Mark

                          ABACUS, Sept 1998 v34 n2 p141(21)

                               The FASB's conceptual framework and political support: the
                               lesson from employee stock options. Haim A. Mozes.

                          Abstract: The FASB's Conceptual Framework of Accounting project has not been
                          able to generate political support because it does not provide an accounting model
                          that takes into consideration various economic events. A detailed analysis of
                          accounting for employee stock options (ESOs) reveals the limitations in the conceptual
                          framework. One such limitation is the fact that there are alternative accounting
                          frameworks for the transaction in which a company grants ESOs to its employees.
                          Thus, the FASB's framework cannot always obtain support for standard-setting
                          proposals.

                          Full Text: COPYRIGHT 1998 Accounting Foundation of the University of Sydney

                          INTRODUCTION

                          The Financial Accounting Standard Board's (FASB) Conceptual Framework (CF) of
                          Accounting project was completed in 1986 at a substantial cost after ten years of
                          effort. Subsequent to the publication of the FASB's CF, the Australian Accounting
                          Research Foundation, the International Accounting Standards Committee, the
                          Canadian Institute of Chartered Accountants, the Accounting Standards Board (U.K)
                          and the New Zealand Society of Accountants began work on their own conceptual
                          frameworks. There are several hypothesized benefits from a CF of Accounting.
                          Horngren (1981) and Dopuch and Sunder (1980) argue that a CF should increase
                          standard-setters' ability to enact standards they feel are consistent with accounting
                          theory. Solomons (1986) argues that a CF should be directed towards the
                          establishment of sound principles for standard-setters to use in shaping accounting
                          practice, and Milburn (1991), DePree (1989), and Chambers (1996) argue that a CF
                          should provide a basis for standard-setters to deduce logical accounting
                          recommendations. Mozes (1992) argues that a CF is useful for organizing and
                          formulating normative accounting research and for defining the terms of debate with
                          respect to various standard-setting proposals.

                          In the U.S. (as well as in other countries), private bodies have had the major role in
                          providing authoritative guidance on the standards that seek to 'present fairly' the firm's
                          financial position and that constitute 'generally accepted accounting principles'. Zeff
                          (1995) explains that in the U.S. the 'generally accepted' clause has come to mean that
                          the principle is acceptable to the FASB and the Securities and Exchange Commission
                          (SEC), and this understanding was codified by the SEC's Accounting Series Release
                          ASR 150 (SEC, 1973). In the majority of cases, the SEC followed the FASB's
                          guidance. However, the FASB was forced to adopt reporting rules which they
                          explicitly acknowledge to be imperfect in several politicized standard-setting issues.(1)
                          For example, par. 5 of SFAS 87, Employers' Accounting for Pensions (FASB,
                          1985), states that the Board 'believes that those conclusions [i.e., SFAS 87] are not
                          likely to be the final step . . . Pension accounting in 1985 is still in a transitional stage'.
                          More recently, par. 61 of SFAS 123, Accounting for Stock based Compensation
                          (FASB, 1995), notes that the political debate over employee stock options (ESOs)
                          became so divisive that it 'threatened the future of accounting standard-setting (par.
                          60)' and the preferred rules could not be adopted. Specifically, the Board maintained
                          that

                          Financial reporting would be improved if all equity instruments granted to employees
                          ... were accounted for on a consistent basis (i.e., expensed). However, in December
                          1994, the Board decided that the extent of improvement that was envisioned ... was
                          not attainable because the deliberate, logical consideration of issues that usually leads
                          to improvements in financial reporting was no longer present.

                          One question that arises is why the FASB has not been able to utilize its CF to obtain
                          support from the SEC in controversies such as those surrounding ESOs. This article
                          provides an in-depth analysis of accounting for ESOs. It illustrates two reasons why
                          the FASB's CF could not deliver regulators' political support. First, there are
                          alternative accounting models or 'views' for the transaction in which a firm grants
                          ESOs to its employees. Each view generates different measurement and timing
                          implications for accounting practice, yet each set of implications may be defended as
                          being consistent with the CF. Second, the measurement and timing difficulties in the
                          FASB's ESO proposal carried equal weight as the FASB's argument that their
                          proposal would result in more relevant financial reporting, because the CF does not
                          provide a mandate for standard-setters to emphasize relevance more heavily than
                          reliability. Hence, there were, and there will continue to be, cases where the FASB's
                          CF cannot generate support for standard-setting recommendations, even if the
                          recommendations are logically consistent with the implications of a given view (see
                          DePree, 1989).(2) The basic feature of these cases is that there are alternative views
                          for the event or transaction in question, the views have different measurement and
                          timing implications, and the CF does not proscribe any one of the views. A second
                          question that arises is what are the implications of the FASB not being able to provide
                          a compelling justification for its recommendations. This article develops the argument
                          that the FASB's inability to articulate convincing justifications for its proposals on
                          controversial topics, due to the weakness of its CF, may result in the locus of
                          standard-setting control shifting from the FASB to the SEC.

                          The primary conclusion of this article is that Conceptual Frameworks of Accounting
                          that are modelled after the FASB's approach are unlikely to prove efficacious. If
                          national (or international) accounting standard-setting bodies desire to develop or
                          improve upon their conceptual frameworks, they would be advised to consider an
                          alternative conceptual framework model. In this model, a conceptual framework
                          would provide specific definitions of economic events in terms of accounting elements
                          and an explicit statement whether the attributes of relevance or reliability are of
                          paramount concern. By doing so, the conceptual framework may provide
                          standard-setters with greater ability to develop logically consistent accounting
                          standards and to effectively respond to politically motivated criticism of their proposed
                          standards.

                          A number of previous studies analyse the usefulness of various CFs by examining
                          whether they allow the logical derivation of accounting rules. As examples of studies
                          analysing the FASB's CF, Dopuch and Sunder (1980) and Ketz and Kunitake (1988)
                          examine whether the U.S. CF is useful for guiding accounting for deferred taxes,
                          Dopuch and Sunder (1980) examine whether the U.S. CF provides useful guidance
                          for oil and gas exploration costs and current value reporting, and Joyce et al. (1982)
                          examine whether the CF's statements dealing with the required qualitative
                          characteristics of accounting data are operational and lead to clear normative
                          recommendations. As an example of a study analysing other conceptual frameworks,
                          Gore (1995) analyses the usefulness of the British and Irish Conceptual Frameworks
                          (Statements of Accounting Practice and Financial Reporting Standards) by examining
                          whether they provide useful guidance for group accounting, deferred taxation,
                          extraordinary items, inflation accounting, funds/cash flow statement, and research and
                          development.(3) Three features distinguish this article from previous research. First,
                          previous research mainly critiques the CF's definition of accounting elements by
                          illustrating the continuing role of judgment on a case-by-case basis in operationalizing
                          the CF's qualitative characteristics of accounting data. The implicit premise of these
                          works is that improved definitions of accounting elements would remedy the CF. By
                          contrast, the conclusion of this article is that CFs that consist only of definitions of
                          accounting elements are inherently flawed. Second, this article relates the weakness of
                          the CF exercise to a highly politicized context of standard-setting. Third, this article
                          analyses the interplay between the weakness of the CF and regulators' role in the
                          standard-setting process.

                          DESCRIPTION OF ESOS AND HISTORY OF ACCOUNTING FOR ESOS

                          ESOs are options to purchase the firm's stock, for a price usually equal to or close to
                          the stock price on the options' grant date. The options generally extend for a period of
                          between five and ten years, and the holder of the options has no voting rights and
                          receives no dividends prior to exercising the options. There is usually a vesting period
                          in which the employee cannot exercise the options, and if the employee leaves the firm
                          prior to the end of the vesting period, the options are forfeited. In addition, employees
                          cannot sell their ESOs, and if they leave the firm after the vesting date they must either
                          immediately exercise or forfeit their ESOs. Because ESOs are non-transferable,
                          employees wishing to reduce their risk from fluctuations in the firm's stock price may
                          exercise their ESOs before the expiration date even if they remain with the firm.(4)
                          Hence, the length of time that ESOs will remain outstanding is unknown until the
                          options are exercised.

