Kip Hagopian

STOCK OPTION EXPENSING:

DEBATE BETWEEN KIP HAGOPIAN AND PROFESSOR MARK RUBINSTEIN

This debate occurred via e-mail, from December 13, 2005-January 31, 2006, between Kip Hagopian and Professor Mark Rubinstein, Paul Stephens Professor of Applied Investment Analysis, The Haas School of Business, University of California, Berkeley. Please note: In this e-mail correspondence, Dean Tom Campbell, Bank of America Dean, The Haas School of Business, acted as intermediary between the parties. Professor Rubinstein’s opening remarks are in response to a January 2006 draft of a paper, entitled, “Expensing Employee Stock Options is Improper Accounting”. The paper was published in the summer edition of the “California Management Review”. The principal author of the paper was Kip Hagopian.

Mark Rubinstein (to Tom Campbell): I think the report you gave me today is very poorly thought out and contains a number of misstatements, which destroy its argument. There is so much wrong with it that a complete response would be quite lengthy. What I will do instead is discuss the general point it seems to make and to extract certain statements and try to show what is wrong with them. The report is so "wrong" that I cannot help but believe it was written with a prior agenda in mind. The report has an impressive number of signers, including embarrassingly Milton Friedman whom I think should know better. You yourself are not primarily an economist and can be somewhat forgiven. Note that the signers do not include a single financial economist. I believe that all past Presidents of the American Finance Association from 1970 onward who are still living would not have signed it, for example.

The basic point that is completely missed is that ESOs are largely given as a substitute for cash compensation and therefore need to be treated in a way similar to cash compensation. At-the-money options are valuable to the employee and we can assume they are not given without the firm assuming there will be some benefit to it, much like cash compensation. At the moment of the grant, not at the moment of exercise, something valuable (just ask any smart employee; would be give these options up for nothing at grant?) has been given and something valuable is naturally expected in return (otherwise the firm would be foolish). What the firm expects I think is that the employee will work harder and/or be less likely to voluntarily leave the firm (which spares the firm hiring and retraining costs).

The Treasury Stock Method does not capture this initial value because it requires no change to the number of shares as long as the option remains at the money (and has other less obvious problems). So that does not deal correctly with the economics of the firm. Something valuable has been exchanged (options for labor services) and the Treasury Stock Method records zero.

The essential observation is that if one firm pays cash compensation and the other compensates with options and if the options are not expensed in some way, the EPS of the two firms will be different and those who focus on EPS will be mislead into thinking one firm has more valuable stock than the other. We want to help people so they will not make this mistake and making an adjustment to the EPS of the firm using options seems a reasonable way to go (whether the adjustment happens in the numerator as an expense, or in the denominator as an increase in the number of shares). Which is best is another issue, but the paper you gave me says do neither.

Here are some misleading statements:

(1) "By its terms, therefore (assuming it is granted at the money) only the spread in an ESO can be realized (not the call premium) and only the original recipient of the ESO can realized that spread."

(2) "In our opinion, ESOs ... are not granted for services but are incentive compensation ..."

(3) "Would anyone other than an employee buy an option that loses its cash value the moment it is purchased, cannot be resold, and can be cancelled at any time? The answer is clearly no."

Comment on (1): Whether or not the spread can be realized is not crucial as the paper goes on to suggest. This is easy to see. When the option is granted at-the-money, there is no spread and yet something of value has been given and something of value is expected to be received (by my above argument). Every financial economist knows that the ability to realize the value of an asset immediately is usually primarily an issue of its liquidity, not its value. True, liquid assets are somewhat more valuable than illiquid assets, but liquidity is usually not the primary determinant of value. We have plenty of models in finance which correctly attribute significant value to at-the-money call options which cannot be exercised or traded prior to expiration (I myself have developed a specific one). Assets primarily take on value from the potential for positive cash flows at some future date. At-the-money ESOs certainly qualify since they can only lead to positive cash flows and never negative cash flows in the future. It would be stupid not to think they are worth something now even if one must have some patience before the contingent cash flows occur.

Comment on (2): Suppose I told my employees that I was not going to pay them any cash salary but that their entire compensation would be in the form of initially at-the-money stock options with terms set so that they have equal value to their former cash compensation. Would the employees now quit because they thought they were not being paid? I don't think so. ESOs are clearly granted for services in this example, and retain a component of that even if they are only part of overall compensation.

Comment on (3): The answer is clearly yes, not no. What the option would be worth would in part depend on whether it was likely to be cancelled. If there were any chance at all it would not be cancelled if it ended up in the money (as the scenario seems to suggest), I would be willing to pay something for it. The wording is, therefore, very misleading since it suggests, I would pay nothing. Technically, the statement is wrong.

Also what could “the option loses cash value at the moment of purchase" possibly mean? If, as the paper seems to maintain, the at-the-money option is not worth anything then how could it lose cash value? If it is worth something because of the usual option premium then the economics of call options is just the reverse: the option is in fact likely to increase in value over time, not decrease -- in general people would not buy them if this were not true. So this statement is ridiculous as far as I can tell and I suggest the author has a fundamental misunderstanding of options.

I could go on, but it feels as if I am beating up a dead horse. Admittedly, I have not dealt with all the arguments of the paper, but the above I think deals with the key one.

Mark Rubinstein to Tom Campbell in an addendum to the above:

As you know, extreme cases are often useful to prove a point. Suppose you were looking at two firms, both reporting the same EPS and both with 10,000 shares outstanding. Neither firm has adjusted its ESO for employee stock options. The firms are identical (same revenues, etc.) except one firm has no outstanding ESOs and the other has 1,000,000 outstanding at-the-money ESOs. Clearly, the firms need different share prices. The share price of the first firm should be much higher than the second. I think that it would be best if EPS could reflect this difference. The Treasury stock method would not. I would have to read the fine print to find out about the overhang of options. But a method that sees the ESO holders as potential shareholders would and this could be accomplished either by reducing earnings or increasing the number of shares.

Kip Hagopian responds to Mark Rubinstein (via Tom Campbell): I appreciate your forwarding Professor Mark Rubinstein's critique of the paper on ESO expensing to which you are a signatory. Here is my response.

