Fair-Value Guesswork Is Best Left to Investors

As I struggled to navigate a roundabout at London’s Heathrow Airport recently, I wondered why the Brits don’t switch to driving on the right side of the road. The main obstacle, of course, is the difficulty of ensuring that all vehicles make the change at the same time. Just a few holdouts, particularly trucks, could cause major problems.

This is an appropriate analogy for the current chaos in companies’ financial statements as standard setters embrace fair-value accounting. Some recent examples:

-- Fair-value loss equals accounting profit: During the third quarter of 2011, growing worries about the sovereign-debt crisis caused Bank of America Corp.’s stock price to plunge to $6.12, from $10.96. Yet the bank’s financial statements for that period show that book value per share rose to $20.80 from $20.29, and the company reported earnings per share of 56 cents, far exceeding analysts’ estimates and setting a new post-financial crisis record.

How did this happen? Current fair-value accounting rules allow the bank to elect to treat the decline in the market value of its debt as revenue, and that is what generated the bulk of the profit. The idea is that the equity holders are better off because there is now a greater chance that the bank will go broke and not pay off its debt. But the presumed reason that the fair value of the debt dropped in the first place was because the value of the bank’s assets declined.

Yet under current accounting rules, the decline in the fair value of the bank’s assets is largely ignored. For example, the loan portfolio, its largest asset, is mostly valued at amortized cost. Perversely, the result is that a decline in the fair value of Bank of America’s assets leads to increased revenue, earnings and book value.

-- Missing fair value: More generally, the current piecemeal approach to fair-value accounting has done little to bridge the gap between companies’ market values and book values.

Take the case of Apple Inc., the world’s most valuable company. Its market capitalization exceeds $500 billion. Yet its book value of equity is only $102 billion. A quick glance at the balance sheet shows that most of this book value relates to cash and investments, which total about $110 billion. The net book value of the assets and liabilities related to Apple’s technology business amounts to minus $8 billion. This total is negative because Apple’s technology assets are valued at less than the amounts owed to suppliers and employees. If we assume that the fair value of cash and investments approximates their book value of $110 billion, the market is essentially valuing Apple’s technology business at more than $400 billion, even as the financial statements value it at minus $8 billion.

The main reason for this gap is that the accounting for Apple’s technology business is still based on the traditional model, which summarizes past transactions. Market values, by contrast, are determined by investors’ expectations about future transactions. The traditional model reports relatively reliable information about past transactions, leaving the more subjective task of forecasting future cash flows to investors.

Nevertheless, the move toward fair-value accounting is increasingly requiring accountants to forecast future transactions and their associated cash flows. It is far from clear that this move is proving helpful to investors. The current method of accounting for intangibles illustrates why.

-- Fair or unfair?: Current accounting rules don’t generally require companies to record their nonfinancial assets at fair value, but there are exceptions. Perhaps the most significant relates to “indefinite-lived” intangible assets, such as the goodwill arising from acquisitions. Current accounting rules make no attempt to systematically amortize these assets against the revenue they are expected to generate. Instead, they are initially valued at cost and then subject to periodic impairment tests to determine whether their fair value has fallen below their book value. If it has, book value is written down to fair value.

So how effective are accountants in making these fair-value adjustments? Take the case of Boston Scientific Corp.’s purchase of Guidant Corp. in 2006. The acquisition raised Boston Scientific’s total assets from $8 billion to $31 billion, adding $20 billion of intangibles. The combination proved to be a dud, generating no significant increase in profit and free cash flows. Investors figured this out, and Boston Scientific’s market capitalization is now around $9 billion, down from more than $25 billion at the end of 2006. But the company is still carrying $16 billion of intangibles on its balance sheet, mostly goodwill from the Guidant acquisition. This despite the losses the company has generated in five of the last six years. As a result, the medical-device maker currently has a book value per share of about $8, and a stock price of about $6.

Other companies with disconcertingly high fair values for their intangibles include Sprint Nextel Corp., Legg Mason Inc., AOL Inc., EchoStar and CME Group Inc.

Of course, estimating the fair value of a complex business isn’t simple. Boston Scientific acknowledged as much on Page 63 of its recent 10-K filing: “Although we use consistent methodologies in developing the assumptions and estimates underlying the fair value calculations used in our impairment tests, these estimates are uncertain by nature and can vary from actual results. The use of alternative valuation assumptions, including estimated revenue projections, growth rates, cash flows and discount rates could result in different fair value estimates.”

Perhaps this is why accounting rules have traditionally left the job of estimating uncertain fair values to investors.

-- Solutions: The traditional accounting model is organized around recognizing revenue from past transactions and matching those with associated expenses to produce periodic earnings. The primary role of the balance sheet is to keep stock of benefits and obligations relating to past transactions. For example, past sales that are owed to a company are recorded in accounts receivable. The key summary measure of performance under the traditional model is the earnings attributable to periodic revenue. Investors use this information as a starting point for projecting future earnings and cash flows, which then are used to estimate company value.

A pure fair-value accounting model attempts to directly value a company on its balance sheet. We currently only see this model in situations in which the fair value of essentially all a company’s assets and liabilities is readily observable, such as with stock mutual funds. Although this model works well for the likes of the Vanguard S&P 500 Index Fund, it is much more problematic for complex companies such as Bank of America, Apple and Boston Scientific.

Current accounting rules are an incoherent mixture of the two models above. They cling to traditional revenue-recognition rules, while introducing ad hoc fair-value adjustments. This results in balance sheets that rarely do a good job of measuring a company’s fair value and income statements that are so peppered with fair-value adjustments that it is hard to estimate a company’s sustainable earnings.

Managers have responded with a bewildering array of “pro forma” earnings adjustments. But in many cases, these adjustments take legitimate costs of doing business and treat them as if they were irrelevant.

An important objective of financial statements is to produce useful measures of company performance. The current accounting system is increasingly falling short in this respect.

Accounting standard setters should stick with the traditional model and leave the task of estimating uncertain fair values to investors. Readily observable fair values are, in many cases, best left to other comprehensive income or footnote disclosures.

In the meantime, investors would be wise to navigate corporate financial statements with extreme caution.

(Richard G. Sloan is a professor of accounting at University of California, Berkeley’s Haas School of Business and a contributor to Business Class. From 2006 to 2009, Sloan was director of equity research at Barclays Global Investors. The opinions expressed are his own.)

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

    To contact the author of this story:
    Richard G Sloan at

    To contact the editor responsible for this story:
    Max Berley at

    Before it's here, it's on the Bloomberg Terminal.