are given to fool people, but I think that many people, including many successful businessmen, underestimate the cost of options, and indeed, of equity itself. The "no cash cost" idea gets in their eyes.

At the small business level I think this does account for some amount of option issuance. With larger public corporations there is not always willful deception, but there is undoubted appeal in being able to give away something valuable that doesn't hit the income statement. Consider the single most spurious argument against expensing options - that it will reduce earnings and unfairly depress stock prices. That this argument is seriously advanced suggests that deception does play a role.

Joe Kristan writes:

Don,

There is still an opportunity cost to issuing new shares. Each share issued dilutes the claim of existing shares in future earnings. Where the shares are issued for less than market value, as in option shares, it dilutes the current equity of existing shares as well. Granted, it never reduces the value below zero (which your check number 1026 might), but it does reduce the value of the old shares nonetheless - just as a payment of compensation in cash would.

By your reckoning, any profitable company could issue an infinite number of new shares. A real economist could say why they can't; I just know they can't.

Kyle Markley writes:

Bernard,

"Not only are opportunity costs real, but in fact ALL costs are opportunity costs. When you spend money on something what you are doing is giving up the chance to spend it on something else, now or later."

No. When I go to the store and buy a gallon of milk for $2.08, my cost is $2.08. It isn't the $2.08 *plus* the apples I might have bought *plus* the bread I might have bought *plus* anything else I might have bought. My cost is simply and only $2.08.

It is true that when I make a purchase, I can't use that money for anything else. But what do we gain by calling it a "cost" and confusing it with _genuine_ costs? How about calling it a "decision made with consideration of the marginal utility of the alternatives?"

Don Lloyd writes:

Jim K,

"There is still an opportunity cost to issuing new shares. Each share issued dilutes the claim of existing shares in future earnings. Where the shares are issued for less than market value, as in option shares, it dilutes the current equity of existing shares as well. Granted, it never reduces the value below zero (which your check number 1026 might), but it does reduce the value of the old shares nonetheless - just as a payment of compensation in cash would.

By your reckoning, any profitable company could issue an infinite number of new shares. A real economist could say why they can't; I just know they can't."

Your arguments are pretty clean, but fail to take two things into account.

First, the supporters of option expensing claim that the (opportunity cost) expense is the full market value of the shares granted. This is of a different order altogether than talking about dilution.

Secondly, in an ideal world, every single share of new stock sold on the market or granted to employees by the company has to result in a net benefit to shareholders, as judged by management. This means that both stock sales and grants are self-limiting and do not actually require an infinite number of new shares to be created. Stock sales only benefit shareholders for as long as the cash raised can be re-invested at an internal rate of return that exceeds the return that shareholders can expect from the current market price of the stock. Stock (or option) grants can only benefit shareholders to the extent that valued employees can be hired and retained at a lower injury to shareholders than would be the case for a total cash compensation. Both of these applications of stock creation are finite in beneficial extent.

Regards, Don

Joe Kristan writes:

Don,

Corporation X has 3 shares outstanding. The company has equity of $3,000. If the corporation pays a salary with $750 cash, equity declines to $2,250 - $750 per share. If the employee then buys a share of stock for $750, the expense doesn't go away, even though equity is restored to $3,000.

If instead corporation X pays the expense by issuing another share - or an option for one share with an exercise price of $0 - we are in the same situation as the reinvested cash example. Why should the effect on reported earnings be any different?

Of course, a sophisticated investor looks beyond the reported earnings to the value of future cash flows represented by the shares, so to some extent the issue is intellectual finger painting; a buyer will take into account the option dilution. Still, if you want to have an earnings number that means something, I think expensing exposes the hidden expense of the option or stock grant.

Certainly nobody who issues options will tell Congress to change the tax law so they won't be expensed.

Regards,
Joe Kristan

Don Lloyd writes:

Joe K,

"Why should the effect on reported earnings be any different?"

I will modify your example to clarify what really needs to be compared.

CASE I --

The company buys $10K worth of its stock on the market and has it delivered directly to an employee without taking ownership itself.

