<<
>>

INFORMIX'S TROUBLES BEGIN

Informix was the best-performing United States software stock for the pe­riod 1990 through 1995.1 By the beginning of 1997, the company was widely regarded as a serious challenger to Oracle’s leadership in the busi­ness of creating large corporate databases.

Informix even taunted its rival on a billboard alongside the interstate highway near Oracle’s office towers, warning drivers that they were approaching a “dinosaur crossing.”

At the same time that it was trash-talking the competition, however, management was taking a huge risk in its operating strategy. Hoping to boost its market share, Informix shifted its marketing resources to Univer­sal Server, a new type of database intended to store different kinds of data, such as images and video. Informix’s strategic initiative proved premature. Not only did the new product remain unperfected, but corporate customers were not yet indicating demand for multimedia database technology.

On top of the product-related woes, some analysts were questioning In­formix’s revenue recognition practices. They grew wary when Informix re­vealed that about $170 million of its 1996 sales to computer makers and other resellers represented software that had not yet been resold to final customers. Roughly $55 million of that total involved agreements by com­puter manufacturers to purchase software equivalent in value to the hard­ware that Informix bought from them.

The Fine Points of Software Revenues

Chairman and Chief Executive Officer Phillip E. White declared that his company’s sales practices were properly disclosed and in compliance with the accounting rules. Several independent accounting experts backed up his statement.

None of this meant, however, that Informix’s revenue recognition policies were the only, or even the best, approach available under GAAP.

For software producers, shipping a product to a reseller does not neces­sarily represent a definitive sale. Just like “old-economy” manufacturers of basic industrial and consumer goods, software producers can expect run-of- the-mill returns of defective items. In addition, a reasonably predictable portion of customers will fail to pay their accounts in full. Deferring a por­tion of revenue for these contingencies is comparatively straightforward, but other recognition issues are more judgmental for a company such as In­formix. Uncertainty lingers over a sale until the software has been installed at the end user and the end user’s staff has been trained in its use. Further­more, resellers often have latitude to return products they cannot sell. At the beginning of 1997, with the accounting profession still developing stan­dards to address these complicating factors, software companies varied widely in their revenue recognition practices. Critics contended that In­formix’s approach was overly optimistic.

The Picture Worsens

As the year wore on, the company’s troubles mounted. On April Fools’ Day of 1997, Informix stunned investors by disclosing that its first-quarter sales had fallen short of analysts’ expectations by $100 million, or 40%. Two days later, the company reported a $140.1 million loss for the quarter, down from a $15.9 million profit in the comparable 1996 period. Sales were 34% below year-earlier levels, at $133.7 million versus $204.1 million. Between March 31 and April 3, Informix’s stock price plummeted by 41%.

Yet another shock arrived on May 1. Chief Financial Officer Alan Hen- ricks resigned after just three months in the position. To seasoned investors, such an abrupt departure by a senior manager represented a telltale sign of trouble.

They were not reassured when Henricks declined to explain his de­cision, even as former Informix executives reportedly said that he had pressed for more conservative accounting policies than CEO White fa­vored.2 White denied that policy differences caused his CFO to step down, instead citing unspecified personal considerations.

Before long, White himself was contributing to the front-office turmoil. In July, under pressure from the company’s board of directors he resigned first as CEO and then as chairman. Meanwhile, Informix failed to meet the deadline for filing its second-quarter financials, another classic red flag for investors. When the quarterly filing belatedly arrived on August 7, it re­vealed a loss of $120.5 million, including a $62.1 million charge for re­structuring costs. Informix’s operating loss of $80.5 million before the special charge was larger than analysts had expected.

The charge for restructuring costs showed that the financial reporting controversy was heating up, rather than simmering down. To correct earlier instances of improper revenue recognition, Informix announced that it would reduce its previously reported 1996 revenue of $939.3 million by $70 million to $100 million. According to new chief executive officer Robert Finocchio, the company had booked some transactions in the wrong periods and others without proper documentation. He added that Informix would recognize some of the revenues in future quarters, but the rest would disappear.3 The company’s investigation of past accounting practices had uncovered forty incidents of improper recognition, primarily in Japan, the United Kingdom, Germany, and Central and Eastern Europe, with a few found in the United States.

Still, the parade of bad news continued. On September 22, Informix’s already depressed share price tumbled 22% as the company announced that the downward restatement of 1997 revenues might be as great as $200 million, twice the previously announced high-end figure of $100 million.