                          In 1972, the AICPA issued Accounting Principles Board (APB) Opinion No. 25,
                          Accounting for Stock Issued to Employees. As a result, virtually no option grant that is
                          out-of-the-money on the grant date will ever be reflected on the income statement.(5)
                          An obvious shortcoming is that this approach ignores that an option to buy stock at a
                          predetermined price during an extended period of time has economic value. In 1982,
                          AcSEC, the AICPA's senior technical committee on accounting matters, submitted to
                          the FASB a lengthy position paper arguing that the cost of employee stock options
                          should be recognized as an expense. Notably, in the more specialized issues that the
                          FASB often delegates to AcSEC, AcSEC's pronouncements are considered
                          'authoritative'. Responding to the perceived shortcoming in accounting for ESOs, in
                          March 1984 the FASB added this to its agenda, and in May 1984 the FASB issued
                          an Invitation to Comment to the public. A preliminary conclusion, voiced by the FASB
                          in a Status Report dated 1 April 1987, was that the cost of an ESO should be
                          measured as the 'Minimum Value'. With the minimum value method, the exercise price
                          is first discounted from the option's expiration date back to the option's grant date
                          using the riskless rate. Then, this figure is subtracted from the stock's market price on
                          the grant date.(6) However, in the third quarter of 1988, the FASB reconsidered its
                          endorsement of the minimum value method, as respondents strongly opposed that
                          method. An expressed concern was that applying this method would lead to a
                          decrease in ESO grants (Poster, 1986). Another was that the non-transferability
                          feature of ESOs is not adequately accounted for under this method (Beresford and
                          Neary, 1986).

                          While the FASB continued its deliberations, in February 1992 the SEC asked its chief
                          accountant to study accounting for ESOs and report on the need for any change. In
                          October 1992, the SEC mandated new Proxy Statement disclosures (see Freher,
                          1992-3, for a description), and in June 1993 the FASB issued an Exposure Draft
                          (ED) on accounting for stock-based compensation that would have required income
                          statement recognition for the cost of ESOs. Given the length of time the FASB had
                          devoted to the stock-based compensation project, the project's very public nature
                          and the SEC's interest in the matter, it is difficult to imagine that the FASB did not
                          receive private expressions of support from the SEC prior to issuing the 1993 ED. In
                          fact, Freher (1992-3) states that the SEC's desire that ESOs should be expensed was
                          one of the primary reasons that the stock option project remained on the FASB's
                          agenda after its 1988 retraction of the minimum value recommendation.

                          Firms whose earnings would be most affected by the 1993 ED's proposal, namely
                          start-up and high-technology firms, mobilized opposition to the ED.(7) The Board
                          received 1,786 comment letters on the ED, with the vast majority opposed to any
                          form of recognition for ESOs (SFAS 123, par. 376). In opposition to the ED,
                          Senator Joseph Lieberman introduced in July 1993 the Equity Expansion Act of 1993,
                          which directs the SEC not to require ESO grants to be recognized as expenses,
                          regardless of FASB pronouncements. In late 1993, AcSEC issued a comment letter
                          to the FASB opposing the ED and recommending that the cost of ESOs be disclosed
                          in the footnotes but not recognized on the income statement. More ominously for the
                          FASB, the Accounting Standards Reform Act of 1994 was introduced in Congress.
                          This act proposed to amend federal securities law by mandating that the SEC
                          explicitly ratify all new accounting standards. The SEC monitored the process
                          surrounding the ED and privately communicated with FASB staff, but did not issue
                          any formal opinion of the ED (SEC Chairman Levitt, 1995). The lack of a publicly
                          stated opinion on the ED is consistent with SEC actions in other controversial cases.
                          Burton (1982, p. 46) explains how the SEC's statutory authority prevents it from
                          expressing official opinions in response to FASB proposals. However, at the
                          SEC-FASB open meeting of 17 November 1993, Commissioner J. Carter Reese
                          expressed strong concerns about the ED's direction and its potential effects on capital
                          formation. These comments received wide publicity in a Wall Street Journal column
                          (p. 18) dated 8 February 1994.(8)

                          Apparently in response to political opposition, the Board withdrew its 1993 ED in
                          December 1994. The SEC then took the very unusual step of filing a Letter of
                          Comment with the FASB in support of the FASB's decision to withdraw the ED. The
                          SEC letter of comment (letter no. 121), dated 3 April 1995, states:

                          The Commission has followed with great interest the FASB's project in accounting for
                          stock-based compensation. We congratulate the Board for arriving at a balanced
                          solution to the policy issues raised during the course of the project after a careful
                          deliberative process in which the views of all interested parties were considered. We
                          concur in the Board's decision that the time has come to end the debate on this project
                          . . . The Commission understands that the Board's tentative conclusions on this project
                          are as follows . . . In closing, the Commission believes that the FASB wisely has
                          recognized that the interests of both providers and users will be served most
                          effectively by concluding the stock-option debate. The Commission is fully supportive
                          of your continuing efforts to issue a final Statement in the relatively near future.

                          The letter of comment is noteworthy for three reasons. First, it appears designed to
                          support publicly the FASB's standard-setting role. Second, it absolves the SEC from
                          any role in the decade-long process in which accounting for stock options was
                          debated. Third, by spelling out its understanding of the final standard to be eventually
                          issued, the SEC provided an implicit endorsement and recommendation for the
                          FASB's final standard.

                          In December 1995, the FASB issued Statement of Financial Accounting Standards
                          No. 123, Accounting for Stock based Compensation (FASB, 1995). SFAS 123
                          requires firms to provide supplementary disclosure in which the compensation cost
                          associated with ESOs is reflected in pro-forma income statements and recommends,
                          but does not require, that firms actually recognize this compensation cost in the income
                          statement. This standard is completely consistent with the 'understanding' in the SEC
                          Letter of Comment to the FASB.

                          Computing the compensation cost associated with ESOs is complicated because
                          employees are likely to exercise their ESOs before the ESO expiration date due to
                          their difficulty in hedging their ESOs (see Lambert et al., 1991, Huddart, 1994,
                          Kulatilaka and Marcus, 1994, and Hemmer et al., 1994).(9) In fact, Huddart and
                          Lang (1996) and Hemmer et al. (1996) provide evidence that executives holding
                          options with a term of ten years will, on average, exercise the options after five or six
                          years. Using the assumption that employees do not hedge their ESOs, Mozes (1995),
                          Huddart (1994) and Kulatilaka and Marcus (1994) illustrate that the firm's cost of an
                          ESO with a contract life of t years may be significantly less than the benchmark
                          Black-Scholes value for an option with a term of t years, because employees'
                          expected holding period will be considerably shorter than the ESO term.(10)

                          SFAS 123 deals with the early exercise issue by requiring the compensation cost of
                          ESOs to be computed using employees' expected holding period rather than the
                          contract period of the ESOs. This assumed holding period is then input into standard
                          options pricing models such as the Black-Scholes or Binomial options pricing model
                          (OPCOST).(11) Among the allowable assumptions in SFAS 123 for employees'
                          expected holding period are that employees will exercise their ESOs: (a) when the
                          stock price reaches a specified multiple of the exercise price, (b) when the stock's
                          volatility reaches a specified level, and (c) when the stock price has increased by a
                          certain percentage within a given period of time. SFAS 123 also requires firms to
                          calculate the number of ESOs granted in which employees are expected to vest
                          (NVEST). The ESO cost is simply OPCOST multiplied by NVEST. Then, the grant
                          date value of the ESOs is prorated to compensation expense over the period between
                          the grant and vesting dates. Subsequent (to the grant date) changes in the stock price
                          and in estimates of employees' expected holding period do not affect the computation
                          of compensation cost. However, changes in NVEST will affect compensation cost. If
                          ESOs are recognized in financial statements, as recommended by SFAS 123, each
                          period between the grant and vesting dates will require the same debit to
                          compensation expense and credit to paid in capital - ESOs (unless there is a change in
                          NVEST).

