If you are a registered user please log in to see more postings.

The AAO Weblog covers accounting issues and current events as they relate the practice of investment analysis.

Author: Jack Ciesielski Created: 10/13/2006 2:54 PM
The AAO Weblog is a weblog published by Jack Ciesielski , dealing with accounting issues and news topics related to investment and finance.

More mail than I ever expected from Monday's posting about the capping of auditor liability in audit client arrangements. This one's from Mark M., a former Big Four partner:

"I believe a point is being missed here. The cap only applies to the Company itself suing the auditors, not to third parties doing so. In effect, the attention in the engagement letters is apparently being directed at companies who make mistakes, intentionally or unintentionally, and then sue the auditors for their own mistakes. This does appear to be an area where some tightening is warranted in the interest of fairness.

As a former Big 4 audit partner, this issue would have no affect on my perceived audit risk due to the specter of the responsibilites to third parties. In general, a real disservice is done to auditors by the notion that the risk of litigation is the only thing standing in the way of auditors abdicating their responsibilities. This is frankly ludicrous. All auditors want to do a good job and, between the internal QC reviews and SEC oversight, all know there are plenty of people looking over their shoulders. I am convinced that the inclusion of such provisions will have no effect on the quality of audits."

Excellent point: this is an arrangement between auditor and firm, not shareholder and auditor. There'd still be a legal avenue available to shareholders in the event of an audit failure where the auditor was culpable. And I'd agree that management is the one party that's primarily responsible to shareholders for management's faults and errors.

At the same time, there's the problem of perceptions. I appreciate the fact that auditors don't usually set out to do a crummy job, and that they know there's plenty of regulatory menace waiting if they fall off their tightrope. And I believe that the system has improved, and is producing far better results than it did just three or four years ago. (Despite news like this. I haven't gotten hold of the report yet; any PwC'ers out there care to share it?) I agree that it isn't fair to paint all auditors badly, as Mark suggests has happened.

But - auditing is a tough business, because perceptions count in auditing more than in any other profession. That's why independence in appearance is just as important as independence in fact. That's where the profession went off the rails in the 1990's with client consulting arrangements - they badly damaged the appearance of auditor independence, if not in fact. Seeking caps on liability doesn't do much to make the public at large think that the auditing profession has moved their image (their appearance) from being lap dogs to being shareholder allies.

* * * * * * * * * * * * * * *

Coming up: the results of the Statement 123R "are you going to use the GAAP figures?" survey. Stay tuned.

I received a couple of comments on yesterday's posting about the capping of auditor responsibility for failed audits. First, a correction, courtesy of John A., about my attribution to Mark Twain about the prospect of hanging:

'Twas Samuel Johnson, via Boswell: "Depend upon it, sir, when a man knows he is to be hanged in a fortnight, it concentrates his mind wonderfully." -- Boswell: Life

I stand corrected. Must have been thinking about Mark Twain's famous quote, "The rumors of my death have been greatly exaggerated." (At least I think it was Mark Twain.)

From accounting doctoral student, Steve S., in California:

I just finished reading about Sun Microsystems in "Capping Responsibility" and your comments about the benefits (or lack thereof) of liability caps to investors. Won't firms pay less for liability-capped audits? And isn't that a benefit to investors? The question is who bears the risk. If investors are well diversified and accounting firms are not, then it might make sense for investors to bear the risk of audit failure rather than the accounting firms, assuming investors are compensated for doing so by way of lower audit fees. Not to mention your point about firms taking financial reporting more seriously. The remaining question, I think, is whether these liability caps reduce the quality of audits since auditor liability is reduced.

I can't agree with very much in this, Steve, but let's start with the "remaining question": whether these liability caps reduce the quality of audits since auditor liability is reduced. That really is the only question, as far as I'm concerned, and what the whole issue is about. So turn it around: just how much inferior quality is acceptable in order to get a lower price on the audit fee? Ask an investor who has a significant stake in a firm and treats it as ownership, not rent. Accepting less quality to save a few bucks would not work in the interests of investors who work in such a "concentrated" style. Not all investors seek diversification.

Of course that's an argument for "well-diversified" portfolios - but if all firms accept shoddier audit work to save on audit fees, does diversification eliminate the blow-up risk? Doubtful; not if all firms systemically have lower-quality financials. Diversification won't help.

Another issue: "investors are well-diversified, and accounting firms are not." In the post-Sarbox era, I think that investors automatically believe that accounting firms do nothing but audits, and that they've given up on consulting services. Wrong. The accounting firms are reaping more audit fees, and they've been barred from consulting for clients - but they are still very active in consulting. It's just not as visible because we don't see any consulting fees paid by audit clients. They're more diversified than you might think - and able to absorb audit risk.

Wouldn't firms pay less for liability-capped audits? Well, they should - they're receiving less for their money. But they probably wouldn't - not if there's very little choice in who will be the auditor. It's tough to negotiate a price down when there aren't many choices to play off the current supplier of services. I'm not suggesting that there's cartel pricing in the auditing world, but you can't deny that pricing and negotiations are tougher when there are fewer alternatives.

Lower audit fees a benefit to investors? I don't think of this as the same thing as raising your deductibles on your insurance. There's a public benefit to having high quality audits: it's called confidence in the markets. Maybe firms would report higher earnings (and I don't think that matters as much as others - more in a minute) if audit fees were lower for liability-capped audits, but markets might demand compensation in the form of lower multiples.

As for the dollar benefit of lower audit fees, lower audit quality: I don't believe audit fees aren't as bad as you might be conditioned to believe by all the griping in the press about higher audit costs after Sarbanes-Oxley. In 2004, after firms should have incurred the brunt of the increased audit vigilance, only 38 firms in the S&P 500 had after-tax audit fees that were greater than 2% of diluted EPS. And 318 S&P 500 firms - 64% of them - have zero impact on EPS from after tax audit fees. So, I'd be less than thrilled as an investor if audit fees went down just because I'm shouldering more risk; there just isn't much bang for the buck.

In fact, maybe auditors should scrap the caps and keep fees and quality high. Now that's a novel idea.

Interesting article in this morning's Wall Street Journal, a maybe-phenomenon that I hadn't come across before.

The article describes an unusual disclosure in Sun Microsystems' proxy describing the "liability cap" arrangement between Sun and its auditor, Ernst & Young. The cap limits the ability of Sun to sue Ernst & Young in the event of an audit failure; as the article puts it, the audit arrangment is "subject to alternative dispute resolution procedures and an exclusion of punitive damages." It also mentions that Silicon Graphics is another liability capper.

(Know of any others? Send 'em in.)

Nothing focuses the mind more clearly than the prospect of hanging, to paraphrase Samuel Johnson. And he was right. Obviously, such an arrangement gives the auditor comfort that a single audit won't drive them into oblivion a la Arthur Andersen - and removes the prospect of hanging. The article mentions that, anecdotally at least, all of the Big Four firms have demanded them, making it impossible for some firms to find an auditor keeping all of their skin in the game.

At first blush, these are not investor-friendly arrangements; it's hard to see where taking the heat off the auditors is going to produce higher-quality audits for investors. (The only investor-friendly rationale I can manufacture is that this preserves the financial health of the auditing profession. But if they aren't concerned with the intensity of audits - how is the investor better off with a financially healthy audit profession?)

There's only one silver lining to this: for every action there's a reaction. If firms lose an avenue of reparation due to shoddy reporting, maybe they'll take more seriously their own first-line responsibility for producing high-quality financial reporting. Take the heat off the auditors, it's going to go somewhere else; this should raise the stakes for the reporting companies. That's not a bad thing. A slim straw to grasp, but a straw nevertheless.

The next meeting of the SEC's Advisory Committee on Smaller Public Companies has been scheduled for December 14, 2005. This is the group that's been championing the rollback of securities laws, especially Sarbanes-Oxley, for the benefit of smaller companies. (Who apparently have no problem wanting to stay small, at least when it comes to self-discipline in the area of financial reporting controls.)

No details yet on what they plan to discuss. More as developments occur.

* * * * * * * * * *

Apologies for the thin postings today. It's a short week, as I'm sure you know, and there's travel ahead - and a torturous Analyst's Accounting Observer to write. Which I will now strap myself into writing.

Happy Thanksgiving to you, and thanks for tuning me in.

[Note: the PCAOB and KPMG subscribe to The Analyst's Accounting Observer. Not that it has any bearing on the comments below, but someone might care.]

Around the beginning of last month, the Public Company Accounting Oversight Board released its annual inspection reports on half of the Big Four: KPMG and Deloitte & Touche. As you'd expect, the reports weren't flattering. The PCAOB does not exist to hand out "attaboys" to the audit profession, it exists to keep spurs under auditors' saddles.

Last week, it was time for the other half of the Big Four to take a turn in the woodshed. Ernst & Young was criticized for not gathering sufficient competent evidence to support its audit opinion for a number of 2003 audits in the PCAOB sample. Some of the issues involved in the audits: improper operating treatment for a lease that should have been capitalized; allowing presentation of only two segments in a client's financials, obscuring operating losses at one segment that should have been displayed; a failure to confirm terms of an agreement amounting to 15% of annual revenue; and the proposal of judgmental audit adjustments for obsolete inventory without support. In auditing the allowance for loan losses for a group of four financial institutions, in three cases the E&Y audit teams had failed to gather documentable evidence that the "unallocated component of the allowance represented a supportable estimate of probable losses inherent in the issuer's loan portfolio." In short, E&Y couldn't prove that it had really examined the squishiest part of the institutions' allowance for loan losses. In another case, E&Y workpapers didn't support that there had been enough analysis done to support the client's treatment of a divestiture as discontinued operation.

PricewaterhouseCoopers didn't fare any better. Deficiencies were found in auditing of accounts receivable (insufficient evidence obtained, improper sample sizes, lack of follow-through on differences found); failure to evaluate whether certain commodity contracts of one client were, in fact, derivatives; and deficiencies in internal control testing (improper reliance on controls that hadn't been tested for several years, reliance on controls tested only in first half of year, tests of controls based on a too-small sample size). Other problems found in audit client financial statements that escaped PwC detection: incorrect tax accounting for foreign subsidiary; improper inclusion of securities as cash equivalents; and recalculation of allowance for doubtful accounts using the client's methodology -without testing the methodology's soundness. There are many more similar findings.

It's not encouraging stuff to read; you'd like to think that the Big Four would do a higher-quality job in what should now be their mainstay business - auditing. But you also have to recognize that the tension between regulator and regulated is what makes the process work for investors. If the day ever comes when the PCAOB doesn't find some flaw in the way public accounting firms operate, investors should be concerned that the regulator isn't doing its job.

No, not the holidays...

It was just about this time last year that the "Great Lease Restatement" was kindled, instead of the fireplace. A year ago this Wednesday, CKE Restaurants filed its 8-K indicating there were problems with its lease accounting, the first one of any note. Gradually, more and more firms started restating; mostly KPMG clients like CKE Restaurants, then other firms. (Note: KPMG is a subscriber to The Analyst's Accounting Observer.) Eventually, the letter that launched a hundred restatements was issued by the chief accountant of the SEC.

Back to the question: is it that time of year again? Maybe; in the last month, there's been a brace of restatements having to do with failed hedge accounting using derivatives.

Last Thursday, Pride International filed a non-reliance 8-K indicating that had incorrectly accounted for swaps and caps as an integral part of certain loans rather than as discrete derivatives. Proper accounting would have marked the instruments to market, with changes in fair value going through other income. (Their auditor was PricewaterhouseCoopers.) In late October, Kilroy Realty also indicated that their hedge accounting needed to be revised: their documentation of hedge accounting was insufficient to allow hedge accounting treatment, but the 8-K offered few details on what was lacking. Their auditor was Deloitte & Touche.

Also in the last couple weeks, three regional financial institutions reported on non-reliance 8-K filings that their use of the "short-cut method" of testing hedge effectiveness was incorrect. The institutions: Pulaski Financial, Provident Bankshares, and Taylor Capital Corporation. Going back into late October, South Financial Group filed a non-reliance 8-K for the same reasons. All four of these institutions share KPMG as their auditor.

Time out for a very quick background on the accounting issue here. When firms use derivatives to hedge a risk - say, the fair value of a brokered CD or an interest rate swap - they have to designate (by documentation) that such a hedge has been created and they must continually monitor the effectiveness of the hedge. "Effectiveness" means that the relationship that was set up between the hedged item and the hedging derivative instrument is still working at later reporting periods. If the effectiveness no longer exists, the hedge accounting is discontinued and any gains or losses on the hedge accounting are reported in period's earnings where the hedge accounting failed. This is a lot of follow-up and documentation that firms would prefer not to do; for obvious reasons, such testing of the relationship is called "long-haul" testing. There is an alternative built into Statement 133 called a short-cut test for effectiveness. At its core, it assumes that a cash flow hedge or fair value hedge is effective if it meets the (many) criteria listed in paragraph 68 of the standard. Meet those criteria, and no long-haul effectiveness testing needs to be done. Effectiveness is presumed.

This is where the four institutions had their problems: their didn't actually qualify for the shortcut treatments. They'll all go back and restate earnings with the changes in fair value of the derivatives flowing through earnings. Two of the firms - Provident and South Financial -said that going forward, they'll re-designate the derivatives as hedges and use "long-haul" testing to monitor effectiveness and regain hedge accounting treatment.

There's certainly a sense of deja vu here, what with the time of year and the KPMG presence. It's way, way too early to expect a restatement trend for termination of "short-cut" derivatives accounting. For now, consider it a coincidence.

Hats off to Standard & Poor's.

(And it has nothing to do with the fact that their equity research department subscribes to The Analyst's Accounting Observer.)

As reported in today's Wall Street Journal, the ratings agency will start to count options expense when it comes to earnings of their S&P indexes - like the mainstay S&P 500.

In my view, they're taking the high road. The focus of accounting standard-setting should be to make financial statements more useful and relevant for the users of financial information. The FASB has done that with its standard on stock option accounting. Earnings will report all costs involved in producing a bottom line after January 1, 2006. Enough is enough. Well done, S&P.

Here's a good question from Ryan S., a portfolio manager:

"Wouldn't it be useful for companies to show a Trailing-Twelve Months (TTM) Income Statement and three different balance sheet dates in their 10-q and/or K's?

Naively, it doesn't seem to me this would entail a lot of additional work on the companies part.

As you know, when a second quarter 10-q arrives it will show the 3 months and 6 months I/S results.

I'd much prefer the 3 months and a TTM I/S.

Then for some industries (retailing comes to mind) with seasonal aspects the presentation of two Balance Sheet dates (current and previous year-end's) isn't all that beneficial (to me). You need the last years comparable quarter to get an idea of how Working Capital is trending and such.

Maybe I'm just getting lazy in my old age. But, it just strikes me that if the goal is to give investors (and others) useful information making these changes would seem a step in that direction."

The easy answer is that financial statements are presented this way because that's all that Regulation S-X requires of publicly-traded companies. I agree: it would be much more useful for investors to spend time with financial statements prepared in the fashion you describe. And there is nothing to prevent companies from preparing them that way, either; the requirements in place are minimums, not prohibitions against releasing additional information. Figure this: companies can present rickety pro forma earnings under Regulation G if they want. They could easily present trailing information on a GAAP basis, too. From a technical standpoint, it would be a lot easier: the firm wouldn't have to be worried about its presentation rationale.

So - what stops companies from presenting information the way you propose? I suspect there are a couple reasons:

- There's nothing in it for them. Grinding out pro forma numbers will entertain analysts and perhaps result in a higher stock price. Presenting more information, as you suggest, will only aid analysis and probably increase the quality of investor and analyst questions and observations.

I don't know of too many managements screaming "Stop tossing me softballs!" to analysts...

- There's increased legal exposure. That's probably the most compelling reason for companies to spurn such presentations. The more information you hang out in public, companies figure, the more grist for the plaintiff's bar.

- If analysts can do it for themselves, why do it for them? Put it another way: if analysts and investors can put this information together on their own (after the conference call, of course) and presentation by the company would increase risks and bother for management, where's the incentive for management to provide the information?

What would make it change? The FASB is working on a financial performance reporting project that takes on many more issues than this, but they are considering some of the issues you mention - but not necessarily going in the direction you suggest. They've decided, in these early stages that "at a minimum, full sets of financial statements for two annual periods (the current and prior annual period) for all business entities. This would mean an entity would present three statements of financial position, two statements of earnings and comprehensive income, two statements of changes in equity, and two statements of cash flows." and "not to provide guidance regarding financial information voluntarily presented beyond the required minimum (that is, full sets of financial statements for two annual periods)." Bye-bye, three year annual earnings presentations.

As for the "trailing" interim earnings reports, they're being considered in a later segment of the project. No preliminary decisions on whether or not they belong in the financial reporting package. But keep your fingers crossed, keep an eye on the project, and let the FASB know your thoughts when they expose a document. And also bear in mind: the SEC would likely have to seriously amend Reg S-X to make FASB's changes stick for publicly-traded firms.

Last month, Dana Corporation reported problems with its 2004 annual report and its 2005 quarterly reporting through June, due to "issues involving customer pricing and transactions with suppliers in Dana's Commercial Vehicle business." About a week later, it pegged the cumulative possible restatement net income effects at $25 to $45 million; subsequently, Dana filed a notification of delay on third quarter 2005 reporting.

Yesterday, Dana filed a non-reliance 8-K covering the financial statements going as far back as 2000. It's not as nefarious as it sounds. How come? From the 8-K:

"The items requiring the restatement of the years prior to 2004 are unrelated to the company's ongoing internal investigation [mentioned above]. During 2004, Dana had recorded net charges totaling approximately $7 million after tax to adjust for items related to prior periods, including certain European benefit plans, state income taxes, inter-company balances, interest expense, and other accrued liabilities. The company had determined that these items did not materially impact the results of operations for the 2004 quarters in which they were recorded or for the full year. As a result of restating its 2004 financial statements for the items identified in the investigation, Dana will now be required to record the items comprising the $7 million in the periods to which they are attributable. Consequently, Dana will be restating financial results back to 2000. Items pertaining to periods prior to 2000 will be recorded in 2000 because they are immaterial to that year."

In short, there were $7 million of assorted charges belonging in prior periods that Dana charged to income in 2004, when its income from continuing operations was $95 million dollars. (Dana figured these were not material to 2004 income, which at first glance seems absurd. It's nearly 7% of income before the prior-period charges. However, they probably figured materiality based on continuing operations income exclusive of $180 million of the nearly-standard amalgam of loss on debt extinguishment, loss on sale of a business and the ever-popular restructuring charge. Add them all back, and the $7 million charge is less than 2.5% of the adjusted earnings.)

Once the 2004 figures are reduced by some $25 to $45 million, however, the $7 million of error correction becomes more material; in either case, the $7 million is much closer to 3% of the revised adjusted earnings. So, the $7 million gets taken out of 2004 and put into the years in which the errors really occurred. The individual years' errors might be higher or lower than the $7 million, obviously, and it looks like they must be material in 2001. The year 2001 was a loss year, making the materiality scope narrower; 2000 was a solidly profitable year ($334 million) making it more likely that recording the cumulative effect of the errors to that year would not make much difference in the figures.

A small cumulative figure can be a net of large positive and large negative figures - and each of those figures, when relating to different prior periods, can carry a lot more importance to each prior year when one of the figures has to be restated. It's a ripple effect, the old butterfly-in-Tokyo flapping its wings and making a storm in Kansas. The difference here is that Dana's restatement of 2004 figures is the butterfly, and the effect is rippling back in time to the year 2000. And that concludes today's lesson on how seemingly small cumulative figures today can be material to prior periods.

Back in May, I mentioned Colgate-Palmolive had two directors on its board that were involved in the New York State Attorney General's investigation of General Re. Mr. Ron Ferguson, audit committee chair and member of the finance committee, and Ms. Elizabeth Monrad, finance committee chair and audit committee member, were being questioned about the General Re transactions with AIG International.

(Same disclosure as before: I am a shareholder of Colgate-Palmolive and I have been for years. I own it, and it's in accounts that I manage for clients.)

In an 8-K filed on Monday, Colgate reported that on November 10, 2005, both resigned from the Board of Directors.