Let me start with a few general comments. 1) I believe a lot of my differences with Mark have to do with vocabulary and our different conceptual frameworks. This is a complex matter that mixes accounting, economics and common sense, but ultimately, it is accounting principles that must prevail. These are often compatible with financial economics but not always. I am focused on accounting because that is what is at issue in this debate. 2) The arguments Mark makes below are some of the most common arguments made by proponents of expensing. I didn't cover them all in the paper because the paper was intended to present "first principles" in an accounting and logical sense; hence no detailed rebuttals of the standard arguments should be necessary. I did, however, address virtually all of Mark's points (and other pro-expensing arguments as well) in my comment letter to the FASB in March 2004. For your information and his, I have attached that comment letter. 3) I think it is curious that Mark selected out a few statements in the paper that didn't comport with his views of why options should be expensed, rather than critiquing the fundamental thesis embodied in the paper. That said let me respond as briefly as possible.

In summary, the thesis of the paper is this: An ESO is a gain-sharing instrument; in a gain sharing instrument, there can be no cost unless and until there is a gain to shared; the cost must be located where the gain is; the gain to be shared is stock appreciation; the stock appreciation is reaped by the shareholder and not the entity; the economic cost of an ESO is fully accounted for by the treasury stock method (TSM); neither the grant nor the vesting of an ESO meets the standard accounting definition of an expense. Attached for easy reference is the latest version of the paper.

I understand that Mark doesn't believe the TSM is adequate to measure the cost of an ESO (more on that later), but I hope he will explain what other of the above statements he disagrees with and why. For example, does he believe that an ESO is not a gain-sharing instrument? (It will be astounding if he does, since ESOs are essentially designedto be gain-sharing instruments.) Does he believe that the employee share of stock appreciation is not a cost to the shareholders? (If so, he stands alone against the entire accounting industry, which believes that the treasury stock method properly measures the economic cost of dilution to shareholders.) Does he believe that the cost of a gain sharing instrument should not be located where the gain is? (If not, where should it be located?). Since he believes that the grant of an ESO is, in and of itself, an expense to the granting entity (is that in addition to the cost to the shareholder due to dilution??), I hope he will explain what specific asset account has been reduced or what specific liability account has been increased when an ESO is granted. Put another way, I hope he will answer the following question: If the granting entity is liquidated the moment after ESOs are granted (before a rise in the stock price), will the assets distributed to the preexisting shareholders be less than at the moment before the ESOs were granted? If the answer is no, how can there be an expense? If the answer to that is that the grant of an ESO without consideration is an opportunity cost, then I hope he will explain what he means by that; or simply tell me what he disagrees with in the "opportunity cost" section of the paper (which starts on page 7).

The following are excerpts from Professor Rubinstein's comments on the paper in which Kip Hagopian has interspersed detailed answers:

MR: "The basic point that is completely missed is that ESOs are largely given as a substitute for cash compensation and therefore need to be treated in a way similar to cash compensation."

KH: I am not sure what point Mark is making. Is he saying that employees accept less cash comp in return for ESOs? If so, see below (which is excerpted from the attached comment letter).

"With the exception of early stage companies, there is little evidence that employees are required to accept lower cash compensation in exchange for ESO grants. In fact, in a study of 490 public companies published in 2000 by Rutgers professors Blasi, Kruse, and Sesil and Professor Kroumova of the New York Institute of Technology, the authors found 'no systematic evidence of any kind that companies that adopted broad-based stock option plans reduced their fixed compensation in any significant way.' Indeed, they found that on average the fixed wage compensation levels of this large sample of companies was 8% higher than comparable firms in the same industries that did not offer broad-based plans. This 8% differential was found to exist in the mid-eighties before the broad-based plans were adopted and the differential remained after the adoption of the plans and through the mid-nineties, the end of the study period."

"But surely it is true that some employees accept less cash compensation in return for ESOs. This does not mean, however, that ESOs should be expensed. Many employees make tradeoffs when agreeing to salary compensation. An executive may either implicitly or explicitly accept a lower salary in return for the opportunity to earn a bonus based on his or her performance, or on the overall performance of the company. A salesperson may accept a lower salary (or no salary) in return for the opportunity to earn commissions on his or her completed sales transactions."

Under GAAP, bonuses and sales commissions are only charged if and when they are earned, not when the bonus or sales commission arrangements are entered into. This is so, despite the fact that it would be infinitely easier to come up with an "expected value" of a bonus or sales commission than it would be to value an ESO (notice I didn't say "option").

However one interprets Mark's statement, it does not follow that ESOs "need to be treated in a way similar to cash compensation". Since bonuses and sales commissions are settled in cash by the company, they are an expense (if and when incurred) to the company. In the case of an ESO there is not a cash outlay (and not an opportunity cost) by the company and, thus, no expense.

While it is true that ESOs are sometimes given as a substitute for cash compensation, the primary reason ESOs are granted is to stimulate employee productivity for the purpose of increasing shareholder value over and above what might be achieved by salary alone. I don't believe that "The basic point ...is completely missed."

MR: "At-the-money options are valuable to the employee and we can assume they are not given without the firm assuming there will be some benefit to it, much like cash compensation. At the moment of the grant, not at the moment of exercise, something valuable (just ask any smart employee; would be give these options up for nothing at grant?) has been given and something valuable is naturally expected in return (otherwise the firm would be foolish)."

KH: I totally agree that ESOs have value and I state that clearly in the first full paragraph on page 10 of the paper, which is printed below:

"The case for expensing is intuitively appealing. When an employee is granted an ESO, he or she does in fact receive something of value. The value is not realizable and is highly conditional, but it is meaningful, otherwise, the employee wouldn't want it. It is natural, therefore, to assume that there must be a cost that is incurred somewhere that corresponds to this value. But because something has a value to the recipient does not mean it has a cost to the donor."

The only thing that is important in this debate is not whether the grant of an ESO has value to the recipient (it does) but whether it is a cost to the granting entity (we know it has a cost to the shareholders, but only if there is a gain to them). Again, I would ask Mark to explain what asset account is reduced or liability account is increased when an ESO is granted. If he says the grant of an ESO is an opportunity cost (this is, in my opinion, the only argument he can make) I hope he will explain what the firm's opportunity cost is. This seems to be the crux of my disagreement with Mark. He believes the cost of an ESO is a cost to the granting entity and I believe it is a cost to the shareholders (but only if the shareholders have a gain). In considering this, I hope Mark will review the analogy to venture capital and hedge funds that is contained in the paper: It is printed below:

"This gain sharing arrangement between employees and stakeholders is not unique to ESOs. A classic form of a gain-sharing arrangement is embodied in almost all venture capital and hedge fund partnerships. Venture capital and hedge fund managers typically receive 20% of the profits derived from investing their limited partners' capital. When these partnerships are formed, the gain-sharing arrangement is consummated. It should be instructive that there is no provision in GAAP to charge an expense to either the enterprise (the partnership) or the individual investors (the limited partners) when these gain-sharing arrangements are entered into. Moreover, if and when gains are taken by these partnerships, GAAP requires that the costs (i.e., the share of profits allocated to the managers) be accounted for on the limited partners' books (as a reduction of their profits), not on the books of the partnership. The same accounting treatment should apply to ESOs, but under FAS 123R it does not."

"(NOTE: The allocation of 20% of the profits of a venture capital or hedge fund is the exact economic equivalent of granting the investment managers a non-transferable option on 20% of the partnership's assets at cost. The limited partners of a partnership are conceptually identical to the shareholders of a corporation, and the capital of an investment partnership is no different in substance than the assets of a corporation.)"

Does Mark think there is any conceptual difference between a partnership and a corporation from an accounting viewpoint? Is there a substantive difference between the "share of profits" contract in a hedge fund and an ESO that is granted at the money? Is there any meaningful difference between the collection of assets under management by corporate employees and the assets of an investment fund?

MR: "What the firm expects I think is that the employee will work harder and/or be less likely to voluntarily leave the firm (which spares the firm hiring and retraining costs). The Treasury Stock Method does not capture this initial value because it requires no change to the number of shares as long as the option remains at the money (and has other less obvious problems)."

KH: This is true and it is a good point. But this is not an argument for expensing; it is an argument for more disclosure, particularly for unsophisticated investors. The possible impact of future dilution (due to changes in the stock price) is not a matter for accountants to deal with; it is for the marketplace to sort out. In order to facilitate this marketplace analysis, I have made a revision to the paper (suggested by Harry Markowitz, Nobel prize winner in economics and a recent signatory to the paper), adding a proposal that a prominently placed disclosure table be included in the financial statements that shows what the potential dilution would be under a range of different prices for the stock (both up and down).

In Mark's hypothetical case (Note: This is contained in Mark’s “addendum” above.) of a company that had 10,000 shares outstanding and 1 million ESOs, this table would show investors what the impact of dilution would be as the share price changes. (By the way, there are conditions under which investors would not punish the stock of this hypothetical company if it had this dilution overhang.)

This might be a good place to talk about the view of economic cost as expressed in the paper. Here is what I say about cost in the paper:

"The objective of an ESO plan, like any other incentive compensation program, is to increase employee productivity, and in so doing, raise the market value of the enterprise to a level in excess of the value that might be attained if the plan did not exist. If this objective is achieved, the use of an ESO plan will result in a net economic gain to shareholders, not a cost. Nonetheless, for the purposes of this paper, shareholder value transfer is treated as a cost because it reduces the preexisting shareholders' gross gain."

Of course it is possible that so many ESOs could be issued that there might be a net economic loss to the shareholders, meaning that the amount of the dilution from ESOs would outweigh the value of the incentive effects of the ESOs. In this case, the shareholders would be worse off because of the excess dilution. But in theory, and I believe in practice (having spent 30 years granting ESOs), in the vast majority of cases, the outcome will be a net gain (at least the shareholders believe that will be the case or they would not allow the issuance of so many ESOs). Again, this is not a matter for accountants to determine. The marketplace will decide.

MR: "So that does not deal correctly with the economics of the firm. Something valuable has been exchanged (options for labor services) and the Treasury Stock Method records zero."

KH: That's because the TSM is a "snap shot" of dilution at a particular point in time. This snap shot is quite accurate and reliable (except that it includes unvested ESOs which causes an overstatement of dilution and, concomitantly an understatement of earnings) at recording the true economic cost of ESOs at a particular moment. The TSM calculates, pro forma, the total assets that would be available for distribution if the company was liquidated, and how many shares are owned by each shareholder of record. This is "mark-to-market", exercise-date accounting done for the shareholder accounts. And the shareholder account is where the cost is. Any charge to the entity would be on top of this cost to the shareholders. This seems to me inappropriate on its face.

MR: "The essential observation is that if one firm pays cash compensation and the other compensates with options and if the options are not expensed in some way, the EPS of the two firms will be different and those who focus on EPS will be mislead into thinking one firm has more valuable stock than the other."

KH: This argument is refuted on page 20 of the attachment. Here is the most pertinent paragraph:

"But should there be a concern about comparability as between two competing companies, if one issues ESOs and the other does not? After all, the existence of ESOs in a company's capitalization represents the potential for dilution at some point in the future. While this is a fact that investors should consider in their analysis of a company, it is not an issue that should concern accountants. It is common for companies competing in the same markets to use different business models and different compensation models. For example, some companies sell their products direct to consumers and others sell through distributors. The income statements of two such companies will look markedly different, particularly with respect to gross margin and selling cost. Or, more on point, consider the case of two commercial brokerage firms, one of which compensates its sales force with salary only, while the other pays no salary, choosing instead to pay commissions based on completed transactions. In neither of these examples is it the accountant's' job to reconcile the statements of the competing companies."

Note that this statement doesn't address whether ESOs are an expense; only that "comparability" is not a reason to make them an expense.

MR: "We want to help people so they will not make this mistake and making an adjustment to the EPS of the firm using options seems a reasonable way to go (whether the adjustment happens in the numerator as an expense, or in the denominator as an increase in the number of shares)."

KH: I would agree with this if ESOs were an expense to the granting entity, but they aren't.

MR (lists what he terms "misleading statements” by KH): (1) "By its terms, therefore (assuming it is granted at the money) only the spread in an ESO can be realized (not the call premium) and only the original recipient of the ESO can realize that spread."

(2) "In our opinion, ESOs ... are not granted for services but are incentive compensation ..."

(3) "Would anyone other than an employee buy an option that loses its cash value the moment it is purchased, cannot be resold, and can be cancelled at any time? The answer is clearly no."

"Comment on (1) Whether or not the spread can be realized is not crucial as the paper goes on to suggest. This is easy to see. When the option is granted at-the-money, there is no spread and yet something of value has been given and something of value is expected to be received (by my above argument). Every financial economist knows that the ability to realize the value of an asset immediately is usually primarily an issue of its liquidity, not its value. True, liquid assets are somewhat more valuable than illiquid assets, but liquidity is usually not the primary determinant of value."

KH:This not about liquidity; it is about "realizeability". As I clearly state, the call premium in an ESO cannot be realized. Only the spread can be realized. This is part of the ESO's design. If the call premium could be realized, the ESO would not be a pure gain-sharing instrument. To wit: The ESO holder would be able to realize a profit even if the shareholders didn't gain. Moreover, since the ESO cannot be transferred, only the original recipient can realize the spread. So what is misleading about that statement? I have not opined in this statement as to what the ESO might be worth to the holder (his own implicit call premium); only that whatever it is, it cannot be realized during the life of the ESO. Accordingly, only the ESO holder's share of the gain (if any) can be realized. Again, from an accounting viewpoint, the hypothetical value of the ESO to the holder is not the same as the cost to the entity.

On page 12 of my comment letter to the FASB (attached) I point out why "realizability" is so important in accounting:

"The critical importance of the concept of "realizability" to the recognition of "revenues and gains" is definitively stated in paragraph 83 of "Statement of Financial Concepts No. 5", entitled 'Recognition and Measurement in Financial Statements of Business Enterprises': 'Revenues and gains of an enterprise during a period are generally measured by the exchange values of the assets (goods or services) or liabilities involved, and recognition involves consideration of two factors, (a) being realized or realizable and (b) being earned, with sometimes one and sometimes the other being the more important consideration'....Revenues and gains are realized when....assets are exchanged for cash or claims to cash. Revenues and gains are realizable when related assets received or held are readily convertible to known amounts of cash or claims to cash (emphasis added)." This pronouncement makes it clear that the receipt of an instrument, whose value is not realizable, would not be recognized under GAAP as a gain to the recipient. By only modest extension of this logic, the grant of such an instrument should not result in a loss or an expense to the grantor."

MR: "We have plenty of models in finance which correctly attribute significant value to at-the-money call options which cannot be exercised or traded prior to expiration (I myself have developed a specific one)."

KH: Again, we are not talking about value to the recipient, only cost to the entity. As a venture capitalist, I assure you I ascribe significant value to my share of profits in my venture partnerships, even though I am not allowed to sell my contracts. Is Mark saying that these partnership entities should take a charge at the closing of the partnership financing? (How have the partnership assets been affected by my profit sharing agreement with the limited partners?) Or should the individual limited partners write down their assets on the day they invest? (If you were a limited partner, would you believe your investment was worth less the moment you invested in the venture partnership? If so why would you invest?) GAAP would not sanction either one of these actions.

By the way, I would be interested in hearing more about the option Mark has developed. For example: Is this option cancelable at the will of the issuer as is the case with an ESO? Is it issued only to employees? (If it is used for financings, it isn't relevant to this discussion.) If these are the terms of the option (granted only to employees; granted at the money; prohibited from being transferred, hypothecated or assigned), it would be an ESO and should not be expensed by the entity that grants it.

MR: "Assets primarily take on value from the potential for positive cash flows at some future date. At-the-money ESOs certainly qualify since they can only lead to positive cash flows and never negative cash flows in the future. It would be stupid not to think they are worth something now even if one must have some patience before the contingent cash flows occur."

KH: As I have said explicitly and clearly in the paper and have said repeatedly in this email, I agree with that. The fact that Mark didn't notice my statement to this effect, makes me realize I need to highlight my position on this so it isn't missed. This is something I have learned from this email.

MR: "Comment on (2): Suppose I told my employees that I was not going to pay them any cash salary but that their entire compensation would be in the form of initially at-the-money stock options with terms set so that they have equal value to their former cash compensation. Would the employees now quit because they thought they were not being paid? I don't think so. ESOs are clearly granted for services in this example, and retain a component of that even if they are only part of overall compensation."

KH: Mark makes a good point here. My statement is not very artfully worded. Obviously "services" are necessary to deliver the results expected in an incentive compensation device, whether it is gain sharing or not. But services per se, while necessary, are not sufficient to produce that result. For example, consider the contingency lawsuit; a law firm may render intensive services but not win the case. In that instance, there would be no payment to the law firm and no cost to the client. That is what I was trying to say but did not say well (I said it better in my FASB comment letter). I will reword this statement. I may also add to it by pointing out some of the logical problems of FAS 123R with respect to its tying of "services" to expenses. Two examples: (1) Assume an ESO doesn't vest for 5 years (this is quite common for existing employees). If the employee leaves at the end of the 59th month, no expense will be charged to the company despite the 59 months of services rendered by the employee. (2) Assume a new employee is granted an option that in part is fully vested on the date of grant (rare but not unheard of). That company will be charged an expense without receiving any services. I would be interested in Mark's thoughts on this. Perhaps I am missing something.

MR: "Comment on (3): The answer is clearly yes, not no. What the option would be worth would in part depend on whether it was likely to be cancelled. If there were any chance at all it would not be canceled if it ended up in the money (as the scenario seems to suggest), I would be willing to pay something for it. The wording is therefore very misleading since it suggests I would pay nothing. Technically, the statement is wrong."

KH: The statement is correct. At the bottom of the first full paragraph on page 9 of the paper, I state: "...ESOs are cancelable at the will of the issuer (by terminating the employee)...". Since we are talking about issuing "similar options to third parties" and not employees, there is no reason for the issuer not to cancel the options immediately after issuance. Thus, there is no chance at all they would not be cancelled. I am on the board of a public company. I would be happy to sell Mark options on the company's stock with identical terms to our employee options (priced at the money, five year vesting, etc). I assure you those options would be canceled immediately. If I have missed something here, I hope Mark will correct me.

Having said all of the above, the issuance of "similar options to third parties" is basically a non-issue and should not even enter the debate. ESOs are intended for employees, directors and consultants, not "third parties". They are designed as incentive compensation for people who can, through their efforts, cause an increase in the value of the enterprise. The notion of issuing them to third parties is nonsensical. So is the notion of selling them to the people (employees) that the company is trying to motivate. The company cannot incur an opportunity cost for not doing something that is contrary to its best economic interest, can it?

MR: "Also what could 'the option loses cash value at the moment of purchase' possibly mean?"

KH: I assume this is the statement Mark is referring to:

"...as stated above, only the spread in an ESO can be realized. This means that if a buyer paid a cash premium for an at-the-money option at the grant date, the cash realizability of the call premium would disappear at the time of purchase."

I don't know what he objects to in this statement. It is very carefully worded. The key words are "cash realizability". Assume I sell you an at-the-money, non-transferable option for a cash premium. Assume the option we are talking about has a $10 strike price, and I sell it to you for $3. Assume the option is vested so it can be exercised immediately. Assume also that the stock rises to $10.01 so it has a $.01 spread. In a tradable option (unless the call premium changed in value for some reason), you would be able to sell the option for $3.01. In the case of an ESO however your only means of realizing cash from the instrument is to exercise the option and sell the underlying stock, thereby realizing only $.01. The cash realizability of the call premium has been eliminated, has it not? There would still be an "implicit" call premium but not one that could ever be realized by the holder. In any event all of this is moot because ESOs, like hedge fund management contracts and contingency lawsuits, are not sold. To do so would defeat their purpose.

MR: "If as the paper seems to maintain, the at-the-money option is not worth anything..."

KH: The paper doesn't say that.

MR: "...then how could it lose cash value?"

KH: Note the above quote from the paper. What the paper says is that if someone paid a cash premium for the option, the cash realizability of the call premium would disappear at the time of purchase. This is explained in my comments above.

MR: "If it is worth something because of the usual option premium then the economics of call options is just the reverse: the option is in fact likely to increase in value over time, not decrease -- in general people would not buy them if this were not true."

KH: I agree. I don't believe I have said anything in the paper or in the above remarks that is inconsistent with that.

MR: "So this statement is ridiculous."

KH: Good thing I didn't make that statement.

MR: "I could go on, but it feels as if I am beating up a dead horse. Admittedly, I have not dealt with all the arguments of the paper, but the above I think deals with the key one."

KH: It is clear that Mark and I are not going to agree on this subject but at least I hope I have clarified my position somewhat. Based on Mark's comments I have identified two areas of the paper that could be improved. As a result, I intend to make some changes. But these changes are minor and will not in any way change the basic thesis of the paper.

I say again: The issue in this debate is not whether an ESO has value; it whether an ESO is a cost to the entity that grants it. Unless a transaction results in a decline in value of an entity (e.g., a decrement to an asset account) it cannot be an expense. An entity cannot incur an opportunity cost for not doing something that is contrary to its best economic interests. The cost of a gain-sharing instrument--whether it is a cost to the entity or the shareholders--should not be recorded unless and until there is a gain. If the gain is taken by the enterprise (as in the case of a contingency law suit), that is where the cost should be charged. If the gain is taken by the shareholders (as in the case of an ESO), that it where the cost should be reflected.

The following are excerpts from Professor Rubinstein’s response (via Tom Campbell) to Kip Hagopian’s rebuttals (above) interspersed with Kip Hagopian’s further rebuttals: 

MR: I don't really want to take the time to continue this. I know that might sound like sour grapes, but it really is just that it will take a lot of time. And when it is over, Kip will probably continue to think what he does now. I continue to think Kip is way off base. Doing this in writing is too time consuming. I would be happy to have a conversation with him over lunch let's say and you could be there and maybe that would help you understand this better.

It is too bad you still can't see what is wrong with what Kip is saying. I've included Richard in this email because I know he is interested. I hope that was ok. If we met, I would hope Richard could join us.

Kip Hagopian (to Tom Campbell) responds: 26th I would have preferred that Mark ask me before circulating the paper to others. Has he sent it to anyone else? I would appreciate it if he would not send it to anyone else without talking to me or to you. Having said that, I am perfectly fine with Professor Stanton having read the paper. Moreover, I would welcome any comments he might have, unless Mark is speaking for him.

MR: I can't resist a short response, which I realize however is terribly inadequate to the number of issues raised.

First, the suggestion that I am an economist and don't understand accounting is laughable. I had three wonderful and penetrating courses in accounting from Robert Sprouse at Stanford (who used to be a member of FASB) and I have read several books about accounting-practice and theory. I have published one paper on accounting (albeit about ESOs) and have a long general section of my forthcoming book on the subject. I think the added fact that I am a financial economist actually gives me an advantage since most accountants do not have the perspective.

KH:  There is nothing I can find in my remarks that suggests that Mark doesn't understand accounting. In fact I was quite aware before I wrote those remarks that he is well educated in accounting as well as economics. Mark's comments on our paper seemed to focus on the economic aspects of the expensing issue (which is appropriate because the economic side of this issue is extremely important in determining the basis for rational accounting). But I was trying to point out that good economics doesn't always comport with traditional accounting practice. For example, when Intel gets an order from Cisco, I suspect that the statistical probability of that order turning in to a sale is at least 95% (this can be very closely determined by the outcome of past orders). An economist might say, therefore, that this $1 million order should be booked as $950,000 in revenue (certainly investors will make that kind of assumption in assessing Intel's business prospects). But this not allowed under GAAP. This particular example points to the issue of recognition of revenues and gains that I mentioned in my last email (where I quote from the Concept Statements) and lends support to the importance of "realizability".

MR:  I should add that there are apparently several real accountants on FASB (perhaps all of them) who would agree I think with most of my interpretations of the accounting issues and would apparently disagree with Kip.

KH: That's true. But I would be more impressed with that if there weren't so many other "real accountants" that disagree with the FASB position. Our list of signatories includes several accounting professionals that are well qualified to serve on the FASB board. Moreover you could have filled the FASB board 10 or 20 times over with the list of qualified accountants that vigorously opposed expensing 12 years ago. This list includes the current FASB Chairman, Bob Herz, who was at PW at that time and helped draft their letter of comment (in fact he was listed as the person to whom responses from the FASB were to be sent). Moreover, in their comment letters to the FASB in 1993 and 1994, the "big six firms" gave very thoughtful and persuasive reasons for their opposition. As I point out in the paper, there have been no changes in fundamental accounting concepts during that time. This raises the question of whether they were right then or are they right now.

MR: So being primarily a financial economist and not an accountant has nothing to do with this. I should add that FASB pushed ahead with their new ESO rules in the face of considerable political pressure against them, which makes me believe they are not just accountants who believe ESOs should be expensed, they passionately believe this. They understand that "comparability" is a basic principle of accounting and as long as ESOs are not expensed that high principle is materially injured. Because of the difficulty of measuring the effects of many transactions, comparability remains an elusive goal, but where measurement is more or less accurate, it is worth pursuing.

KH: I covered the comparability issue in my last email. It is only an issue if ESOs are really an expense and are not treated as such. But if we determine that ESOs are NOT an expense it is not an issue, is it? The overhang of options is required disclosure for all public companies. We think this disclosure could be even more comprehensive and make proposals to that effect in the paper. If full disclosure is made, investors should have no problems making proper comparisons between companies.

MR: Second, the idea that a firm can magically convey something that is of immediate value to an employee without generally creating an immediate economic cost to its concurrent shareholders does not make economic sense.

KH: This is an important statement because it suggests that Mark and I may not be as far apart on this than it appears. Mark refers to there being "an immediate economic cost to its...shareholders." If he really means "shareholders" then we may find common ground. If he means "the company" then we disagree.

As I say in the paper, when an ESO is granted it does have immediate economic value to the recipient on the date of grant. This value is not realizable but it is implicit. The question of whether there is an immediate economic cost that corresponds to that economic value is complex. On one level, there is such an immediate cost. Since an ESO is a gain-sharing instrument, the cost must be located where the gain is, and inasmuch as the gain is shareholder stock appreciation, the cost is incurred by the shareholders. The implicit value of the ESO is the DPV of the expected profit in the option at the end of its expected term. The

implicit economic cost, therefore, is the DPV of the expected amount of the value transfer from the shareholders to the ESO holders. In other words, whatever the ESO's implicit call premium is to the employee, is also an implicit drag on the shareholders' future stock appreciation. Ergo, we have a symetrical transaction at the date of grant.

But there is another way of looking at economic cost that is even more salient than the one described above. I discuss it briefly in the paper. To wit: An ESO is an incentive compensation device that is designed to stimulate employee productivity, and in so doing, to increase the value of the shareholders' ownership to a level in excess of the value that might be attained if the plan did not exist. In order to do this, the market value of the enterprise must rise enough to offset the dilutive effects of the ESO program. If this objective is achieved, the use of an ESO plan will result in a net economic gain to shareholders, not a cost. (Note: For the purposes of the paper, shareholder value transfer is treated as a cost because it reduces the preexisting shareholders gross gain.) The point is, if ESOs are not issued in such abundance that their dilutive effects out weigh the incentive effects, there will be no net economic cost--there will be a net economic gain. (Note: The distinction between "gross" costs and "net" costs has been described well in a paper by Prof Charles Calomiris, PhD Econ at Columbia U.; I will happy to send it to you and Mark if you are interested.) Achieving a net economic gain is, of course, why ESOs are granted in the first place. I hope Mark will be willing to comment on this.

To illustrate the above point, allow me to pose a hypothetical situation. Assume you own a sole proprietorship that manufactures "widgets". For 20 years it has been appraised as having a book value and market value of $1million. Now assume you hire a new GM and, in an effort to increase the value of your company, you grant this GM an option to buy 50% of the business at its current appraised value. Have the assets in this business declined in value on the day you grant this option? Under FAS 123R they have. This is counterintuitive to me.

Now assume that the GM you have hired is highly motivated by the opportunity to make money on the option. After a lot of thought, he makes a design change to your widget-making machine that doubles its output and, concomitantly, doubles the value of your business. Now the business has increased in value by 100%. But, because of the gain-sharing arrangement (the option) with the GM, the value of your interest in the business has only increased by 50%. Have you suffered a loss due to the sharing of your gain with the GM? Or have you reaped a net gain because of the incentive effects of the option?

MR: That would be like creating something out of nothing, which, except for the Genesis story of God's creation of the universe "ex nihilo," violates classical physics. The idea that there is a contingent future cost and not an immediate cost see above does not make sense either from present value reasoning and since the employees have been given a instrument they see as immediately valuable for good reason, whether it ends up in-the-money in the future or not. If there were no immediate cost to the firm, that would suggest that firms should issue unlimited options to their employees at no cost to their current shareholders and their employees would be overjoyed.

KH: I hope I have answered this in my comments directly above. As I say in the paper and have said more than once in prior e-mails, issuing options does have economic consequences to the "current shareholders". And issuing "unlimited options" could have serious economic consequences. That is why for the entire history of ESOs, during which time they were not expensed, they have always been granted judiciously.

MR: As a shareholder, if I saw this coming, I would sell my stock as fast as I could, as I think Kip would. If it is an economic cost (at the time of grant or more generally by its vesting date, even if the options remain at-the-money) to its current shareholders, and if accountants can capture the cost in their statements in an accurate way (which is for me, by the way, where the real controversy lies), why not? In Kip's accounting world, the income and balance sheets of the firm in my example and its earnings per share would not immediately reflect this and would, by themselves, be terribly misleading.

KH: With all due respect, in “Kip's accounting world”, since what the employee gets in profit is a share of the shareholders' stock appreciation, the cost would have to be located where the appreciation occurs: on the balance sheets of the shareholders. (BTW, this cost has always been and will continue to be located there anyway. What FAS 123R does is charge an additional cost to the financial statements of the enterprise. I don't understand this.)

MR: The response that this information can be handled by a footnote is of course always possible, but is not I think, for significant matters, an acceptable way out if there is a decent way to measure the cost of the omitted material transactions. This I think accords with the basic principles of accounting, at least the ones I was taught (and that make good economic sense).

KH: It seems clear to me that where Mark and I disagree is where the economic cost (and therefore, the accounting cost) of an ESO is to be located. Since the profit an ESO holder earns comes right out of the shareholder's pocket (I believe this is an irrefutable fact. Does Mark agree?), I don't understand why Mark believes that on top of that, an additional cost should be charged to the entity that grants the ESO.

Here is something else to ponder: Since Mark believes that the "real controversy" lies in how to value the ESO, why value it at all? Why not just wait and see what happens? (The standard way GAAP deals with items that, "due to economic uncertainties are difficult to value", is to wait until the uncertainties are removed. The FASB made an exception with ESOs.) Why not simply use "mark-to-market" exercise-date accounting. That way, if the stock price goes up, the entity's expenses would go up. If it goes up a lot, the expense would eventually wipe out all of the company's profit. Alternatively, if the stock plunges, the company could report a very large increase in profits. In all of these scenarios, the outcome would be perverse and utterly nonsensical because the ESO gains and "expenses" caused by the changes in stock price would bear absolutely no relationship to the operations of the business. So my question is: If expensing the actual employee profit in the option is so clearly wrongheaded, why does it make sense to expense a difficult-to-estimate fraction of that same profit on the date of grant? Doesn't the clear fact that charging the actual spread in the option makes no sense suggest that we are charging the cost in the wrong place?

Now I will beat the dead horse, as Mark put it in a prior email. An ESO's theoretical call premium cannot be converted to cash under any circumstances. It is designed to prevent such a conversion. Consequently, there can be no expense. The same accounting logic that applies to bonuses and sales commissions, hedge fund profit sharing agreements and all other gain-sharing arrangements should apply to ESOs.

A final word. Mark disagrees with me and all of the 25 signatories to the paper. He also is at odds with the multitude of big six accountants from 12 years ago (many of whom are still around and still hold the same views) who made very solid arguments in opposition to expensing. Leaving me out of it, it seems to me that the credentials of all of these other people are worthy of some respect. Is it possible that both sides of this debate have respectable cases to make?

Once again, I appreciate this exchange of views. Even though Mark has not convinced me I am wrong (nor, I'm sure, have I convinced him), I have learned something from this that will make the next draft of the paper better.

MR: This is why Kip needs to be present at the meeting. However, maybe Tom has absorbed Kip so he doesn't need him. If so, the meeting will be useful.

I apologize for sharing Kip's writing with Richard, and please let him know that. I do this sometimes without asking if I am almost 100% certain there is no problem, and in this case I think I am right. First, I did not give Richard or anyone else his paper, I have just included him in on the emails. Second, Richard and I have been working on these issues together and he is my co-author on our recent "position paper" on this which can be found at my web site (www.in-the-money.com -- which is required reading for Tom Campbell in preparation for our meeting). Third, clearly I need all the help I can get. and Richard has a very clear and orderly mind which makes him very helpful. Fourth, I thought Richard would appreciate being involved at least a little bit.

So Tom, read our joint position paper, it deals with some of Kip's responses below, and when you are ready, let's get together.

One response to what is written below. Kip makes the claim that because ESOs are motivational devices, if they work that way, they will give rise to revenue not a cost.

KH: Not to be picky, but I never mentioned revenue. What I said is that the incentive effects of an ESO could produce a net economic gain to shareholders rather than a net cost. I also said that the nominal cost of an ESO is the employee's share of the shareholders' stock appreciation. This is a wholly different concept from revenue.

MR: To me that is like saying if I buy a factory, I can ignore the cost and just recognize the revenue.

KH: Of course that's not true. Perhaps I don't understand Mark's point here. I don't see how he can possibly make this comparison and I can't find anything I said that would lead to this inference. Mark is referring to comments I made on economic cost, not accounting cost.

MR: In the case of an ESO, although I just focused on costs, of course the firm is expecting increased revenues otherwise it would not have used the ESOs and these revenues will come in gradually over time and to some extent offset the cost, but both have to be in the financial statements. When is it ok to recognize revenues but not costs?

KH: Never. Matching revenue with expense is a fundamental accounting principle. BTW, a gain-sharing instrument adheres to this principle. The cost (employee profit in the option) is matched with the gain (the shareholder's stock appreciation).

MR: I want to recognize both.

KH: Me too.

MR: You don't have to amend current accounting rules to recognize the revenues so that is why I did not focus on that.

KH: I am confused about the point Mark is trying to make. The costs of running a factory (wages, depreciation, etc.) are obviously direct expenses of the company because the costs occur on the company's income statement. Those costs are matched to revenues of the company. The factory costs described above are direct costs to the company and indirect costs to the shareholders. ESO costs are different. The costs of an ESO are not costs to the company; rather they are direct costs to the shareholders. This is where Mark's opinion and the opinion expressed in the paper seem to differ.

If I am right as to our main point of contention I would like to ask Mark to comment on the following:

1) Mark has never challenged the basic premise of the paper. In the Exec Summary there are 6 statements, which we believe are facts that cannot be persuasively refuted. Which of these statements does Mark disagree with and why? We need to know where we are wrong.

2) In the example of the "Widget" business: When the business was sold for $2 million, the owner's share was diluted and he "only" received $1.5 million, thereby incurring a nominal economic cost of $500,000 because of the option (I say "nominal" because he actually reaped a $500,000 net gain). This cost was a mathematical outcome of the gain-sharing arrangement. The GM's profit on his option was $500,000. Had the GM's option been expensed at grant date, the expense would have equalled the DPV of the $500,000 and would have been charged to the company. Thus, the owner would have incurred both a direct cost (the GM's share of the gain) and an indirect cost (the charge to the company's income statement and balance sheet) for the same thing. What is the rationale for this?

Tom Campbell (to Mark Rubinstein):  I look forward to our meeting. I'll read your paper this weekend. And I'll convey to Kip your response regarding Richard. I recall from Kip's email, "I would have preferred that Mark ask me before circulating the paper to others… Having said that, I am perfectly fine with Professor Stanton having read the paper. Moreover, I would welcome any comments he might have, unless Mark is speaking for him." So, I'm sure Kip will accept your explanation.

Regarding your hypothetical, if a firm buys a factory, it books the cost of course. Eventually, it will book the revenue from the goods manufactured there. It doesn't book the revenue right away, because it would be speculative. But suppose the cost is speculative. Symmetry would suggest the cost not be booked either, if it is speculative. I think that is 1/2 of Kip's point. The cost of granting employee stock options is speculative. The second half is that it is a cost, but not to the company. Comparable to your example, suppose the company receives the factory as a gift from its existing shareholders. Then, I believe, one would not book any cost to the company. The existing shareholders are out the value of their gift, but the company isn't. The only oddity is that the gift of the value of the employee stock options from the current shareholders is not a voluntary gift from them. It's a gift compelled from them by the company. But it comes from them nonetheless.

Mark Rubinstein (to Tom Campbell): 27th I don't really think Kip should object to my distributing his paper since in fact I didn't, even to Richard. I hope you don't mind if I include Richard in this email since again I trust he will be joining us. Incidentally, I am delighted to see you are taking such a persistent interest in this.

I trust you realize there are two things to read, the one I sent you as an attachment which relates to accounting in general, about which I would hope there would be little controversy, and a second short position paper Richard and I have written recently on ESOs reflecting my comments at the FASB round table which you can download from my web site http://www.in-the-money.com by clicking on the Mark Rubinstein button on the home page.

I can see that you are up to speed. Your argument regarding the hypothetical is a good one and well put. I need to consider the best way to reply to it. A first answer is this: when you say an item of revenue is "speculative" that could mean two things: (1) it is hard to know its current value, or (2) it is hard to know its future value. Firms tend not to report revenues and assets of type (1) until the speculative condition has been removed. But firms do tend to report assets of type (2) provided they are not also of type (1). For example, when the firm buys a factory at that point it only faces type (2) uncertainty and it does record that as an asset even though the eventual selling price of the factory is unknown.

Similarly, for ESOs, It is argued by FASB and several financial economists that with current option valuation techniques they are only speculative in the type (2) sense, therefore they should be expensed. Actually, I am much more of a skeptic about this and my original article on ESO accounting practice (1995, I think) says that it is not true that option valuation techniques are good enough and ESOs are type (1) speculative instruments and therefore should not be expensed in the way FASB has actually proposed. However, at that time, I lacked an important insight that reduces ESOs to only type (2) speculative instruments using newly developed valuation techniques. This is explained in our recent position paper. Therefore, Richard and I come down saying they should be expensed (albeit with a different schedule than given by FASB).