                          As an example of the recommended accounting for ESOs in SFAS 123, assume that
                          the firm grants 1,100 ESOs, the value of one ESO on the grant date is $20, the
                          vesting period is four years, and employees are expected to vest in 1,000 ESOs. Each
                          year between the grant and vesting dates, the firm would debit $5,000 (1,000 x 20 x
                          1/4) of labour services and credit $5,000 of equity.(12)

                          POSSIBLE VIEWS OF THE ESO TRANSACTION

                          Accounting standard-setting requires a model that represents the economic event in
                          question in terms of accounting elements. This analysis refers to such a model as a
                          view. The bulk of the FASB's CF describes objectives of financial reporting and
                          qualitative qualities of accounting data; defines the elements of financial statements
                          such as assets, liabilities and equities; provides recognition criteria; and discusses
                          concepts of income. At the very end of the CF, pars 229-254 of Concepts No. 6
                          provide views of several transactions and events including instalment sales, discounted
                          debt issues, differences between accounting and taxable income, investment tax
                          credits, and certain costs incurred to procure future benefits. The objective of these
                          examples is 'to emphasize concepts and sound analysis and to foster careful
                          terminology, classification and disclosure (par. 229)'.(13) However, in par. 255, the
                          Board states that 'the examples in paragraphs 232-254 are intended to illustrate the
                          definitions and related concepts, not to establish standards for accounting practice
                          (par. 230)'. Furthermore, par. 229 suggests that the examples were deliberately
                          selected precisely because they were trivial in that 'they appear to make little
                          difference to practice'.

                          Since the CF does not provide a view for many transactions and events, standards
                          that are subsequently adopted provide a model for representing the transaction or
                          event in question in terms of the CF's accounting elements. The problem arises when
                          there are a number of plausible views for a particular event or transaction. Each view
                          may generate different measurement and timing implications, yet each set of
                          implications can be defended as being consistent with the CF. In these cases, the
                          FASB's support for the 'correct' view must be obtained and justified exogenously to
                          their CF, and regulators cannot defend a particular recommendation as being
                          mandated by the CF. An example of such a case is accounting for ESOs. This section
                          analyses several views of ESO grants, corresponding to those considered in the 1993
                          ED and in the discussion of alternatives in SFAS 123. The selection of an ESO view is
                          crucial because each of the three plausible views generates different measurement and
                          timing implications.

                          Liability View

                          The firm pays employees a receivable as payment for their labour services. The
                          receivable is the difference between the option's exercise price and the stock's market
                          price, and the receivable fluctuates in value based on ensuing stock price movements.
                          The basic issue in the liability view is whether ESOs can be recognized as liabilities
                          under the CF.

                          Statement of Financial Accounting Concepts No. 6, Elements of Financial Statements
                          (FASB, 1985) defines a liability to be 'probable future sacrifices of economic benefits
                          arising from present obligations of a particular entity to transfer assets or provide
                          services to other entities in the future as a result of past transactions or events (par.
                          35)'.(14) According to this definition of a liability, the firm's future opportunity cost,
                          which represents a legal obligation to dilute the value of a share of its common stock
                          and to transfer equity value from existing shareholders to ESO holders, cannot be
                          considered a liability. Stock that is not outstanding is not an asset of the firm, so the
                          firm does not transfer any assets when it sells stock to its employees at a
                          below-market price. Therefore, the firm's obligation to sell stock at a below-market
                          price does not represent a liability.(15)

                          Hence, if the liability view provides the accounting model for ESOs, no expense would
                          be recognized until the ESOs are exercised because there cannot be any
                          corresponding credit to a liability. The remaining question would be how to account
                          for the exercise transaction. Compensation expense would normally arise when the
                          firm sells stock at a below market price in exchange for labour services. According to
                          SFAC 6, an expense is an 'outflow or other using up of assets or incurrence of
                          liabilities . . . that have occurred or will eventuate as a result of central operations (pars
                          80-81)'. Selling stock for less than market value arguably represents an 'outflow or
                          using up of assets'. However, in the case of ESOs the labour services are provided in
                          the entire period between the grant and exercise dates. Recognizing compensation
                          expense during the exercise period would distort the intertemporal allocation of
                          income by creating a mismatch between the exercise period's revenues and expenses.
                          For example, with a ten year option that is not exercised until the expiration date,
                          recognizing an expense in year 10 would compress ten years' expense into one year.

                          As a result, the accounting model represented by the liability view would treat the
                          exercise transaction as a sale of stock for the exercise price, with a debit to cash and a
                          credit to capital stock for an amount equal to the exercise price. Alternatively, cash
                          would be debited for the exercise price, capital stock would be credited for the
                          market value of the stock, and a contra-equity account (i.e., discount on
                          paid-in-capital) or retained earnings would be debited for the difference between the
                          market and exercise prices. However, the contra-equity or retained earnings entry
                          would not affect income.

                          Advance View

                          The firm grants equity to its employees in the form of ESOs in exchange for
                          employees' future labour service, and the ESOs that are issued represent an advance
                          payment for services to be rendered. As collateral for employees' performance of their
                          employment obligations, the firm retains the right to cancel the ESOs if employees
                          leave the firm prior to the vesting date.

                          The amount the firm realizes from the ESO grant is the value of the labour services to
                          be provided by employees. According to the advance view, this value is the market
                          value of the ESOs on the grant date, because that is the amount employees receive for
                          their labour services when they contract with the firm. Hence, the amount the firm
                          receives from issuing the options is established on the grant date and subsequent
                          changes in the stock price do not affect that amount. As a result, all ESOs granted
                          would be valued on the grant date. This argument is similar to that in Robbins (1988).

                          However, employees are not legally obligated to perform any future services for the
                          firm when they receive ESOs. In the event that employees do not perform any future
                          services, they forfeit their ESOs but they are not required to compensate the firm with
                          any assets or services. While the options are likely to entice employees to remain with
                          the firm until the vesting date if the market price exceeds the exercise price, that
                          incentive does not qualify as an asset for the firm. Since the firm does not receive any
                          net assets on the grant date in exchange for the ESOs, the firm cannot possibly have
                          increased shareholders' equity on the grant date.(16) Hence, one could not rely on the
                          advance view to recognize a prepaid compensation asset and an increase in equity on
                          the grant date.(17) Rather, the appropriate accounting treatment would be to debit
                          compensation expense and credit equity each period between the grant and vesting
                          dates. Because the advance view maintains that the options are received by
                          employees on the grant date, all the ESOs would be valued on the grant date, and an
                          equal amount of compensation expense and equity issuance would be recognized each
                          period between the grant and vesting dates. This corresponds to the valuation method
                          required by SFAS 123.

                          Executory View

                          When granting ESOs, the firm legally commits to provide ESOs (equity) to its
                          employees at the vesting date if the employees perform specific labour services
                          between the grant and vesting dates.

