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.

Now that my coot attack has subsided, let me tell you about one of the more interesting exchanges during the conference. It encapsulated a lot of the "who's at fault discussions" going on. And it brought out an interesting angle on auditing that perhaps should be pursued by the Big Four auditors.

Moderator Katherine Schipper asked PCAOB board member Daniel Goelzer if the PCAOB will take auditors to task in their reviews of registered auditing firms, for not noticing backdated options in the first place. Goelzer's answer: there are understandable reasons for missing the misdating. Most of these firms were using APB Opinion No. 25 accounting for their option plans with a resulting compensation balance of zero being reported. Why would they spend much time and effort auditing such an amount if they didn't suspect malfeasance in the first place?

(Not discussed: during the periods in question in the backdating scandals, the auditing business was a much different animal than it is now. Post-Sarb-Ox, there's a lot more focus on documentation, and no "cross-selling" efforts between auditors and the consulting side. Pre-Sarb-Ox, it was a lot wilder and woolier. When we look at the issue now, I'm not always sure observers remember that today's auditors are much more sober bunch than what existed in the heyday of backdating.)

Ms. Schipper then asked if the auditors should have caught the misdating of their audits of footnoted Statement 123 information. Theoretically, auditing that footnote information vigorously could have uncovered contracts with incorrect dates - and this has been a fixture of financial statements since 1996. The whole thing could have been avoided. Goelzer's reply was that an audit is an document-driven process, and most auditors might have had no reason to believe that the documents they were examining were false at the time. Another member of the panel, Peter Klinger of BDO Seidman, chimed in that auditors don't usually devote the same amount of attention to footnotes as they do to financial statement amounts. (I think Goelzer agreed with him.) That's not a comforting thought to investors who often rely on the footnotes as much, if not more than, the actual financial statement amounts.

Professor David Larcker, one of the conference organizers, reminded the panelists that there was a body of academic literature existing as far back as 1997 showing that opportunistic pricing of options had been found. (Surprise: Erik Lie was not the first academic to discover it. Take a look at "Good Timing: CEO Stock Option Awards and Company News Announcements" in the Journal of Finance, Vol. 50 No. 2, June 1997. Available on the SSRN Network.) Why didn't the auditing profession take note and sharpen their focus on option awards?

Good question, and no really clear answer emerged from the discussion. My own interpretation: we're looking at things with perfect hindsight, and everything looks so obvious when you look at the past. For instance, some of the graphs shown at the conference were amazing: it was a plot of cumulative stock returns before and after the grant date of options for a sample of 7,786 grants from 1,970 companies from 1996-2005. Wish I could show you the graph here, but let me just say that it looked a lot like a "V": the returns sloped down just before the award date, then sharply curved upward. That's not a small sample either: the sample covered approximately 89% of the U.S. market capitalization. Shouldn't it be obvious to auditors?

Well - no. We're looking at the forest now, with perfect hindsight. And auditors, if they were looking at all in the late 1990's were looking at trees, and probably not too closely. You almost have to ask why everyone didn't notice - investors and regulators included.

That's not an excuse for auditors: Larcker's point is a good one. Maybe auditing firms do look at the academic literature for an auditing advantage already - but if they do, they seem to have missed a pretty good way to have sharpened up their auditing processes. After all, as the investigations now show, the SEC uses academic literature for scoping things out. (Lesser-known example: Professor Carol Marquardt of Baruch College brought regulatory attention to EPS management through her work on the infamous "CoCo" bonds.) Maybe the Big Four's national office think tanks will recruit more heavily from the ranks of academe in the future.

Spent the day yesterday at one of the best train stations in the country: Union Station, Washington DC. Why? The Arthur and Toni Rembe Rock Center for Corporate Governance, a joint effort of Stanford Law School, Stanford's Graduate School of Business and the Engineering School, held a conference entitled "Lucky Strikes: Public Policy Issues in Backdating and Springloading." Catchy, eh?

The conference featured speeches by SEC chairman Christopher Cox and enforcement director Linda Chatman Thomsen. No surprises in either speech, really. (Although it was the first time that I'd heard anyone trace the rise of options issuance back to the employee stock option plans of the 1980's when they were part of the anti-takeover landscape. That was an assertion of Ms. Thomsen.) More than just speeches by SEC staff, the conference was a feisty (well, to an accountant) blend of commentary and discussion from academics, members of the law profession, and other regulators like the PCAOB's Daniel Goelzer.

