Our November column challenged the status quo by using the “Four-Way Test” to show that GAAP-based financial reporting is unethical. As if that situation isn’t bad enough on its own, GAAP’s shortcomings also contribute to inefficiency in the capital markets and the economy.

Our December column, “Debunking myths about market values — Part 1,” continued the challenge by identifying the overall lack of market value measurements as GAAP’s most significant ethical flaw. We suggested this defect can be overcome only when accountants fully comprehend that their reluctance to use values is sustained by at least 14 false myths.

That column debunked two myths by showing that market values aren’t single points but distributions of results of multiple transactions, and that assets’ original costs (and liabilities’ proceeds) don’t necessarily equal their initial market values.

Let’s look at four more.


“Users don’t want to know market values.”

We have no clue who conjured this bonehead idea, and we can’t comprehend how it has remained entrenched.

Those points don’t matter, though, because this myth was totally debunked over 20 years ago in Financial Reporting in the 1990s — and Beyond, written by highly credible statement users and published by what is now the CFA Institute. These quotes prove this myth is completely false:

  • “It is axiomatic that it is better to know what something is worth now than what it was worth at some moment in the past.”
  • “There is no financial analyst who would not want to know the market value of individual assets and liabilities.”
  • “In Financial Accounting Standard No. 33, we were provided with information that, although imprecise, was a godsend to those financial analysts who understood it and were able to use it in their work.”
  • “In a phrase, analysts prefer information that is equivocally right rather than precisely wrong. Inexact measures of contemporaneous economic values generally are more useful than fastidious historic records of past exchanges.”

In light of these clear-cut statements, we suspect that only two kinds of users would not want market values to be reported to the public. They consist of those who barely understand financial statements, or those who have somehow developed workable value-estimation techniques. The former would resist any change that would add to their bewilderment, while the latter wouldn’t want to lose their profits from market inefficiency.

We shed no tears for either group.


“Market values for assets (or liabilities) are irrelevant if management doesn’t plan to sell (or settle) them.”

We speculate that most defenders of this dog-eared myth simultaneously mutter something about professors and an ivory tower. However, reality contradicts its message.

First, market values of individual mission-critical assets are relevant even if selling them would disrupt operations, because that more-complete information enables more reliable estimates of an entire entity’s going-concern value. To explain, users begin by assessing a company’s future cash flows. Generally, those flows are achieved by using its assets, not by selling them. Because market values represent a consensus about the assets’ potential cash flows from being used, their amounts provide powerful insight into the company’s aggregate future cash flows and its intrinsic value. Therefore, assets’ market values are useful even when they aren’t for sale.

Second, liabilities’ market values provide more useful information about the current economic impact of future cash outflows than obsolete book values can ever deliver.

Third, we dispute the premise that rational managers would ever blindly commit to use assets or leave liabilities outstanding indefinitely. Because business conditions continuously evolve, market value changes cannot be safely ignored. Suppose that real estate under a factory has appreciated because it’s located in a newly booming residential area; a wise manager would sell it if current operations produce an inadequate return on its now-enhanced value. Bottom line: It’s folly to keep reporting cost-based numbers for assets and liabilities because doing so not only misleads users but also shields managers against fiscal discipline.

Finally, we remind readers of the declaration that, “There is no financial analyst who would not want to know the market value of individual assets and liabilities.”


“Market values are unreliable because they’re hypothetical predictions.”

This myth mistakenly associates market values with future events.

Just like original costs, market values are historical facts produced through observed past events, with the key difference that the events are recent. Because financial reporting is supposed to help statement users assess future cash flows, they should be provided with these clear, dependable and up-to-date facts about the most current conditions.

Unfortunately, we believe the Financial Accounting Standards Board fell victim to this myth in SFAS 157 (ASC 820) when it awkwardly defined “fair value” as (emphasis added): “The price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date.”

By using the conditional verb tense (“would be”), the board pushed the time frame for measurement into the unobservable future, thereby forcing accountants to predict amounts that might possibly happen, instead of observing those that have already happened. The board could have avoided this flaw if it had acknowledged that values are past facts, not future guesses.

With emphasis added, we suggest this phrasing is better: “The price that multiple other market participants actually received upon selling similar assets or actually paid to settle or transfer similar liabilities in orderly transactions at or near the measurement date.”

This definition would compel accountants to observe recent market transactions, instead of vainly trying to predict what might occur in the future.

For what it’s worth, our personal conceptual frameworks include this axiom: “Market values are reliable because they’re actually observed.”


“Market values can be reliably calculated.”

This claim greatly oversimplifies the multiple simultaneous and dynamic factors that create market values. Simply put, a transaction’s amount emerges from a confluence of many supply and demand conditions at a particular time.

In reality, supply and demand continuously shift because of changes in such things as buyers’ and sellers’ assessments of the exchanged items’ values, alternative means for meeting the two parties’ needs, the symmetry of information between them, and competition from other buyers and sellers. At best, very few situations allow anyone to know enough about all those factors to calculate a precise market value.

Nevertheless, if there are no observed recent transactions, it may be helpful to carefully estimate market values after identifying and quantifying some of those factors. However, the makers and users of those estimates should not attribute very much precision or accuracy to them.

To explain one factor that limits precision, we learned long ago in high school science about the significant digits constraint. Briefly, it holds that a calculation can be no more precise than the least precise of its inputs. For example, the sum of 2.1 and 5.7753 is 7.9, not 7.8753; also, their product is simply 12 with no decimal places, not 12.12813. Both results reflect the fact that 2.1 has only two significant digits.

Alas, accountants always disregard this limitation. For example, IBM recently reported a projected pension benefit obligation of $97,837 million. Surely, a factor with only one significant digit lurks in the long chain of actuarial calculations behind this spuriously precise number.

If so, the only legitimate measure would be $100 billion ± $50 billion and IBM’s reported net income and earnings per share would have only one significant digit. That outcome should rock the whole financial reporting world!

Because many GAAP measures have only this kind of illusory precision, we believe those who present them are acting unethically, even if they’re unaware of what they’re doing.

Also, other errors abound because most accountants don’t fully grasp how to estimate market values with present values. To usefully estimate, say, a liability’s market value, the present value calculation must apply the discount rate that reflects the risk inherent in both the reporting company’s own creditworthiness, and in the uncertainty surrounding the amounts and timing of the liability’s future payments.

We find that FASB significantly erred in both respects for pension liabilities. First, the board mandates a discount rate based on other companies’ creditworthiness.

In particular, SFAS 157 (ASC 715) requires employers to use the return rate from a “portfolio of high-quality debt instruments.” Unless a specific company’s credit rating is “high-quality,” there’s no way under the sun that using that broad market rate will produce a fully useful estimate of the liability’s market value.

Second, FASB’s specified discount rate is derived from traded bonds with virtually certain contractual payments. We think the board stumbled badly when it chose that rate because the expected cash outflows for investment-grade debt are much less risky than the unknown benefits an employer must pay to an unknown number of retirees for unknown periods of time.

In summary, no one can reliably calculate precise market values. It’s far better to find ways to observe recent exchanges of similar assets and liabilities. If that option is infeasible, those who compile estimated values must be careful and ever-mindful of the applicable constraints.

Likewise, no user should ever assume that any estimated values are precise beyond one or at most two significant digits.


We’ve now debunked six myths that have misled accountants and inappropriately discouraged them from incorporating market values in GAAP financial statements. We’ll keep speaking out until better practices make those statements both fully ethical and fully useful.

Doing anything less makes no sense whatsoever.

Paul B.W. Miller

Paul B. W. Miller is an emeritus professor at the University of Colorado at Colorado Springs.

Paul R. Bahnson

Paul R. Bahnson is a professor at Boise State University.