                          The accounting model represented by this view is that ESO grants are executory
                          contracts. An executory contract specifies the goods or services to be provided by the
                          outside party and the compensation to be provided by the firm. As with other
                          executory contracts, the firm's obligation to issue ESOs can only be recognized when
                          the employees actually provide labour services. The reason is that U.S. GAAP does
                          not allow the firm to recognize the portion of executory contracts for which
                          performance has not yet occurred.(18)

                          According to this logic, no transaction occurs on the grant date. Because employees
                          are free to leave the firm and contract with another firm if the stock price drops prior
                          to the vesting date, employees can, in effect, reset the exercise price on their ESOs
                          prior to the vesting date and avoid the negative outcomes associated with a decline in
                          the value of the underlying stock. Furthermore, employees may attribute a higher
                          future value to their ESOs because they believe that their efforts will increase the firm's
                          future stock price. This seems plausible, given that ESOs are often exchanged for
                          incentive purposes. Hence, the grant date value of the ESOs does not represent the
                          value that employees expect to receive in the future for their labour services. Since
                          there is no evidence on the grant date of the value employees attach to their labour
                          services, the only logical date to value the ESOs using the executory view would be
                          the vesting date, when the ESOs are received by employees.

                          Under the executory view, the treatment of partial performance during the vesting
                          period would be analogous to the treatment of defined benefit pension plans with cliff
                          vesting. In cases with cliff vesting, the employee either vests in all the pension benefits
                          or in none of the pension benefits. Paragraph 42 of SFAS 87, Employers Accounting
                          for Pensions, requires that every year during the pension vesting period the firm must
                          recognize as pension expense an amount equal to the pro-rated pension benefits
                          earned during the year, despite the fact that the employee receives zero benefits if he
                          departs prior to the vesting date. Similarly, the fact that employees do not vest in any
                          ESOs prior to the vesting date should not preclude accounting recognition of
                          compensation expense prior to the vesting date. After partial performance, employees
                          have a receivable representing a pro-rated number of ESOs and the firm has an
                          obligation to issue a pro-rated number of ESOs.

                          Hence, the executory view would lead to the conclusion that the compensation
                          expense and equity issuance that are recognized each period in the interval between
                          the grant and vesting dates should be based on the expected vesting date value of the
                          ESOs and on the number of ESOs in which employees are expected to vest. As a
                          result, periodic compensation expense would differ every period between the grant
                          and vesting dates, as estimates of the vesting date value of the ESOs are revised.(19)
                          Consider the options which have a three year vesting period, where employees are
                          expected to vest in all options, and the options' value is $150 after the first year, $300
                          after the second year and $230 at the vesting date. The ESO expense recognized
                          would be $50 for the first year (1/3 x $150), $150 for the second year (2/3 x $300 -
                          $50), and $30 for the third year ($230 - $150 - $50).

                          Each of the three models is deficient. The error of modelling ESO grants as liabilities
                          of the firm (the liability view) is that an obligation based on equity values is more
                          commonly modelled as an equity instrument. The error of modelling ESO grants as
                          advances to employees (the advance view) is that: (a) the firm has the right to cancel
                          the ESOs if employees leave the firm prior to the vesting date, and (b) employees
                          have the right to leave the firm before the vesting date. These rights make it
                          problematic to model ESO grants as an exchange on the grant date between the firm
                          and its employees. The error of modelling ESO grants as executory contracts (the
                          executory view) is that most executory contracts obligate both parties to perform their
                          contractual obligations. However, ESO grants are unilateral contracts that are binding
                          on the firm but not on employees.

                          In summary, there are three basic timing and measurement issues in accounting for
                          ESOs. First, should the ESOs be measured on the grant date or on the vesting date?
                          The advance view implies grant date valuation, the executory view implies vesting date
                          valuation and the liability view implies exercise date valuation. Second, should the
                          ESO valuation be based on a suitably modified option pricing model or on the options'
                          intrinsic value (stock price minus grant price)? The advance and executory views imply
                          the use of options pricing models, but the liability view implies the use of intrinsic value.
                          Third, if options pricing models are to be used for ESOs, how should the ESO
                          valuation be affected by the early exercise problem? While each of the three ESO
                          views has different measurement and timing implications for financial accounting, none
                          of the views is compelling or mandated by the FASB's CF. The analysis of the three
                          ESO views can be generalized to suggest that the Board should have crafted the CF
                          to provide views for as many relevant transactions as possible. While the Board's
                          decision not to provide views in its CF reduced the controversy surrounding the CF, it
                          also limited the CF's usefulness in providing solutions to politicized standard-setting
                          problems.

                          THE ROLES OF RELEVANCE AND RELIABILITY IN ACCOUNTING FOR
                          ESOS

                          Despite the lack of an ESO view in its CF, the FASB could possibly have obtained
                          regulators' backing for the 1993 ED by arguing that its proposal provides more
                          relevant and comparable data to users of financial statements than provided under
                          APB 25. This section develops the argument that, even if one accepts the assumption
                          that the data required to be reported by the 1993 ED is more relevant than the data
                          required to be reported by other ESO reporting rules, a second flaw in the FASB's
                          CF prevented it from using the relevance argument to support the 1993 ED.
                          Specifically, the FASB's CF provides little guidance for dealing with conflicts between
                          the qualitative data characteristics of relevance, comparability, and reliability.

                          According to Statement of Financial Accounting Concepts No. 2, Qualitative
                          Characteristics of Accounting Information (FASB, 1980), reliability is based on
                          verifiability and representational faithfulness. One requirement of verifiability is that:
                          'Verifiable financial accounting information provides results that would be substantially
                          duplicated by independent measurers using the same measurement methods (par. 82)'.
                          Similarly: 'Verifiability implies consensus. Verifiability can be measured by looking at
                          the dispersion of a number of independent measurements of some particular
                          phenomenon. The more closely the measurements are likely to be clustered together,
                          the greater the verifiability of the number used as a measure of the phenomenon (par.
                          84).' Given the latitude firms have in dealing with the early exercise problem (i.e., the
                          measurement problems) and the question of when ESO costs should be determined
                          (i.e., the timing problems), there is a question concerning the reliability of ESO costs.

                          The FASB attempted to minimize the issue of reliability with respect to the timing and
                          measurement problems by arguing that its proposal (a) results in the most relevant
                          information set being reported, and (b) improves users' ability to make cross-sectional
                          comparisons with accounting data. SFAS 123 states that: 'The Board continues to
                          believe [sic] that financial statements would be more relevant if the estimated fair value
                          of employee stock options was included in determining an entity's net income, just as
                          all other forms of compensation are included (par. 61)'. In addition, SFAS 123 states
                          that ESO recognition would improve the comparability of financial statements because

                          Some entities use fixed stock options more extensively than other enterprises do, and
                          reported operating expenses thus are understated to different degrees. Comparisons
                          between entities of profit margins, rates of return, income from operation, etc. are
                          impaired to the extent that entities continue to account for their stock-based employee
                          compensation according to the provisions of APB 25 (pars 97-98).(20)

                          Further, SFAS 123 explicitly states that the FASB emphasizes comparability more
                          than reliability in the debate about accounting for ESOs. 'Increasing the comparability
                          of financial statements is a worthy goal, even if all entities use a method that is less
                          precise than the Board or its constituents might prefer (par. 39).'

                          Another approach taken by SFAS 123 to the issue of reliability is the argument that
                          no reliability problem exists because any value for ESOs is more reliable than a value
                          of zero:

                          All estimates, because they are estimates are imprecise. Few accrual-based
                          accounting measurements can claim absolute reliability, but most parties agree that
                          financial statement recognition of estimated amounts that are approximately right are
                          preferable to the alternative - recognizing nothing - which is what Opinion 25
                          accounting recognizes for most employee stock options. Zero is not within the range
                          of reasonable estimates of the value of employee stock options . . . (par. 111).

                          SFAS 123's primary line of argument, emphasizing comparability and relevance
                          benefits, is not compelling because it attempts to justify minimizing a primary attribute
                          (reliability) for the benefit of another primary attribute (relevance) and a secondary
                          attribute (comparability). Specifically, SFAC 2 presents a hierarchy of accounting
                          qualities. Reliability and relevance are both a 'primary decision-specific quality', while
                          comparability is a 'secondary and interactive quality'. The argument that no reliability
                          issue exists is easily countered simply because the fact that a number is more plausible
                          than zero does not make that number inherently reliable. Though estimates are
                          routinely used in pensions, allowance for uncollectibles, depreciation, deferred taxes,
                          etc.,(21) those estimates will eventually be 'trued up' and poor estimates will only
                          distort the intertemporal allocation of income. With ESO estimates, the equity value
                          that is initially provided will never be 'trued up' to reflect employees' actual holding
                          periods. As a result, poor estimates of employees' actual holding periods will
                          permanently distort financial statements.

