by Alfred M. King

That I would agree with Messrs. Miller and Bahnson may strike some readers as a surprise, given our past histories, but fair is fair. They are not totally wrong about valuation.

In their recent articles “Value-based accounting II: Timeliness and reliability” (Feb. 23-March 14, 2004, p. 14) and “Value-based accounting III: What do users demand?” (March 15-April 4, 2004, p. 14) the authors make the following points about value information:


  • Information about current market values tells users what they need to know.
  • Market values represent consensus valuations for assets or liabilities that reflect the amount, timing and uncertainty of the future cash flows related to them.
  • Market values are neutral — that is, not biased — because consensus exists among multiple participants, including buyers and sellers separate and distinct from the reporting entity.
  • Market values can be verified.
  • Market values are a faithful representation of cash flow potential.
  • Market values can be represented by a statistical distribution of amounts paid and received in a large number of transactions.
  • The distribution’s mean is an expected market value that provides superior predictive information.

I agree with many of these points. However, while sound in theory, the arguments made, and the conclusions drawn, all rest on one fatal flaw.With the exception of marketable securities, there is rarely a single price or a single value for any specific asset. The following example is somewhat complex, but actually is the point of my argument. The world, and business, is complex and cannot be simplified through theoretical assumptions.
Take a 1,000-ton press. It has been installed, debugged and broken in. The press produces a limited number of high-value parts for the manufacturer, a defense supplier. Unfortunately, demand for the parts appears to be diminishing, so future cash flows are fraught with uncertainty. In short, the future cash flows associated with the press are very hard to estimate. In fact, they are dependent on the willingness of the Department of Defense to purchase the items; the DOD’s future demand, in turn, is dependent on future congressional appropriations.

There is an inactive market in 1,000-ton presses. A manufacturer will sell a new one at, say, $100,000 (completely made-up numbers). Freight, installation and debugging would run $50,000. Absent any inflation, the company has an original cost of $150,000 on the books and, let’s say, a current depreciated net book value of $110,000.

A used equipment dealer would pay $35,000, as is and where is. He would have to dismantle the press, hold it in his warehouse and hope to sell it to another buyer within three years for possibly $75,000. Put a different way, a used press could be bought for $75,000 and, with freight and installation of $50,000, would have a total value to a buyer of $125,000.

Last year’s volume of parts produced on the machine had a net contribution margin of $50,000, which provides sound economic support for the current $110,000 book value. Current projections, however, show that anticipated future sales of the parts may generate contributions of only $15,000 to $20,000 a year for the next seven years, at which time the parts may cease to be produced. There may be an asset impairment, but SFAS 144, on an un-discounted cash flow analysis, does not apply at this point. So under current generally accepted accounting principles there is no writedown.

Under the approach recommended by Miller and Bahnson, at what amount should the company value the press?

They argued that the relevant information for users is the anticipated future cash flow. So at this point, using a 10 percent discount rate and a $15,000 contribution for seven years, the present value of the future cash flow would be $73,000; if the sales generated $20,000 of future cash flows, the PV would be $98,000. To this we would add the anticipated sale of the press in seven years at $35,000, which has a present value of $18,000. So the PV of the future cash flows might be somewhere between $91,000 and $116,000. At $50,000 of annual cash flow for seven years, the PV would be $243,000.

What do we report if we anticipate, or hope, that Congress increases appropriations and last year’s $50,000 contribution appears sustainable for seven years? Should the asset be written up to $261,000, written up to $116,000 or written down to $91,000? Maybe the most conservative approach would be to put the press on the books at what it could be sold for to the dealer, $35,000!

This example is not ridiculous. It comes from real life. The Financial Accounting Standards Board, in Concepts Statement 7, said that in this situation one should apply probabilities. There might be a 25 percent chance of no future sales, in which case we sell to the dealer, a 20 percent chance of $15,000 contributions for seven years, a 40 percent chance of a $20,000 contribution, and a 15 percent chance of current sales levels being retained. If those probabilities are reasonable, the Concepts 7 fair value is $112,500, a write up of $2,500, even though that is the one value we know will never be realized.

So we have a choice, if we go to the Miller and Bahnson fair value model, of either writing up the asset, or writing down the asset. The only way a reader of the statements could make sense of these numbers would be if a footnote laid out the assumptions and probabilities, including the discount rate, just as shown above.

But this example, complex as it seems, is but one asset. Most manufacturing companies have many assets. There are not enough financial analysts or appraisers in the world to go through this exercise every year, much less every quarter for all assets of every company. The footnote disclosures would be incomprehensible! Following the recommendations of Miller and Bahnson, would we disclose all this information? Would it even be considered relevant, much less reliable?

I submit that the company’s current net book value of $110,000, combined with a discussion of the DOD sales outlook for the underlying product in the management discussion and analysis, will suffice for virtually all investors. Quoting from a 10-year old Association for Investment Management and Research study to support their position lends it little probative value.

Going to a fair value approach will make the complexities of today’s financial statements seem like child’s play. We already are at a point where the only people who can hope to understand a typical 10-K are professional security analysts. Even they will be baffled if companies have to go to the Miller-Bahnson fair value approach, with all its required disclosures.

What Miller and Bahnson propose is a triumph of theory over reality. In theory, I admit, they may be on to something. In practice, statement users will plead for mercy. If Miller and Bahnson get their way, security analysts will request that, as supplemental disclosure, companies go back to or retain the old-fashioned historical cost model. And if the historical cost information is what they really want, why should accounting theorists tell them they are wrong?

Miller and Bahnson have presented a “solution” that is still in search of a real-world problem.

Register or login for access to this item and much more

All Accounting Today content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access