International Accounting Standard 39 has been the most controversial of all IFRS rules. Fair value accounting requires that for each class of financial assets and liabilities a firm must disclose value in a way that permits it to be compared with the corresponding carrying amount in its balance sheet. By definition, fair value is the value paid by a willing seller to a willing buyer under other than fire sale conditions.


Fair value accounting is by now second nature to American companies and to those foreign companies quoted on US stock exchanges, and therefore using US GAAP. The concept of fair value accounting was introduced in 1999 with SFAS 133. But this notion is rather new to other firms. As such, it represents significant cultural change requiring:

  • Marking to market inventoried derivatives and other instruments for fair value reasons; and
  • A departure from the classical accruals method leading to book value, which can be quite inaccurate.


Under IAS 39 some of what is currently shareholder funds will be classified as liabilities. For instance, if a bond pays no cash but shares, it should be seen and reported as a liability. As far as the banking industry is concerned, this can have significant impact on the method of meeting Tier-1 capital requirements.

For financial instruments such as short-term trade receivables and payables, when the carrying amount is a reasonable approximation of fair value, no disclosure of fair value is required. Otherwise, like SFAS 133 in the United States, IAS 39 expands the use of fair value for measuring and reporting on:

  • Financial assets
  • Financial liabilities, and
  • Derivative instruments.


IAS 39 also provides for limited use of hedge accounting, but sets criteria for recognition and derecognition. This comes beyond IAS 32, which calls for compound instruments such as embedded derivatives, to be split into their components, and accounted for accordingly.

Volatility in fair value is primarily, but not exclusively, due to market risk. For instance, currency exchange risk arises on financial instruments that are denominated in a currency other than the functional currency of the entity. Other examples of fluctuation in fair value are commodity price risk and equity price risk [In fact, equity itself is a commodity].

The rules of IFRS require that, in disclosing fair value, a company must group financial assets and liabilities into classes, and offset them only to the extent that their related carrying amounts are offset in the balance sheet. Moreover, in its financial statement an entity shall disclose:

  • Method and assumptions applied in determining fair values of financial assets and financial liabilities; and
  • Other hypotheses, such as estimated prepayment rates, rates of projected credit losses, interest or discount rates, and so on.


Behind these requirements lies the need outlined in the preceding two sections: that every entity must disclose information enabling users of its financial statements to appreciate its capital and evaluate the nature and extent of risks arising from financial instruments to which it has been exposed:

  • During the period, and
  • At reporting date.

As contrasted to US GAAP, IAS 39 integrates seven or eight different SFAS standards by FASB; not only the better known FAS 133 which addresses derivatives. For instance, part of IAS 39 are some procedures prescribed by FAS 91 regarding accounting for new loans. References to this broader coverage have been made in a meeting at IASB. It is proper to bring to the reader’s attention, however, that in the United States several agencies develop reporting rules. For instance, the 2005 Annual report by Microsoft states:

We account for the licensing of software in accordance with American Institute of Certified Public Accountants (AICPA) Statement of Position (SOP) 97-2, Software Revenue Recognition. The application of SOP 97-2 requires judgment, including whether a software arrangement includes multiple elements, and if so, whether vendorspecific objective evidence (VSOE) of fair value exists for those elements.


Microsoft’s 2005 Annual Report further states that changes to the elements in a software arrangement, the ability to identify VSOE for those elements, the fair value of the respective elements, and changes to a product’s estimated life cycle could materially impact the amount of earned and unearned revenue. It also points out that judgement is also required to assess whether future releases of certain software represent:

  • New products, or
  • Upgrades and enhancements to existing ones.


The same Annual Report underlines that Microsoft accounts for research and development costs in accordance not with one but with several accounting pronouncements, including SFAS 2, Accounting for Research and Development Costs, and SFAS 86 Accounting for the Costs of Computer Software to be Sold, Leased, or Otherwise Marketed. SFAS 86 specifies that:

  • Costs incurred internally in researching and developing a computer software product should be charged to expense until technological feasibility has been established for the product.
  • But after technological feasibility is established, all software costs should be capitalized until the product is available for general release to customers.


Judgement is, most evidently, required in determining when technological feasibility of a product can be established. And as the reader is aware from section 3, judgement is also present with practically every interpretation and implementation of accounting and financial reporting rules.

Finally, an interesting case to bring to the reader’s attention is that of fair value of debt. The fair value of a company’s debt is determined using pricing models that reflect one percentage point shifts in appropriate yield curves. The fair value of investments is determined through a combination of:

For starters, duration is a linear approximation that works well for modest changes in yields and generates a symmetrical result. In 1938, Frederick Macaulay developed the duration algorithm that bears his name. Its output can be used as a proxy for value over the length of time a bond investment is outstanding.

Over the years, the duration algorithm has evolved different versions. Modified duration, with greater price sensitivity, is an example. Another is average duration, assumed for positions that fall into each time band. More sophisticated pricing models are reflecting the convexity of the price/yield relationship. As such:

  • They provide greater precision, and
  • Take account of asymmetry of price movements for interest rate changes in opposite directions.


The impact of convexity is more pronounced in longer-term maturities and low interest rate environments. As this example demonstrates, modern financial reporting and management control not only involve assumptions that may be subjective, but also require appropriate choice of models. This brings into perspective model risk. The reader should keep these references in mind when evaluating an accounting system’s dependability, as well as the action of its implementers.