                          Due to the inconclusiveness of its arguments, the FASB did not successfully
                          demonstrate that the objectives of financial accounting, as agreed upon in its CF, are
                          best served by recognizing ESOs. Had its CF explicitly stated that relevance is the
                          more important attribute and should be given precedence in a conflict between
                          relevance and reliability, the FASB might have obtained political support from
                          regulators for their focus on the benefits of the increased relevance in financial
                          accounting from ESO recognition. Instead, SFAC 2 defends its approach to
                          preferability questions by stating that 'The basis for those decisions should be better
                          understood if they can be seen to be aimed at obtaining an optimal mix (as judged by
                          the Board) of certainly defined informational characteristics (par. 157)'. However,
                          statements that a certain trade-off is or is not optimal can never be shown to be true or
                          false because they are opinions rather than facts. Moreover, the CF's notion that there
                          is an appropriate trade-off between relevance and reliability is logically inconsistent.
                          As Chambers (1996) argues, trading off relevance and reliability results in financial
                          statements that are nether relevant nor reliable.

                          Hence, it is not surprising that the FASB's ESO proposal was not accepted or
                          regarded as compelling by affected parties, and that the FASB failed to obtain the
                          SEC's political support for recognizing ESOs. Basically, the FASB's CF calls for the
                          Board's judgement in cases where plausible alternatives exist (i.e., 'as judged by the
                          Board'). However, standard setters' personal preferences are not compelling
                          arguments for regulators' support. Quite simply, the FASB could not demonstrate that
                          their ESO proposal was mandated or logically required by their CF.

                          AcSEC and SEC Comments on the 1993 ED

                          An analysis of AcSEC's comment letter to the 1993 ED on Stock based
                          Compensation indicates that AcSEC framed its criticism of the ED in terms of the
                          reliability of ESO costs computed using option-pricing models. The AcSEC letter
                          states that:

                          This majority concludes that the usefulness of financial statements for investment and
                          credit decisions will not be improved by including the expense amounts determined
                          under these models. It believes current or improved disclosure requirements will
                          inform financial statement users adequately about stock options granted to employees
                          . . . AcSEC continues to believe methods other than a minimum value method will not
                          produce reliable results that can be compared among similar enterprises . . . While a
                          majority of AcSEC reached the conclusion that the fair value issue remains intractable,
                          AcSEC continues to support fair value in concept. If such values become objectively
                          more determinable through changes in the structure of options, AcSEC will reconsider
                          its views. (emphasis added)

                          This reinforces the argument that the theoretical hurdles which the 1993 ED could not
                          surmount were timing and measurement issues, and the relevance-reliability trade-off.

                          An analysis of SEC Commissioners' remarks provides a similar inference. In
                          explaining the difficulties that led to the demise of the 1993 ED, SEC Commissioner
                          Richard Roberts (1994) remarked that:

                          The first and largest problem with this approach is how to value stock options. Should
                          they be valued at grant date, vesting date, or exercise date? . . . There has been no
                          consensus on the timing or the formula. While the Black-Scholes and binomial pricing
                          models are the most popular for valuing exchange-traded options, they are less
                          effective in valuing options that are non-transferable.

                          These remarks suggest that the SEC had timing and measurement concerns with the
                          1993 ED. First, because there is no established view of the ESO grant transaction, it
                          is not clear when the ESOs should be measured. Second, even if one could determine
                          when the ESOs should be measured, the measurement may not be sufficiently reliable.
                          The following remarks by SEC Commissioner Steven M. H. Wallman (1995) also
                          suggest that the SEC was not persuaded by the FASB's arguments: 'I understand the
                          arguments made both for and against a charge to income for stock option
                          compensation. It is wrong to conclude that there can be only one correct answer.' On
                          the other hand, the FASB had argued that there is only one answer - theirs.

                          Commissioner Roberts further suggests that the measurement and timing problems
                          were exacerbated by the political nature of accounting for stock options, and that the
                          SEC may apply different reliability standards in different cases. He continues, noting
                          that:

                          A new accounting standard for employee stock options imposing a charge to earnings
                          could potentially have a substantial adverse impact on small, fledgling companies,
                          Therefore, any standard imposing such a charge should be very carefully drawn, or it
                          could have a negative impact on our capital formation system to the detriment of both
                          companies and investors. The problem is that it appears no one can agree on a
                          formula, and I am not sure that a good one exists for this purpose at the present.

                          By interspersing the theoretical measurement concerns (i.e., 'no one can agree on a
                          formula') with concerns for small company capital formation, Roberts' remarks suggest
                          that in the absence of the political and resource allocation issues surrounding
                          accounting for stock options, the timing and measurement problems would have been
                          acceptable. A similar sentiment can be found in Commissioner Wallman's remarks:
                          'Given the highly charged context of the stock option debate, the FASB approach of
                          requiring disclosure - as opposed to requiring recognition-of the fair value of stock
                          options is an acceptable solution' (emphasis added).

                          The conclusion from the Commissioners' remarks is that, given the political and
                          resource allocation issues, the FASB had a greater burden of demonstrating that its
                          formula for ESO recognition satisfactorily resolved the timing and measurement
                          problems. In the absence of a strong theoretical justification for the FASB's 1993 ED,
                          the SEC was not willing to risk the possibly negative effects on capital formation and
                          to expend the political capital needed to support the FASB's ED. However, in other
                          cases, where the political and economic stakes are smaller, the timing and
                          measurement issues raised by the 1993 ED would not have been problematic.(22)

                          SEC PERSPECTIVE ON ITS RELATIONSHIP WITH THE FASB

                          The lingering questions raised by the process in which SFAS 123 was adopted are
                          which group is responsible for considering the political and economic costs of
                          accounting standards, and how the SEC's approach to standard-setting differs from
                          that of the FASB. Dopuch and Sunder (1980, p. 18) struggle to find a role for private
                          standard-setters. They suggest:

                          First, the Board (FASB) could explicitly recognize the nature of financial accounting as
                          a social activity which affects a varied set of interests, both of those who actively
                          participate and those who do not . . . In its statement of objectives, the Board could
                          define mechanisms for arriving at a compromise ruling after a hearing has been given to
                          all affected groups in society. The Board's primary objective would simply be to arrive
                          at a compromise ruling after considering various points of view on each issue.

                          However, Dopuch and Sunder (p. 19) conclude by asking: 'Once this role were
                          recognized, what would be the advantages and disadvantages of allowing a private
                          board like the FASB to make compromise decisions?'

                          A conclusion from the SEC actions and comments on the SFAS 123 project implies
                          that the role Dopuch and Sunder suggest for the FASB has been appropriated by the
                          SEC. In commenting on how the SEC regards its relation with the FASB, SEC
                          Chairman Levitt (1995) states:

                          Both the Commission and the FASB consider accounting proposals in the context of
                          ensuring investor protection and providing unbiased information to the securities
                          markets. The Commission does have a broader mission, which includes, among other
                          things, enforcement of the federal securities laws, market regulation, and facilitation of
                          capital formation. (emphasis added)

                          While the facilitation of capital formation is accomplished by reflecting economic
                          activity faithfully, this relation is viewed only as a general, but not a binding guide.
                          Chairman Levitt further states: 'The increased transparency provided by unbiased
                          financial reporting generally has been considered to facilitate capital formation, not
                          diminish it' (emphasis added). In other words, the SEC expects the FASB to provide
                          an authoritative practitioner perspective of how accounting data should be presented
                          to report on economic activity as faithfully as possible, and the SEC balances that
                          input with concerns such as the facilitation of capital formation which they, but not the
                          FASB, need to consider. This conception of the FASB's role is consistent with the
                          FASB's stated mission, which is

                          to establish and improve standards of financial accounting for the guidance and
                          education of the public, including issuers, auditors, and users of financial information . .
                          . To be objective in its decision making to ensure, insofar as possible the neutrality of
                          information resulting from its standards. To be neutral, information must report
                          economic activity as faithfully as possible without colouring the image it communicates
                          for the purpose of influencing behaviour in any particular direction.