Moreover, management revealed that the improper recognition practices extended back to 1995, resulting in a possible revenue reduction of as much as $50 million for that year. In certain instances, said the company, sales representatives had made informal side agreements to provide resellers extra funds to finance sales.4 Informix’s 1997 10-K elaborated:

The unauthorized and undisclosed agreements with resellers introduced acceptance contingencies, permitted resellers to return unsold licenses for refunds, extended payment terms, or committed the Company to as­sist resellers in reselling the licenses to end-users. Accordingly, license revenue from these transactions that was recorded at the time product was delivered to resellers should have instead been recorded at the time all conditions on the sale lapsed.5

Lessons for Analysts

In the end, the company restated its revenues downward by a total of $244 million for the period 1994 through the second quarter of 1997. Informix ac­knowledged that it had overstated its revenue by 4.0% in 1994, 12.9% in 1995, and 29.1% in 1996. This year-by-year escalation followed a frequently observed pattern of falsification. Bonus-seeking managers may initially veer off the straight-and-narrow by “borrowing” a small amount from future rev­enue, intending to “pay it back” the following year, but they instead fall fur­ther and further behind. Eventually, the gap between reported revenues and economic reality grows too large to sustain. Outside analysts’ skepticism mounts and closer auditing of the books unmasks the false reporting.

The Informix affair also teaches analysts not to construe auditor- certified compliance with GAAP as an assurance of integrity in financial re­porting. If senior managers set the tone by pushing the rules to their limits, executives a step or two lower in hierarchy may feel emboldened to break the rules altogether.

Liberal reporting differs from fraudulent reporting, in a legal sense, but a corporate culture that embraces the former can foster the latter.

In October 1997, the American Institute of Certified Public Accoun­tants reduced the wiggle room for companies such as Informix. The ac­counting body’s Statement of Position 97-2 (SOP 97-2), “Software Revenue Recognition,” superseded SOP 91-1 and at long last addressed the industry­specific complexities of determining when a sale was truly a sale. SOP 97-2 required companies to recognize revenue from a software arrangement by allocating a fair value to each element, such as software products, enhance­ments, upgrades, installation, training, and post-contract customer support. Determination of each element’s fair value was to be based on objective ev­idence specific to the software vendor. The AICPA subsequently deferred the effective date of SOP 97-2 until, with SOP 98-4 (“Modification of SOP 97-2, Software Revenue Recognition with Respect to Certain Transac­tions”) in February 1998, the accounting body was able to specify what constituted “objective evidence.”

As a consequence of adopting SOP 97-2 as of January 1998, Informix began deferring some license revenues that it formerly would have recog­nized on delivery of the software. That is to say, the company took consid­erable latitude in booking revenues until the accounting profession developed rules to deal with all the complexities of the software business. Investors should not have been lulled by the fact that Informix’s pre-1998 financial reporting satisfied the minimum requirements of GAAP. Manage­ment’s aggressiveness painted a rosier picture than it would have been al­lowed under standards better adapted to the intricacies of the software business.

Neither did the certification of Informix’s statements by independent auditor Ernst & Young prevent investors from receiving revenue figures that proved too high, in retrospect.

As the reliability of the company’s past fi­nancial statements became increasingly suspect during 1997, Informix felt obliged to hire a second accounting firm to advise its directors. Meanwhile, both Informix and Ernst & Young became targets of a class action suit by shareholders. The plaintiffs alleged that the company’s executives had over­stated revenues by claiming to have sold software that was merely shipped to resellers temporarily and then returned. On the strength of these faked sales, the shareholders claimed, the officials had boosted Informix’s share price to facilitate their personal sales of stock. In May 1999, Ernst & Young agreed to pay $34 million as its share of a $142 million settlement of the lit­igation. The auditor stated, as defendants often do under such circum­stances, that it settled to avoid costly litigation.6

By the time Ernst & Young accepted the financial consequences of its role in the matter, Informix had replaced the firm with a new auditor. In dismiss­ing Ernst & Young, Informix’s board did not criticize the firm for signing off on 1994-1997 revenue figures that subsequently turned out to be overstated. The board did, however, mention a dispute over recognition of revenue from industrial manufacturers in the first quarter of 1998. The disagreement, in short, arose well after the earlier improper booking of sales came to light and new management controls were implemented. According to Informix, the company resolved the dispute to Ernst & Young’s satisfaction, deferred ap­proximately $6.2 million of revenue, and underwent yet another restatement of its financials. Ernst & Young responded that Informix’s statement was in­complete. Shortly before its dismissal, said the auditing firm, it had informed the board’s audit committee that a lack of necessary resources in Informix’s financial reporting departments had created difficulty in accumulating the accurate information required for timely disclosure. This condition was a ma­terial weakness, in Ernst & Young’s view.7

To an outside analyst, it appeared that Ernst & Young ran into conflict with Informix for enforcing strict standards in 1998 after being sued by shareholders for alleged laxity in earlier years. Dismissal of Informix’s audi­tor was one more classic warning sign for analysts, on top of the senior management upheavals and delayed financial statement filing of 1997. Events throughout 1997 and 1998 reinforced an essential point about rev­enue recognition: Even when an independent accounting firm certifies that a company’s financials have been prepared in accordance with generally ac­cepted accounting principles, the analyst must stay alert for evidence that the numbers misrepresent the economic reality.

<< | >>
Source: Fridson M., Alvarez F.. Financial Statement Analysis. John Wiley & Sons, Inc.,2002. — 413 p. 2002
More financial literature on Economics.Studio

More on the topic INFORMIX'S TROUBLES BEGIN: