Market conditions change — sometimes a lot — but the strange quarterly theater of earnings season is a constant.

The banking industry's next show begins in a month. Corporate executives will present earnings under generally accepted accounting principles, accompanied by a barrage of figures put forth as core, operating, and cash performance. They claim the various metrics serve to illustrate performance, while some analysts and investors complain that the alternative methods are intended chiefly to obfuscate performance.

The subjective nature of accounting lends ammunition to both sides of the debate, and the looming question of the efficacy of fair-value measurements is likely to get a more extensive airing in the coming months. Fair value's application in particular to familiar banking instruments such as asset-backed securities, structured instruments like collateralized debt obligations, and even leveraged loans introduces additional fudge factors in determining earnings.

Given the current unease in credit markets, investors may worry less about income and focus their attention more squarely on the quality of balance sheets. And that is just as well, because in their study of earnings per share, companies and investors are searching for an objective truth that probably does not exist.

"There is an understandable desire among the investment community to ask companies to quantify the quality of the economic value added during a quarter in a single number," said Howard Mason, an analyst at AllianceBernstein LP's Sanford C. Bernstein & Co. LLC. But "there is no such thing as a true, core number."

The elusiveness of a simple measure of performance promotes management posturing and Street resistance. In sanguine market conditions, parsing what constitutes extraordinary expenses and what constitutes operating costs may amount to nothing more than polite conversation about the nature of financial disclosure. When conditions sour, the arguments take on more urgency.

This summer's volatile markets have virtually guaranteed that the industry will report a rash of hits to earnings in the third quarter. Deteriorating credit quality will dictate large loan-loss provisions, and illiquid markets will force marks against the value of financial instruments. Volatile interest rates have prompted portfolio restructurings, and companies have shut down wilting business lines, recording special costs in doing so.

"You are going to see these types of charges this quarter, and you will hear banks urge investors to overlook them as almost one-time in nature, and to look elsewhere for the results of the core business," said Gerard Cassidy, an analyst at Royal Bank of Canada's RBC Capital Markets.

Some of those charges may legitimately distort companies' results — particularly discontinued businesses — but Mr. Cassidy said banking companies that stretch credulity in casting operating expenses as special items aren't fooling anyone.

"The market sees through this stuff in a nanosecond," though that won't deter companies, he said. "They are all going to try to dress up the pig" because "most management teams seem to think they have to."

Mark Fitzgibbon, the director of research at Sandler O'Neill & Partners LP, said analysts are gearing up for a "sloppy" quarter that will generate plenty of discussion about core performance.

"There is definitely some gamesmanship and there will probably be more gamesmanship this quarter than is typical," he said.

Fair value is likely to be at the center of some of those games, particularly at the large banking companies that run businesses and hold portfolios that fall under that accounting regime. Though the methodology is not new — banking companies have applied it to their trading portfolios for years — the Financial Accounting Standards Board issued a standard last September that spelled out how it should be applied throughout financial statements. Most large banking companies, including Citigroup Inc., Bank of America Corp., and JPMorgan Chase & Co., adopted the standard earlier this year.

The statement's most simple articulation of fair value is the price of a financial instrument "in an orderly transaction between market participants," and calculating fair value for securities in liquid markets requires little heavy lifting. Liquid securities in active markets fall in the top level of the standard's hierarchy, in which valuation may be as easy as multiplying the number of the securities by quoted prices.

Bankers and their lobbyists have long complained that the bulk of their assets aren't "Level One" assets. Loans, securities, and other financial instruments held by banking companies tend to fall in the two lower levels of the hierarchy. Instruments that are not actively traded but are similar in many respects to those that are fall into "Level Two"; instruments that have what the standard calls "significant unobservable inputs," and therefore rely on extensive modeling to determine values, are "Level Three."

The standard's dependence upon internal valuations has meant that mounds of securities held by banks are marked to model, rather than to market. That has raised substantial questions about the reliability of the models and the internal controls surrounding model development and input.

According to the standard, "unobservable inputs shall reflect the reporting entity's own assumptions about the assumptions that market participants would use in pricing the asset or liability."

In other words, the instruments are worth whatever the company says they are worth, with the significant conditional that the company must document and support its conclusions.

Not surprisingly, the instruments at the center of this summer's liquidity crisis — asset-backed securities and collateralized debt obligations — frequently fall under the two lower levels of the standard's hierarchy.

The standard requires that value be "considered from the perspective of a market participant that holds the asset or owes the liability" — meaning that the instrument is valued at an exit price, rather than an entry price. That's an important distinction in credit markets flooded with bids and few matching asks.

The difficulty of applying fair value to illiquid assets is feeding an army of audit consultants that specialize in the task, and even the most seasoned professionals — most of them die-hard fair-value supporters — admit that its application this quarter has been a maddening task. Competing well-documented, good-faith efforts at applying the standard to similar securities can result in wildly varying values.

"There is a lot of interpretation in accounting, and this is a very good example of it," said Mr. Cassidy. "Without anybody doing anything wrong or illegal, you have one entity that believes that its portfolio should be accounted for at a certain value, and provides its justification, and another entity comes up with a totally different value with a totally different justification — and both of them could be legitimate."

How banking companies reflect fair-value measurements in financial statements depends largely upon the portfolio in which the instruments are held, and companies have some discretion in making that decision. A company that elects to hold an instrument until maturity in most cases will not use fair-value accounting. Another company could hold the identical security in a trading portfolio and recognize gains or losses on the security through marks to revenue.

Fair value therefore jumbles the models and calculations of a generation of investors that has long looked to loan-loss provisions as the ultimate barometer of credit costs. That in itself does not undermine fair value's legitimacy, and may in fact accurately reflect the conglomeration of credit and liquidity risk.

"If a structured-credit desk in a trading business took a hit on an instrument that it bought and sold in two days, you wouldn't expect it to show up in provisions," said Mr. Mason. "You're very aware that it's a credit-related loss, but you are perfectly comfortable with the idea that it doesn't show up in provisions."

"Do you interpret the trading losses from a CDO portfolio as credit losses?" asked Mr. Cassidy. "We might say the root of the loss is due to credit, but to the owner of that security, it is a trading-revenue loss."

When the losses are netted against trading gains, it can become virtually impossible to detect the magnitude of a company's credit costs — unless banking companies offer more help than they have in the past, and walk investors through their quarterly hits and misses.

Companies active in leveraged lending are likely to face some of the most vexing accounting questions. They must account for commitments using fair value, so deterioration in credit markets can trigger a mark against earnings before the loan is even extended. When the borrower draws upon the loan, the lender may have a measure of flexibility in choosing where it will hold the credit — and that may determine whether the lender recognizes losses through higher provisions or in trading accounts.

Bank managers and investors will continue to wrangle over the precise earnings per share every quarter, even as the futility of the exercise is increasingly apparent to all parties. But in the current environment, the balance sheet is in the ascendancy, and the income statement secondary.

"If the quality of the balance sheet is questioned, there is no reason to look at earnings or buy the stock," said Richard Bove, an analyst at Punk, Ziegel & Co. "I think this quarter banks are going to focus on defending the valuations they've used for their loan portfolio and their securities."

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.