The more a company pays auditors for consulting and technology services, the more likely it is to meet or beat earnings projections, concluded a recent study undertaken by professors at Stanford University, the Massachusetts Institute of Technology and Michigan State University.
And Wall Street notices; a company's stock price tends to fall if it files a proxy statement that reveals large fees for nonauditing services--basically consulting and information technology services--from the accountant.
Wall Street's five largest accounting companies--Andersen (formerly Arthur Andersen), Deloitte & Touche, KPMG International, Ernst & Young and PricewaterhouseCoopers--all carry out nonauditing work, such as consulting or installing new information systems. According to the university study, proxies filed between Feb. 5 and June 15 indicate nonauditing business made up 67 percent to 74 percent of the fees paid by publicly traded companies to the Big Five. Information-systems projects generated 6 percent to 13 percent of total fees.
"Our results suggest that the provision of nonaudit services impairs independence and reduces the quality of earnings," write authors Karen Nelson, Richard Frankel and Marilyn Johnson.
Sounds dry, but it's a touchy topic on Wall Street.
Auditors are supposed to be unbiased as they pore over a company's books, but Nelson and her peers suggest that accountants with nonaudit businesses might be less critical of their clients' bookkeeping. The higher the nonaudit fees, the less independent the accountant tends to be.
The notion of independence itself is a point of contention. The term usually refers to an auditor's ability to make a judgment without being influenced by the client's desires, but the university study lacks a good way to measure independence, critics said.
"Although auditor independence is a critical variable in the study, it is never adequately defined by the authors," said Alan Anderson, a senior vice president with the accounting industry's trade group the American Institute for Certified Public Accountants (AICPA). "Instead the study assumes that the ratio of nonaudit services to total fees is an acceptable proxy, and that the higher the ratio the less independent the audit. The use of this ratio as a proxy for independence is questionable."
Naturally, accounting companies reject even the hint of a whiff of a scent of any notion that they're anything but objective.
"This academic study asserts that some auditors may be more or less independent, but in the real world of the auditing profession there are no shades of independence--an auditor is either independent or he is not," Anderson said.
The university study assumes companies that frequently meet or beat Wall Street forecasts by a penny are "managing" estimates, or adjusting their books or operations solely to meet analysts' financial targets. It's a game that securities regulators and many investors frown upon.
There's nothing intrinsically wrong with topping analysts' profit estimates. Ideally, all corporations would surpass expectations through normal business practices.
A company could "manage estimates "through bookkeeping tricks such as shifting a few dollars to the revenue line from accounts receivable; extraordinary actions such as cutting research and development spending temporarily; or unusual sales tactics to boost sales in the final week of a quarter--a particularly notorious practice in the enterprise-software industry. Or companies simply convince analysts to set earnings forecasts at a level that can be easily hurdled.
So when the university researchers looked at five months of data and found that companies meeting or beating estimates tend to pay their auditors more nonauditing fees, it appeared that managing estimates goes hand-in-hand with technology implementation services and consulting. Or at least auditors that provide those services are more likely to turn a blind eye to accounting practices that give an artificial boost to earnings per share.
Defenders of the practice argue that consulting work gives an auditor a deeper look at a company's business and thus a better way to evaluate the books. The fact that companies paying higher consulting fees to auditors tend to meet or slightly exceed analyst estimates might simply reflect the fact that a well-managed company uses consultants.
But probably not, in Frankel's view. "That's kind of a thin reed to me," Frankel said. "I don't necessarily buy into that argument."
Critical look at cross-consulting
Frankel and AICPA's Anderson agreed on at least one thing about consultants: They could be a sign of poor management, more than anything else. "Investors may see the purchase of significant nonaudit services as a sign of challenges within the company, not impaired auditor independence," Anderson said. "The (university) report itself makes this concession."
Taken to its logical conclusion, the study seems to suggest that accounting companies shouldn't be doing things like giving operational advice or installing enterprise resource planning systems for clients. At least two of the Big Five--Andersen and KPMG International--have separated their accounting arms from their original consulting businesses in the past year, though the university study indicates Andersen and KPMG continued to get paid for information systems services and other nonauditing fees.
Andersen still has another unit with 11,000 consultants, a company spokeswoman said.
Yet the professors are quick to say they don't believe the Securities and Exchange Commission needs to require an absolute division between consulting and auditing.
"I don't think our results could at all support a policy that extreme," Nelson said. "You could imagine cases where it (paying one company for auditing and consulting) is not a bad thing."
Nelson and Frankel back the SEC's current policy of disclosing information that investors can use to make their own decisions.
In fact, maybe the market does just that. According to the study from Stanford, MIT and Michigan State, companies with the "least independent" auditors underperformed the Standard & Poor's 500 or the S&P 400 Midcap Index by an average of 0.44 percent on the day their proxies were filed.
"It does appear the market is concerned with this information," Nelson said.
Increased government regulation would imply that investors need protecting. However, a large portion of stock is held by professional money managers--presumably trained, informed individuals who know how to read proxy filings.
"If you're a believer in market efficiency...you believe the market can put a proper value on these things," Nelson said.