                          Hence, if expensing ESO grants would interfere with capital market activity for small
                          firms, the SEC, but not the FASB, must consider that effect.

                          This is reflected in comments by SEC Commissioner Wallman and SEC Chief
                          Accountants Walter Schuetze and Michael Sutton. Wallman (1995) comments on the
                          stock option debate that:

                          I believe FASB's fundamental independence, integrity, and wisdom are essential to its
                          ability to develop high-quality accounting standards. I also prefer to have the private
                          sector set accounting standards subject to Commission oversight, rather than having
                          that responsibility revert to the Commission. Unless this process remains apolitical,
                          however, that may become difficult. (emphasis added)

                          Schuetze (1994), commenting on AcSEC's change of position on ESOs, states:

                          If public companies are pressuring their outside auditors and AcSEC to take particular
                          positions on financial accounting and reporting issues and outside auditors are
                          subordinating their views to those of their clients, can the outside auditor community
                          continue to claim to be independent? . . . Could such a trend be anything other than an
                          invitation to Congress, the SEC and other regulators to regulate more heavily, and
                          more directly? (emphasis added)

                          Sutton (1996) emphasized the need for professionalism in the U.S. regulatory system:

                          Implicit in all this is the assumption that private sector standards setters and auditors
                          will act independently... Historically, we have relied heavily on the professionalism of
                          registrants, auditors, and standard-setters to make the system work... the
                          Commission's ability to depend on the professionalism of private sector participants is
                          critical to maintaining the credibility of our system. (emphasis added)

                          Given the obvious political nature of standard setting in general, and of ESOs in
                          particular, the implication of these remarks is that the SEC, and not the FASB, is the
                          appropriate forum for the political and resource allocation issues to be debated.
                          Moreover, from the perspective of the SEC, the only role for the FASB is to provide
                          an authoritative position exclusively representing that of the profession. SEC Chairman
                          Levitt (1996) specifically notes that the FASB's value to the SEC is derived from the
                          fact that it does not consider political ramifications: 'While I freely believe in the need
                          for free and unfettered debate, there's a process that is insulated from politics... This is
                          one reason I'm not receptive to requests that the SEC intervene to soften or delay
                          FASB projects before the process has reached out to all constituents.'

                          Because it argues for an increase in the political and regulatory importance of the SEC
                          relative to that of the FASB, the SEC's role in the debate on accounting for ESOs
                          may be interpreted as an attempt to increase its regulatory domain. This interpretation
                          of the SEC's behaviour is consistent with Bealing et al. (1996). They interpret the
                          SEC's early regulatory actions in the 1930s and 1940s as being motivated by the
                          desire to institutionalize its role and establish itself as part of the political and legal
                          landscape.

                          SUMMARY AND CONCLUSIONS

                          This article argues that the effectiveness of the FASB's Conceptual Framework (CF)
                          is undermined by the level of abstraction with which it was written, and to a lesser
                          extent by the lack of guidance on how strongly to emphasize the attribute of relevance.
                          The FASB's CF primarily provides accounting concepts or definitions of accounting
                          elements, while its standards provide views, or models that relate an economic event
                          to the related accounting elements definitions in its CF. As a result, questions
                          continuously arise about the appropriate accounting model for various economic
                          events, and the FASB's CF may not deliver regulators' political support for its
                          advocacy of a particular reporting rule.

                          Even without providing definitive views, a CF can provide greater political support to
                          standard-setters by explicitly stating that the attribute of relevance should be more
                          heavily emphasized than the attribute of reliability. Such an approach would provide
                          standard-setters with the ability to enlist regulators' support in reporting issues
                          involving uncertainties such as pensions, deferred taxes, estimated current asset values
                          and ESOs. Specifically, regulators could counter arguments that a proposed rule
                          would create unreliable accounting data by pointing to the CF's mandate to pursue
                          relevance even at the cost of reliability. However, the FASB's CF does not provide
                          any support for emphasizing relevance over reliability.

                          In cases such as employee stock options, accounting for research and development
                          cost (Bierman and Dukes 1975), and capitalization of interest cost (Mozes and Schiff
                          1995), the FASB appears to have claimed a mandate from its CF when no such
                          mandate exists. When the political opposition to a standard proposed by the FASB is
                          not intense and there are no serious issues concerning resource allocation and capital
                          formation, as in the latter two cases, the proposal may be acceptable to the SEC
                          despite its weak justification. However, the greater the political activity surrounding a
                          proposed accounting standard, the better the FASB must justify its recommendation
                          to the SEC. In the highly politicized case of ESOs, the FASB did not provide a
                          compelling rationale for its proposals, and its proposals were ultimately rejected.

                          The FASB's experience with accounting for ESOs leads to the conclusion that it
                          would have been better served by a CF that leads to clearer standard-setting
                          recommendations.(23) To the extent that the CF cannot provide compelling arguments
                          in favour of the FASB's proposals in politicized issues, the political and economic
                          issues will dominate the standard-setting process and the FASB will risk ceding
                          control of these issues to the SEC. If U.S. accounting standards are going to continue
                          to be set in the private sector, it may be necessary for the FASB to revise its CF to
                          provide explicitly accounting models for a number of troublesome events and
                          transactions, and to provide a method for resolving tensions between the relevance
                          and reliability attributes.

                          The more general conclusion of this paper is that, ideally, conceptual frameworks
                          should contain specific definitions of economic events in terms of accounting elements
                          and an explicit statement whether the attributes of relevance or reliability are of
                          paramount concern. Clearly, the economic meaning of various transactions and the
                          relative importance of the relevance and reliability attributes to financial market
                          participants vary across countries. Hence, one can imagine conceptual frameworks
                          based on the suggested model varying substantially from another, and one can
                          understand why standard-setters in different countries would attempt to 'reinvent the
                          wheel' by developing their own Conceptual Framework of Accounting. By contrast,
                          countries following the U.S. approach should have similar conceptual frameworks
                          because there is little substantive variation in the general description of accounting
                          elements and required data attributes. Hence, if standard-setters in other countries
                          wish to emulate the FASB's approach, they might avoid the cost of developing a
                          conceptual framework by simply adopting the FASB's CF.

                          1 Accounting researchers have long noted the political factors that shape financial
                          reporting standards (e.g., Gerboth 1973, Solomons 1978, and Watts and Zimmerman
                          1978).

                          2 Past examples other than ESOs include: (a) accounting for pensions, where the
                          value of the firm's unfunded or overfunded pension obligation does not appear on the
                          balance sheet as assets or liabilities; (b) accounting for marketable securities, where
                          the change in the value of 'available-for-sale' marketable securities does not appear on
                          the income statement; and (c) accounting for foreign subsidiaries, where the
                          exchange-rate related change in the translated asset values of the subsidiary does not
                          appear in the parent's income statement. A recent example is the FASB's proposal for
                          accounting for derivatives, where the change in the value of derivatives that hedge
                          future cash-flows would not appear on the income statement.

                          3 Dopuch and Sunder (1980), Joyce et al. (1982), Ketz and Kunitake (1988) and
                          Chambers (1996) all criticize the FASB's CF and conclude that it is inadequate. Gore
                          (1992, 1995) concludes that the British and Irish Conceptual Framework is modestly
                          useful because it provides guidance on some, but not all, of the issues examined.