The conference covered more than I can cover in a single posting. After hearing about the problems that can be caused by even automatic grant plans and transactions in options during blackout periods (if the managers know that there's material information afoot, shouldn't they halt the auto-pilot transactions? Or depend on the auto-pilot as a defense?) and the varying length of blackout periods (there's little uniformity from company to company, and it's unclear why it should vary) and prepaid variable forwards (contracts that let executives get upfront cash from their option/stock holdings by pledging them for delivery in the future - without reporting a sale in the exec comp reporting regime), I'm more convinced than ever that most option plans inherently work against the long-term equity investor and no amount of regulation will change that.

The Princeton Review, sponsor of the leading SAT preparation course in the United States, needs a little refresher course in Statement 133. (And it has lots of company. Statement 133 has been a sore spot for companies over the last twelve months, generating many restated financial statements.)

The company filed a non-reliance 8-K on Friday, due to "embedded derivatives" in an issue of preferred stock that need to be de-embedded and accounted for as stand-alone derivatives. Why? That's because Statement 133 requires contracts for financial instruments to be evaluated by the issuer to see if they contain derivatives that wind up getting historical cost treatment just because they're plopped into just such a contract. Put it this way: a company might enter an onerous derivative contract that would have to be accounted for at fair value, with the result that investors would see changes in that contract each quarter. To avoid that, the terms of the derivative might be embedded in another contract - like preferred stock or convertible debt - that doesn't get re-measured at fair value. Statement 133 prevents such end runs by requiring separation of such an embedded derivative from a host contract when all three of these conditions are met:

The economic characteristics and risks of the embedded derivative instrument are not clearly and closely related to the economic characteristics and risks of the host contract.

The contract containing both the embedded derivative instrument and the host contract is carried at historical cost unless GAAP requires it to be carried at fair value with changes in fair value reported in earnings.

If the embedded derivative instrument was a separate instrument in the first place, it would be considered a derivative instrument under Statement 133.

The company issued convertible preferred stock in June 2004 - without the proper separation of embedded derivatives. It also issued warrants which should have been classified as liabilities, instead of equity. The firm hasn't finalized its figures yet, but from the ranges given in the 8-K, it looks like reclassification will slightly improve the net loss in 2004, increase it in 2005, and improve it again slightly in 2006. The balance sheet is more leveraged in 2004, less leveraged in 2005, and unchanged in 2006.

It's already a fair value reporting world, and only likely to become more so with the issuance of Statement 157, "Fair Value Measurements," which will pave the way for more fair value reporting standards. Statement 133 was one of the more broadly-sweeping fair value standards issue to date. Before new ones get issued, the SEC might be doing some broad sweeping of its own to make sure that current GAAP is being followed before new fair value GAAP arrives.

Interesting 6-K filing by Japanese electronics giant NEC: it seems that the company is having trouble preparing its financial results on a United States accounting principles basis. This isn't something new challenge for NEC, so it's a little unusual to see it happening now.

The company will be reporting results on a Japanese basis for the current year; they gave no indication that they'd be switching back to U.S. GAAP. NEC's financials for the fiscal year ending March 31, 2006 were audited by Ernst & Young ShinNihon under Japanese auditing standards - not the same standards as required in the United States by the Public Company Accounting Oversight Board. After the report was filed with the Japanese regulators, the auditors "requested further analysis" for a revenue recognition issue; this caused a delay in filing the firm's 20-F with the SEC, and was announced on September 28. It still will file (eventually) the 2006 20-F on a U.S. GAAP basis; although they make no commitment to further filings, it would be strange not to continue reporting on a U.S. basis once the bugs have been worked out.

The company admits that it will take "considerable time" to prepare its financials on a U.S. basis for the first half of fiscal 2007 (the current year) because of the problems in getting resolution on 2006. So, it's switching to Japanese GAAP in its future financial releases. Needless to say, this is going to hose the analysts who have been following the company on a U.S. basis - and don't know reporting under Japanese GAAP.

What's the hold-up? Being a foreign issuer, NEC doesn't report on its internal controls yet. But the auditors have found problems neverheless: The September 28 filing states that the current internal controls do not "ensure that all significant U.S. GAAP adjustments and reclassifications, presentations, and disclosures" are included in U.S. GAAP financials; the company lacks specific policies and procedures for dealing with transactions under U.S. GAAP; and the company lacks personnel with the "appropriate knowledge, experience, and training in the application of U.S. GAAP at its headquarters' corporate controller division or its various business units and subsidiaries."

Damning stuff, and not uncommon in the United States when firms were going through their first Section 404 reviews. The net result of those weaknesses led to the delay in filing: problems were found in revenue recognition, income taxes, accruals and reserves, and research and development expenses. Basically, in much of the income statement.

The real sticky issue, as noted in yesterday's filing seems to be the revenue recognition issues in NEC's "IT Solutions" group. Apparently, this is a services group that enters into customer arrangements that have more than one facet to them: the firms is committed to delivering multiple elements of a contract over various stages of an agreement's life. The iron rule of revenue recognition (one of the iron rules, anyway) is that you don't recognize revenue before you've delivered goods or provide services - even if you've been paid in advance. NEC's auditors "requested further analysis to support the relative fair value of maintenance and support services provided as part of multiple element contracts": it's what's needed to parse out the different components of a contract, each component with a possibly different revenue recognition pattern. So the relative fair values of each can ultimately have a big impact on the timing of revenue recognition. Given that these contracts often span more than one year, it's not unlikely that there could be issues with previous years' documentation that have an effect on FY 2006 also. That would only add to the time needed to finish the job.

F5 Networks filed a non-reliance 8-K on Wednesday; MIPS Technologies also filed a non-reliance 8-K that day. I'll cut to the chase: both companies announced that they'll be restating back as far as 1999 when they're finished their investigations of options backdating.

That's on top of a few other recent disclosures (Monster, Integrated Silicon, Valeant Pharma) that their revisions will be presented on a restated basis rather than a catch-up basis. F5 and MIPS won't be restating as far back as the others but back to 1999 is still a long time.

It's safe to say a pattern is emerging: if you buy off on these plans, restatements will be done on a fully restated basis - something that would be much more useful for investors trying to assess how far off base managements led them during the "anything goes" years. (At least for this week, there's a pattern. More informal monitoring to come.)

Old news now, but worth a mention anyway: Ford announced on Monday that it would be restating its financials from 2003 to 2005 in an amended 2005 10-K (and the "selected financial data" therein, from 2001 to 2005) for incorrect accounting related to derivatives. Per the 8-K filing:

"During the preparation of its response to a comment letter from the Division of Corporation Finance of the Securities and Exchange Commission a routine review of its Annual Report on Form 10-K for the year ended December 31, 2005, our indirect wholly-owned subsidiary, Ford Motor Credit Company ("Ford Credit"), became aware of a matter related to accounting for interest rate swaps under Statement of Financial Accounting Standards No. 133, Accounting for Derivative Instruments and Hedging Activities, as amended ("SFAS 133"). Specifically, Ford Credit discovered that certain interest rate swaps it had entered into to hedge the interest rate risk inherent in certain long-term fixed rate debt were accounted for incorrectly because they did not satisfy the technical accounting rules under SFAS 133 to qualify for exemption from the more strict effectiveness testing requirements. PricewaterhouseCoopers LLP, our independent registered public accounting firm, audited our 2001 through 2005 financial statements, which included a review of these swaps.

These interest rate swaps were entered into as part of Ford Credit's asset-liability management strategy. As noted above, the swaps economically hedge the interest rate risk associated with long-term debt issuances, and we continue to believe that these swaps have been and will continue to be highly effective economic hedges. The correction to the accounting does not impact the economics of the hedges, nor does it affect cash."

A few things: first, the fact that Ford had a "routine review" by the SEC infers nothing special. Companies of Ford's size are frequently reviewed by the SEC as a matter of course. Second, while they "became aware of a matter" in responding to the SEC's comment letter on the 2005 financials (filed March 1, so it had to occur sometime afterwards), it's hard to believe they couldn't have been aware of the matter beforehand - companies had been filing restatements for derivatives do-overs as far back as May 2005, like General Electric did, and there was plethora of such restatements in the late fall of 2005. Third, it sounds like the company tried for shortcut effectiveness testing without meeting the "technical requirements" - which are pretty prescriptive, and they're still requirements, whether derided as "technical" or not. Because they went the shortcut route, they couldn't have had contemporaneous documentation to justify the hedge effectiveness once denied the shortcut treatment. Fourth, it sounds like they're saying "it was our auditors fault." Actually, it sounds like both parties missed it.

Is there another wave of these restatements coming? Doubtful, even though Farmer Mac went through the process a couple weeks ago. Being a smallish company, Farmer Mac's controls - including a review of such details - might not yet be up to the same speed as the others making the restatements. And Ford's finance and accounting staff, coping with $5.8 billion losses, succession changes at the top, bond rating changes and market share changes, just might have put changing its derivatives accounting near the bottom of its "to do today" list.

There's a long restatement period facing Integrated Silicon Solution, Inc. and Valeant Pharmaceutical International: as far back as 1997 financial statements.

Today, Monster Worldwide joins the ranks of long-term restaters. In their 8-K filing of current financial statements, they also provide an update on their option investigation:

"The Company is still conducting its accounting analysis and has not yet determined definitively the impact of these differences on the Company's historical financial statements. However, the Company expects that it will restate its previously filed financial statements for the years 1997 through 2005... the Company's board of directors has concluded that the Company's previously issued financial statements and other historical financial information and related disclosures relating to periods through December 31, 2005 contained in the Company's filings with the Securities and Exchange Commission (the “SEC”), including applicable reports of its independent registered public accounting firm and press releases, should not be relied upon."

It's way too soon to call this a trend. The approach makes sense though, from the standpoint of giving shareholders information about the economics of pay within these firms during those long-past years. While there may be nothing material in 2006, as Monster notes, the roots of these problems might have an effect on years to which 2006 might be compared. When the facts are all displayed for shareholders, the revised history of all these firms is going to be interesting.

If a movie is ever made about the options backdating scandal (not likely), it might open with a crawling text message on the screen that reads like this:

"A long time ago in a bubble far, far away..."

The options investigations are stretching way, way back in time - and they'll produce some restatements that go way, way back too, if the differences between right and wrong accounting are material enough in prior years. Yesterday, Integrated Silicon Solution, Inc. filed a non-reliance 8-K warning investors that its special committee investigating options grants had found that "... the actual measurement dates for financial accounting purposes of certain stock options granted primarily during fiscal years 1997-2005 differ from the recorded grant dates of such awards. Consequently, new measurement dates for financial accounting purposes will apply to the affected awards, which will result in additional and material non-cash stock-based compensation expense... ISSI's Board of Directors, with the concurrence of the independent committee, determined that ISSI should restate its financial statements for various periods since ISSI's initial public offering in February 1995."

No specifics were mentioned on the years to be restated, but the tone of the statement makes it sound like 1997 through 2005 could be fair game. Though the company came public in 1995, even 1997 is a long way back for restatement.

Different firm, different industry, same message: another recent non-reliance filer is Valeant Pharmaceuticals International. In its 8-K it disclosed quite the same issue: its special committee had determined that options granted between 1997 and "subsequent years" had incorrect measurement dates and correcting them will cause financial statements to be restated for certain, yet-to-be-determined periods.

As these investigations get worked out, there will likely be many auditor-client controversies over what constitutes a big enough difference in past history to require restatement. Firms would prefer to see as few restated back periods as possible, and play catch-up with their corrections whenever they can justify it: the less information given about the past, the fewer questions that can be asked by investors. (And trial lawyers.) At the other end of the information spectrum, investors should prefer more details about the past. After all, the accounting that was violated (APB 25) was the accounting of choice by these firms and it should have presented a picture of the compensation cost to shareholders. (It was an incomplete picture: SFAS 123 accounting would have been better.) APB 25 was, however, what companies had agreed to abide by, and it should have showed compensation. As it turned out, companies didn't abide by it, and compensation went unnoticed. Investors should care.

On October 12, the SEC reached a settlement with William Beecher, the former CFO of i2 Technologies for his participation in scheme to fraudulently report revenues and earnings over a five-year period while taking part in massive insider trading while the misleading reporting occurred. Amazingly, he got off with only a cease-and-desist order and a $1.3 million settlement, and is barred from serving as an officer at a public company for five years. Beecher must have had one sharp attorney; what he and his compadres, CEO Gregory Brady and President Reagan Lancaster, did at i2 was pretty amazing by today's standards.

Maybe it wasn't so amazing by standards of the late 1990's, which is to say, no standards. According to the SEC's complaint in the matter, the trio engaged in revenue inflation from 1997 to 2002 by recognizing revenue for deliveries of software that they knew did not work, and approving "side agreements" with customers that indicated they knew the delivered software did not work. The company delivered faulty software to the likes of Nike, Kmart, Best Buy, Corporate Express and Procter & Gamble. In a strange brush with history, the firm even engaged in barter transactions for non-existent work done for Enron Broadband Services, working with Jeffrey Skilling (whose sentencing is to be announced today.) The arrangement was essentially a "you scratch my back, I'll scratch yours" deal: i2 needed revenues, so Enron Broadband Services helped them in return for help they could receive later in making their own revenues. At one point, there was even discussion of adding Ken Lay to i2's board of directors.

The trio clearly understood the accounting principles involved, as evidenced by their e-mails. Also evidenced in their e-mails was the fact that they knew they were selling useless products: in 2000, Lancaster had e-mailed Brady "about i2 building "bullshit demos" for prospective customers that showed all functions working together, which was not reality." He went on to state that "I am not pointing fingers but I will tell you that you can only sell vapor for so long and then it catches you. Right now we have vapor in CM [Customer Management, an i2 software product], Procurement, Marketplace deals, etc. .. . We are selling our stuff with a good pitch but there is no substance behind anything ..."

At one point, the company brought in Michael Cusumano, an MIT professor, "to analyze the firm's structure and processes," according to the SEC's complaint. A memorable conclusion in his findings: "Two core competencies in i2: (1) Can sell anything to anyone. And (2) delivery guys can make any piece of crap work, given enough time."

When all was said and done in the company's restatement in 2003, the revenues were overstated in 2000 by 41%; in 2001 by 14%; and understated in the first half of 2002 by 27%. For the entire period, the net earnings effect was a negative $207.1 million. During 2000 and 2001, Beecher exercised options and sold $19.9 million of stock; Brady did the same and reaped $92.1; and Lancaster cashed out $27.9 million. At the same time, CFO Beecher and CEO Brady signed off on financial statements and representation letters to auditors indicating that the financials were properly prepared.

The story has a couple of lessons in it. Lesson one: notice that this case was just settled with one of the principal players in 2006; it was the restatement in 2003, after an audit committee investigation beginning in 2002, that got the matters to the forefront. Three years later, closure arrives. So it will probably be with backdating investigations - total closure on them will not likely occur any time soon. It'll be years in the making; until then, you'll have to live with the backdating twists, turns and surprises.

Second lesson: this was a blatant case of skunky revenue recognition. You just don't hear that many instances of it these days, and it was the gimmick of choice in the late 1990's, early 2000's era. There seems to be fewer 8-K non-reliance filings on revenue recognition issues than in past years. Could it be that the auditing profession has been strengthened by the Sarbanes-Oxley reforms? That's a benefit inuring to investors that they might take for granted amid all the gabble about rolling back the act.

In the past year, many pundits have expected wholesale terminations of defined benefit pension plans. IBM got the ball rolling in January when they announced their plans to freeze defined benefit pension plans beginning in 2008. It's a move that earned IBM plenty of criticism and perhaps inspired more firms to look at the possibility of triggering their own pension Ice Age.

So far this year, there have been a number of new freezes - but they haven't reached avalanche status yet. Check out this list of pension freezes noted in 2006 filings so far: there have been 40 announcements captured among SEC-registered firms. The majority of the firms - 21 of them - are fairly small, with market caps of less than $1 billion. (Two aren't publicly-traded: Remington Arms and Harry & David Operations.) In fact, only 9 firms have market caps over $5 billion: IBM, Hewlett-Packard, DuPont, Alcoa, Alltel, General Motors, Hershey, NCR, Lexmark, and Northeast Utilities. That makes for less drama in the press: if smaller companies are freezing plans, there are less people affected per company - meaning headlines are less dramatic.

There were 20 freezes announced in the first half of 2006 (that we could find). There are another 20 in the second half, with two-and-a-half months to go - so the snowball is picking up some momentum.

While the idea of a pension freeze sounds cruel - i.e., heartless managers thrusting faithful employees out onto the frigid tundra - shareholders should consider it to be a sign of effective managers. Defined benefit pensions create cash obligations that are uncertain; they put the firm in the role of guarantor. Defined contribution plans can provide similar benefits to employees, make them responsible for their own investments, and most importantly for managers, make a volatile cost a controllable one. Notice from the list in the file that 29 of the 40 firms are increasing their 401K contributions or adding other employee benefits at the same time they're freezing defined benefit plans. It's uncertain from the filings that the other 11 aren't going to do the same thing. The point is, managers aren't necessarily hosing their employees; they're trading one kind of retirement cost for another that's more controllable.

Big firm, little firm - they've all got to manage costs. Defined benefit pension plans create costs that are hard to control, so it's no surprise that firms are freezing no matter their size. The snowball might start rolling down the hill faster once firms have had longer to digest their burdens under the new pension act - and after they see the effects of the new FASB rules on their balance sheets.