Result -- The company incurs a $10K compensation expense. The shareholders do not suffer a dilution of ownership as no new shares have been created, and suffer only indirectly as the company that they own has expended $10K.

CASE II --

Part 1 -- The company creates and grants $10K worth of new shares to an employee.

Result of Part 1 -- The shareholders suffer a direct cost due to the ownership dilution of their shares. This dilution can be quantified by the fact the shareholders would have to expend $10K collectively to restore their original ownership positions in the company by buying shares on the market from either willing sellers or from the employee himself. The company itself suffers no cost or expense.

Part 2 -- The company subsequently repurchases the newly created shares from the market for $10K, assuming that the market price has not yet changed.

Result of Part 2 -- When a company buys its own stock, this is equivalent to destroying the shares and reduces the outstanding share count as any company shares owned by the company itself are really owned by the external shareholders.
The shareholders have now had their direct cost removed and their ownership of the company has been restored to the original levels. They now suffer the indirect cost of the $10K expended by the company to repurchase the shares.

Summary -- At the end of CASE I and at the end of CASE II, part 2, both the company and the shareholders end up in exactly the same place. The number and distribution of shares is the same and the company is out $10K.

The question is why these two cases end up with different accounting treatments when only the order of actions is different between them. The answer is that the treatment should NOT be different. The only way to treat the two cases the same is to record a company compensation expense of $10K when the company repurchases shares to reverse the ownership dilution of the shareholders as a result of employee grants of new shares. This is complicated by the fact that shares can be repurchased for other reasons, primarily when shares are thought to be undervalued and worthy of investment and when repurchasing shares represents a tax-advantaged alternative of distributing unneeded cash to shareholders instead of taxable cash dividends.
This complication should not be fatal.

By following the above procedure, the employee compensation expense is either borne directly by shareholders in ownership dilution or indirectly by a reduction of company value. It is an equivalent expense of $10K in either case, and must not be recorded as both a direct and an indirect cost at the same time. Until the company actually repurchases shares, if it chooses to do so, the company itself cannot be charged with a compensation expense.

Regards, Don

Joe Kristan writes:

Don, your modification is interesting, but I believe it fails to meet the accounting goal of "matching" an expense with the period in which it occurs. By deferring the expense to an eventual repurchase (which may never happen), you will deduct the expense years after the work was done.

It still seems that breaking the transaction out into it's implied cash components gets a more straightforward and economically accurate reflection of the expense.

However it's treated on financials, the smartest investor I've met doesn't much look at reported earnings. He subscribes to the "cigar box principle"- how much cash is in the cigar box at the end of the day? Reported earnings are useful only to the extent they help the investor answer that question.

Don Lloyd writes:

Joe K,

"Don, your modification is interesting, but I believe it fails to meet the accounting goal of "matching" an expense with the period in which it occurs. By deferring the expense to an eventual repurchase (which may never happen), you will deduct the expense years after the work was done."

Actually, the expense is not being deferred at all. All the repurchase does is move the expense from the shareholders directly back to the company. The problem now is that companies can start with the expense falling directly on the shareholders and then move it back to the company without recording it. If the companies actually had to record buybacks for this purpose as expenses, they would become extremely reluctant to do so. This in turn would mean that the extent of dilution over time would not be able to be disguised by ongoing 'free' repurchases.

Regards, Don

Bernard Yomtov writes:

Kyle,

I didn't say that the cost of the milk was $2.08 plus the apples plus the bread, etc. I said that the $2.08 "genuine" cost is an opportunity cost because spending $2.08 on milk means you give up the chance to spend it on something else. That's exactly what it means to spend money.

It's important to think in terms of opportunity costs because that helps you make decisions about spending. Is $87 bilion spent on Iraq worthwhile? Well, that depends on comparing what we get for it with whatever else we could get for $87 billion. I'm not trying to hijack us into a discussion of Iraq, just saying that spending decisions are choices. The opportunities we give up are the cost.

Kyle Markley writes:

Bernard,

I don't think it's useful to think in terms of opportunity costs because it causes confusions in exactly this issue.

Where is the cost in awarding stock options? There is none _to the company_. The cost is felt, simply and only, by existing stockholders through dilution.

When the PDF authors say "it bears an economic cost equal to what an outside investor would pay for the option" they're not talking about cash. They're talking about some nebulous opportunity cost that is not a _real_ cost.

The company actually creates the stock out of thin air, it doesn't cost them anything, zip, zilch, nada. The cost is $0.00. From where comes the idea that the cost is equal to what an outside investor would pay for the option? That outside investor has absolutely nothing to do with the situation.

It sneaks in under the guise of opportunity cost -- the company could have sold an option to an outside investor. The cash they could have had by selling it is their opportunity cost, but this isn't a real cost.

If I have the authority to give away someone else's property, and I do, _their_ net worth is affected but _mine is not_.

Bernard Yomtov writes:

Kyle,

Suppose a company has equipment lying around that it is not using. The equipment could be sold for, say, $1000. But the company gives it to an employee instead. Would you say the company incurred no cost?

(To avoid confusing accounting costs with economic costs, assume the equipment has been fully depreciated and has zero value on the company books.)

Joe Kristan writes:

Bernard is exactly right. It is a myth that there is no cost to issuing an option. Any time that you give a piece of the company for less than fair value, you lose something - you incur an expense.

If you own a $200,000 house and write an option enabling somebody to buy half of it for $1, good luck convincing your wife that you have created wealth out of thin air.

Ownership interests in a corporation work in much the same way. While the diffuse ownership of the corporation makes it hurt less, you are still giving away valuable equity.

Don Lloyd writes:

Joe K.

"If you own a $200,000 house and write an option enabling somebody to buy half of it for $1, good luck convincing your wife that you have created wealth out of thin air."

Your example can easily be used to show just how illogical the expensing of stock options can be.

Start with your $200K house. If you give a one half ownership to Joe Smith, the following is what happens if you follow your expensing scenario --

1. Your share of the house is now 1/2.
2. Joe Smith's share is now 1/2.
3. The house value has been reduced by $100K (the transfer of half of the original value) to $100K.
4. Your 1/2 share of the $100K house is now $50K.
5. Joe Smith's 1/2 share is now also $50K.

This is, of course, nonsense.

It should be clear that the value of the house has no reason to be reduced at all. It is exactly the same with a company. A company's value is not affected by who owns how much of it.

Regards, Don

Joe Kristan writes:

Don,

The option reduces the shareholder value just the same as a cash expense. We must remember that the shareholders are the owners, just as the married couple owns the house in the example. Giving away shares of the stock is conceptually the same as giving away part ownership of each corporate asset, as Bernard notes.

Regards,
Joe

Don Lloyd writes:

Joe K,

"The option reduces the shareholder value just the same as a cash expense...."

That's exactly the point, but it only does it ONCE.

There is a fixed logical relationship between an economic asset and its owner.

Assume that I have a business envelope that contains two $5 bills. Because it exists and I own it, it represents $10 of my net worth.

If I open the envelope and take out one of the $5 bills and give it away, lose it, or destroy it, my net worth has been decreased by $5.

If instead I grant my brother a half share in the contents of the envelope, my net worth is once again reduced by $5. The contents of the envelope is NOT changed by the fact that its ownership is now split. My net worth is ONLY reduced by my loss of total ownership. If the value of the contents of the envelope were decreased by a change in ownership, or for any other reason, then my net worth would fall even further.

When assets fall in value, their owners are also directly impacted. The reverse is NOT true. I.E., a loss or redistribution of ownership share does not affect the value of the asset in any way.

Regards, Don

Joe Kristan writes:

Don,

I guess the difference is that I don't draw the distinction you do between the envelope and its owners. I think of shares as merely a convenient way to share the envelopes contents. Out of respect for Mr. Kling's bandwidth, and the tolerance of his readers, we should probably exchange any further notes via email.

Regards,
Joe

Guni writes:

Hi there, this comments are very usefull . I love to know further on how the discussion went on between Joe and Don. Cheers.

Comments for this entry have been closed
Return to top