                          4 By contrast, the assumption for marketable options on stocks that do not pay
                          dividends is that they will never be exercised prior to the expiration date (Merton
                          1990).

                          5 APB 25 requires the firm to calculate its compensation expense on ESOs as the
                          option's intrinsic value (i.e., the stock price minus the exercise price) on the
                          measurement date. Since the measurement date is almost always the date on which the
                          ESOs are granted and the intrinsic value for an at-the-money grant is zero on that
                          date, no expense is ever associated with the option grant.

                          6 If the firm pays dividends, the present value of the dividends expected to be paid
                          from the grant date through the expiration date is subtracted from the minimum value.
                          Noreen (1979) suggests that the minimum value is the firm's net cost to acquire
                          treasury stock and hedge the ESOs. The stock's market price on the grant date
                          represents the cost to acquire the stock and the discounted exercise price represents
                          the firm's proceeds from the options' exercise. Whether or not the employee
                          eventually exercises the ESOs, the firm will have only lost the time value of money
                          represented by the minimum value.

                          7 Foster et al. (1991, 1993) provide evidence on how various methods of calculating
                          the firm's ESO cost would reduce earnings.

                          8 The SEC governing board consists of five commissioners and a chairman.

                          9 There are two reasons why employees are unlikely to hedge their ESOs. First, the
                          ESOs may represent a large proportion of employees' wealth, and they may not have
                          sufficient additional wealth to hedge the options. Second, execution of a hedging
                          strategy requires employees to sell the underlying stock short, and such action may
                          conflict with insider trading restrictions. These issues do not arise with transferable
                          options, as long as there exists at least one investor with the ability to hedge the
                          options and earn risk-free arbitrage profits.

                          10 Huddart (1994) and Kulatilaka and Marcus (1994) show that the greater the
                          employee's risk aversion the lower the firm's ESO cost, and the greater the
                          employee's wealth the greater the firm's ESO cost, because more risk-averse
                          (wealthy) employees will hold their ESOs for a shorter (longer) period of time. Mozes
                          (1995) provides an upper bound for the firm's cost of ESOs that is independent of
                          employees' risk aversion. He shows that for firms with high unsystematic risk relative
                          to systematic risk, such as small high-tech firms, the minimum overstatement in using
                          the Black-Scholes benchmark to determine the firm's ESO cost often exceeds 50 per
                          cent.

                          11 Empirical support for applying Black-Scholes options valuation techniques to
                          long-lived options is provided by Noreen and Wolfson (1981). They find that market
                          prices of publicly traded warrants are approximated by the Black-Scholes model.

                          12 In most cases, the labour cost would be immediately expensed. However, the cost
                          would be capitalized if the firm could identify that the labour was for the construction
                          of long-term assets.

                          The cost of nonqualified stock options (NQOs) is deductible by the firm when the
                          employee exercises the ESOs. However, the cost of incentive stock options (ISOs) is
                          not deductible by the firm unless the options are disqualified by the employee. For
                          NQOs, a deferred tax benefit is recognized each period, since the compensation
                          expense that is recognized for financial accounting purposes is not yet deductible for
                          tax purposes. Hence, if the corporate tax rate is 30 per cent, each year the firm would
                          recognize a deferred tax benefit of $1,500 ($5,000 x .3) and a corresponding
                          decrease in tax expense of $1,500. Any difference between the tax benefits actually
                          realized from employees' exercise of the ESOs and the cumulative amount recognized
                          on the balance sheet as a deferred tax benefit would be a direct adjustment to
                          shareholders' equity.

                          13 Consistent with the usage in this article, the term 'view' is used several times in par.
                          240 to denote an accounting model that represents the economic event in terms of
                          accounting elements.

                          14 The definition of 'other entities' is given in SFAC 6 (par. 24) as: 'employees,
                          suppliers, customers or beneficiaries, lenders, stockholders, donors, and governments
                          are all other entities to a particular entity' (emphasis added).

                          15 Stock appreciation rights (SARs) are similar to ESOs except that, in most cases,
                          employees can demand an amount of cash equal to the difference between the
                          option's exercise price and the stock's market price on the date the options are
                          exercised. SFAS 123 requires that SARs be recognized as compensation expense
                          and liabilities if payment in cash is required at the employee's discretion. Balsam
                          (1994) criticizes the distinction between ESOs and SARs and argues that ESOs
                          should also be treated as liabilities. However, unlike ESOs, SARs are liabilities under
                          the CF because the firm's obligation is to transfer cash to ESO holders. While the firm
                          can, in theory, issue stock to obtain the cash to pay the SAR obligation, the firm can
                          also do so with other monetary liabilities such as bonds or notes payable.

                          16 A analogous situation exists when investors subscribe to acquire stock, but have
                          not yet paid for the stock. Because investors have no enforceable obligation to pay the
                          balance of the subscription amount, the SEC does not allow firms to report the
                          subscriptions receivable as an asset. Rather, the firm must show them as a
                          contra-equity account. The net effect is that when investors subscribe to acquire the
                          firm's stock, the firm does not report any increase in either assets or equity until the
                          investors actually pay for the stock.

                          17 Interestingly, the 1993 ED recommended that equity and prepaid compensation
                          cost (equal to the value of the ESOs on the grant date) should be recognized on the
                          grant date, and that the prepaid compensation asset should be amortized to
                          compensation expense over time.

                          18 Operating leases and take-or-pay contracts are examples of executory contracts
                          that are not recognized in financial accounting.

                          19 The FASB's terminology and reasoning in its final standard is somewhat confusing.
                          Although SFAS 123 requires the accounting method implied by the advance view,
                          par. 95 of SFAS 123 actually refers to ESO grants as executory contracts.

                          20 Under the reporting that follows from the executory view, the compensation
                          expense in the interval between the grant and vesting dates will change from period to
                          period, as the current value of the ESOs changes. The resulting volatility in
                          compensation expense may harm the time-series and cross-sectional comparability of
                          income and compensation expense. As a result, the comparability argument in favour
                          of ESO recognition applies more readily to the implications of the advance view than
                          to the implications of the executory view. It is arguable that this line of reasoning led
                          the FASB to adopt the advance view.

                          21 For example, in pension accounting the firm must estimate employees' tenure at the
                          firm, employees' final pay, employees' longevity, the appropriate discount (settlement)
                          rate, and the long-run rate of return on plan assets. In addition, it is not clear if the
                          liability should be based on the accumulated benefit obligation or on the projected
                          benefit obligation. Deferred taxes involve reliability issues because estimates must be
                          made for when temporary differences will reverse, and whether deferred tax benefits
                          will be realized.

                          22 Gorton (1971) provides a somewhat similar conclusion from his in-depth analysis
                          of the SEC's actions in the standard-setting process surrounding SFAS 19,
                          Accounting and Reporting by Oil and Gas Producing Companies. In that case, the
                          SEC took the unprecedented steps of holding public hearings on whether to accept
                          the FASB's standard and then issuing its own standard. based on interviews with the
                          principal SEC figures involved, Gorton concludes that the SEC did not find the
                          justification offered by the FASB for its standard to be compelling or even convincing.
                          Further, Gorton argues that in the absence of political pressure, the SEC may have
                          accepted the FASB's standard because of the lack of a mandate or an excuse to
                          create a better standard. However, Gorton (p. 41) states that the intense political
                          pressure on oil and gas accounting created a 'destabilizing force which provided the
                          opportunity for the Commission's leadership to extend its current-value initiatives by
                          proposing reserve recognition accounting'.

                          23 Gore (1992) argues that another major error in developing the CF was the
                          extravagant expectations the FASB initially associated with the CF project.

                          REFERENCES

                          American Institute of Certified Public Accountants, Accounting Principles Board
                          Opinion Number 25, Accounting for Stock Issued to Employees, AICPA, 1972.