A second answer is this. I like to think of stock option grants as in effect two separate transactions: the firm pays its employees in cash then some employees use the cash voluntarily to buy non-transferable stock options at fair market value from the firm. From the firm's point of view, to first order, the employees are then no different than outsiders who have purchased these options. In my perfect accounting world, the firm will report two types of earnings. Earnings based on the cash transactions (which I would call operating earnings or undiluted earnings), and extraordinary earnings based on changes in the value of the options after they are issued. This could be accomplished at least in part by fully diluted earnings. This answer also makes clear why the crucial distinction for accounting is really whether a cost or asset is type (1) speculative or not. An asset which is only type (2) speculative can always I think be reduced to an equivalent set of two transactions as I have here, and in that case, I think we would clearly want to report the cost or asset in the balance sheet and income statements. Accounting needs to be designed so that two firms, one which directly issues ESOs to its employees in lieu of salary, and one which issues ESOs in this round-about way look the same to investors.

I have not run this particular way of putting this past Richard, before writing it here, which is dangerous; and if he has any problems with it, I assume he will chime in.

In reading this over I realized at least one point is unclear. When I say in the 5th paragraph, "earnings based on the cash transactions," I mean earnings based on the cash salary payment to employees. The other transactions are handled as if the employees were outsiders.

Kip Hagopian (to Tom Campbell):  31st FYI, the "two separate transactions" concept is commonly used as a way of thinking about the issuance of stock or tradable options (I sell you an equity security for cash and then turn around and use the cash to pay you for your services) to employees. But (in my humble opinion) the concept doesn't apply to an at-the-money, non-transferable option. When I grant you an ESO, I don't want you to pay me for it. I want it to be an incentive to you to increase the value of my stock. If you are successful I share the gain in the value of my stock with you. If you pay for it, the economic value of the instrument to me will be eroded and ultimately nullified. What if you were a salesman and I wanted to incent you to sell my widgets. Would I try to sell you the right to earn a commission? Am I missing something here? Stock and tradable options are the same as currency. ESOs are not.

Kip Hagopian replied to Dean Campbell that he would be unable to go to lunch and suggested that Tom Campbell attend the lunch by himself. He then sent the following (Note: KH does not know whether Professor Rubinstein received this.):

Kip Hagopian (to Tom Campbell): Here is MY view on the accounting vs. the economics of ESOs.  As an economist, you are far better qualified than I to opine on the latter, so I hope you will tell me where I go wrong.

Assume an at-the-money ESO is granted to an employee and is vested on the day of grant.

First, the accounting perspective:  On the date of grant, no value can be realized by the recipient; therefore no recordable cost is incurred by any other party to the transaction.  But as the ESO appreciates, the spread in the ESO will become realizable and it will be recorded on the employee's books as a profit.  It will also be accounted for as cost to the shareholder in the form of a value transfer from the shareholders balance sheet to the employee's balance sheet.  In other words the value and the cost will be recorded if and when they actually occur.  As in most transactions, the profit to one party will be a cost to the other.  Thus the transaction will be symmetrical, as the cost and the profit are actually occurring.  This is the usual way accountants account.  They wait until something has happens and then they record it and report it.

Here is the economic perspective:  When the ESO is granted it has economic value to the recipient on the day of grant.  This value is not realizable but is implicit.  This implicit value is the DPV of the expected gain in the ESO at the end of its expected term.  The economic cost of the ESO is equal to the economic value to the employee.  It is also the DPV of the employee's eventual profit and it is also implicit (not realizable).  Since the value is measured in future stock appreciation, the cost will take the form of a reduction in future stock appreciation.  But where will this cost located?  Is there any place on the balance sheet of the granting entity that is a stock account that will be reduced as a result of the employee's profit?  I argue that the answer to that is no.  But we do know that if the stock of the entity appreciates, the shareholders' accounts will be reduced by the exact amount of the employee's profit.  In this way the shareholders will share their gains with employees. So at the date of grant, what is the economic cost to the shareholders?  I would argue that when ESOs are granted, the shareholders do have an economic cost equal to the DPV of the expected reduction in their expected future gain.   In other words, whatever the ESO's implicit call premium is to the employee is an implicit future drag on the shareholders' stock appreciation.   Thus we have a symmetrical transaction at the date of grant.

The real economic cost of an ESO:  There is another economic aspect to this issue that Mark has never mentioned and that I only mention in passing in the paper.  The economic costs described above are really gross costs.  They do not take into account the incentive effects of the ESO.  Here's what I say in the paper:

"The objective of an ESO plan, like any other incentive compensation program, is to increase employee productivity, and in so doing, raise the market value of the enterprise to a level in excess of the value that might be attained if the plan did not exist.  If this objective is achieved, the use of an ESO plan will result in a net economic gain to shareholders, not a cost.  Nonetheless, for the purposes of this paper, shareholder value transfer is treated as a cost because it reduces the preexisting shareholders’ gross gain."

The point is, if ESOs are not issued in such abundance that their dilutive effects out weigh the incentive effects, there will be no economic cost--there will be an economic gain.  (Note: The distinction between "gross" costs and "net" costs has been described well in a paper by Prof Charles Calomiris, PhD Econ at Columbia U.)   This, of course, is why they are granted.  The question as to which effect will win out will be determined by the market place.  Since Mark is an economist, I wonder why he never talks about this aspect.

Dean Campbell subsequently met with Professor Rubinstein and Professor Stanton. The following is Tom Campbell’s recollection of Professor Rubinstein’s comments at that meeting.

According to Tom, Mark started the meeting with something of an apology, saying that he had concluded that we know a lot more about this subject than he initially thought. He acknowledged specifically that the exchange between us had been productive, and that he had learned from it. While he stopped short of saying he agreed with our position (he still thinks some kind of expense needs to go on the entity's statement), he did say that he intended to refine his own published statements on this issue to include a recognition that if option costs are booked, the gains (stock appreciation) on which those costs are dependent should also be booked and spread over the same time period as the costs. He apparently wasn't clear on where he would book those gains (we are not sure he has thought this through; when he does he may realize that the net effect of doing this will be to record a gain, which, of course, would be inappropriate). He also indicated that he intends to state in his writings that this refinement had not been part of his thinking prior to our dialogue. In fact, he said that he might revise his paper on this subject and footnote our paper in the revised version.