                          Balsam, S., 'Extending the Method of Accounting for Stock Appreciation Rights to
                          Employee Stock Options', Accounting Horizons, December 1994.

                          Bealing, W. E., M. W. Dirsmith and T. Fogarty, 'Early Regulatory Actions by the
                          SEC: An Institutional Theory Perspective on the Dramaturgy of Political Exchanges',
                          Accounting, Organizations and Society, Vol. 21, No. 4, 1996.

                          Beresford, D. R., and R. D. Neary, 'FASB Examines Compensation for Stock
                          Options Plans', Financial Executive, April 1986.

                          Bierman, H. Jr, and R. E. Dukes, 'Accounting for Research and Development Costs',
                          Journal of Accountancy, April 1975.

                          Burton, J. C., 'The SEC and Financial Reporting: The Sand in the Oyster', Journal of
                          Accountancy, June 1982.

                          Chambers, R. J., 'Ends, Ways, Means and Conceptual Frameworks', Abacus,
                          September 1996.

                          Collett, P., 'Standard Setting and Economic Consequences: An Ethical Issue',
                          Abacus, March 1995.

                          DePree, C. D., 'Testing and Evaluating a Conceptual Framework of Accounting',
                          Abacus, March 1989.

                          Dopuch, N, and S. Sunder, 'FASB's Statements on Objectives and Elements of
                          Financial Accounting: A Review', The Accounting Review, January 1980.

                          Financial Accounting Standards Board, Statement of Financial Accounting Concepts
                          No. 2, Qualitative Characteristics of Accounting Information, FASB, 1980.

                          -----, Statement of Financial Accounting Concepts No. 6, Elements of Financial
                          Statements, FASB, 1985.

                          -----, SFAS 123, Accounting for Stock based Compensation, FASB, 1995.

                          Foster, T. W., P. Koogler and D. Vickrey, 'Valuation of Executive Stock Options
                          and the FASB Proposal', The Accounting Review, 1991.

                          -----, 'Valuation of Executive Stock Options and the FASB Proposal: An Extension',
                          The Accounting Review, January 1993.

                          Freher, E., 'What the SEC's Proxy Disclosure Rules Mean to Your Executive
                          Compensation Practices', Journal of Corporate Accounting and Finance, Winter
                          1992-3.

                          GeRboth, D, 'Research, Intuition, and Politics in Accounting Inquiry', The Accounting
                          Review, July 1973.

                          Gore, P., The FASB Conceptual Framework Project 1973-1985: An Analysis,
                          Manchester University Press, 1992.

                          -----, 'The Practical Use of Conceptual Frameworks For Financial Accounting',
                          Advances in International Accounting, Vol. 8, 1995.

                          Gorton, D. E., 'The SEC Decision not to Support SFAS 19: A Case Study of the
                          Effect of Lobbying on Standard Setting', Accounting Horizons, March 1991.

                          Hemmer, T., S. Matsunaga and T. Shevlin, 'Estimating the "Fair Value" of Employee
                          Stock Options with Expected Early Exercise', Accounting Horizons, December 1994.

                          -----, 'The Influence of Risk Diversification on the Early Exercise of Employee Stock
                          Options by Executive Officers', Journal of Accounting and Economics, Vol. 21, 1996.

                          Horngren, C. T., 'Uses and Limitations of Conceptual Frameworks', Journal of
                          Accountancy, April 1981.

                          Huddart, S., 'Employee Stock Options', Journal of Accounting and Economics, Vol.
                          18, 1994.

                          Huddart, S., and M. Lang, 'Employee Stock Option Exercises: An Empirical
                          Analysis', Journal of Accounting and Economics, Vol. 21, 1996.

                          Joyce, E. J., R. Libby and S. Sunder, 'Using FASB's Qualitative Characteristics in
                          Accounting Policy Choices', Journal of Accounting Research, Autumn 1982.

                          Ketz, J. E., and W. K. Kunitake, 'An Evaluation of the Conceptual Framework: Can
                          it Resolve The Issues Related to Accounting for Income Taxes?', Advances in
                          Accounting, Vol. 6, 1988.

                          Kulatilaka, N., and A. J. Marcus, 'Valuing Employee Stock Options', Financial
                          Analysts Journal, November-December 1994.

                          Lambert, R. A., D. F. Larcker and R. A. Verrecchia, 'Portfolio Considerations in
                          Valuing Executive Compensation', Journal of Accounting Research, Spring 1991.

                          Levitt, A, 'An Exchange Between the Honorable Edward J. Markey, Chairman U.S.
                          House of Representatives Subcommittee on Telecommunications and Finance,
                          Committee of Energy and Commerce and the Honorable Arthur Levitt, Jr., Chairman,
                          Securities and Exchange Commission', Accounting Horizons, March 1995.

                          -----, 'The Accountant's Critical Eye', Remarks at the 24th Annual National
                          Conference on Current SEC developments, AICPA, Washington, D.C., 10
                          December 1996.

                          Merton, R. C., Continuous-Time Finance, Basil Blackwell, 1990.

                          Milburn, J. A., 'Building a Better Conceptual Framework', CA Magazine, December
                          1991.

                          Mozes, H. A., 'A Framework for Normative Accounting Research', Journal of
                          Accounting Literature, Vol. 11, 1992.

                          -----, 'An Upper Bound for the Firm's Cost of Employee Stock Options', Financial
                          Management, Winter 1995.

                          Mozes, H. A., and A. I. Schiff, 'A Critical Look at SFAS 34: Capitalization of
                          Interest Cost', Abacus, March 1995.

                          Noreen, E., 'Comment: Measuring the Compensation Element in Employee Stock
                          Option Plans', Journal of Accounting, Auditing, and Finance, Fall 1979.

                          Noreen, E., and M. Wolfson, 'Equilibrium Warrant Pricing Models and Accounting
                          for Executive Stock Options', Journal of Accounting Research, Autumn 1981.

                          Poster, C. Z., 'A Look at the Future of Stock Options', Financial Executive, April
                          1986.

                          Robbins, B., 'FASB's Long Look at Stock Compensation Plans', Journal of
                          Accountancy, August 1988.

                          Roberts, R., 'Remarks on Current Disclosure Issues', Institute of Management
                          Accountants Annual SEC/FASB Accounting and Reporting Conference, New York,
                          14 December 1994.

                          Securities and Exchange Commission, Accounting Series Release No. 150, Statement
                          of Policy on the Establishment and Improvement of Accounting Principles and
                          Standards, SEC, 20 December 1973.

                          Schuetze, W., 'Remarks to the 1994 AICPA National Conference on Current SEC
                          Developments'. Cited in Journal of Accountancy, March 1994.

                          Solomons, D., 'The Politicization of Accounting', Journal of Accountancy, Vol. 146,
                          No. 5, 1978.

                          -----, 'The FASB's Conceptual Framework: An Evaluation', Journal of Accountancy,
                          June 1986.

                          Sutton, M. H., 'Remarks at the 24th Annual National Conference on Current SEC
                          Developments', AICPA, Washington, D. C., 10 December 1996.

                          Wallman, S. M. H., 'Remarks Before the American Society of Corporate Secretaries
                          49th Annual Conference,' White Sulphur Springs, 1 July 1995.

                          Watts, R. L., and J. L. Zimmerman, 'Towards a Positive Theory of the Determination
                          of Accounting Standards', The Accounting Review, Vol. 53, No. 1, 1978.

                          Zeff, S. A., 'Commentary: A Perspective on the U.S. Public/Private-sector Approach
                          to the Regulation of Financial Reporting', Accounting Horizons, March 1995.

                          HAIM A. MOZES is Associate Professor of Accounting at Fordham University. The
                          author would like to acknowledge the input provided by professors Robert Halperin
                          and Georgia Saemann, and the helpful suggestions provided by the Editor.
 
                              Article A21266064