CALLING THE SIGNALS
Aspiring analysts should extract at least one invaluable lesson from the study of Informix’s aggressive revenue recognition: Staying alert to evidence of flawed, or possibly fraudulent, reporting is essential, even when the auditors put their blessing on the numbers.
Exactly what sort of evidence should one look for, however?The specific answer is not the same in every case. As a rule, though, distorting one section of the financial statements throws the numbers out of whack in some other section. Assiduous tracking of financial ratios should raise serious questions about a company’s reporting, at a minimum. To illustrate this point, let us consider another example from the computer software industry.
KnowledgeWare’s chief executive officer was Pro Football Hall of Fame quarterback Fran Tarkenton. After retiring from the National Football League in 1978, the one-time record-holder in touchdowns and passing yards founded Tarkenton Software, which he merged with KnowledgeWare in 1986. Specializing in client-server development tools, a type of software used in creating programs for networked personal computers, the company came to be regarded as an emerging star in its industry. In the early 1990s, however, KnowledgeWare fell behind its competitors in adopting new tech- nology8 and overextended itself through a series of acquisitions.9 As the company’s new products failed to take off, KnowledgeWare began reporting losses and in July 1994 laid off a quarter of its workforce. With operations reporting negative cash flow, KnowledgeWare “appeared to be close to being out of money,” according to an industry analyst.10
At that point, Tarkenton, who was renowned as a scrambler during his professional football days, managed to get KnowledgeWare out of an extremely tight spot.
On August 1, 1994, the company announced an agreement to be acquired by Sterling Software. Sterling, a leader in electronic commerce and systems software, agreed to pay 0.2893 shares of its common stock for each KnowledgeWare share, producing a purchase price of approximately $143 million.The rebound in KnowledgeWare’s fortunes proved short-lived, however. Within a month of its pact with Sterling, the company announced a $19.0 million loss for the fiscal year ending June 30, 1994. In addition to recording $15.4 million of red ink in the fourth quarter, KnowledgeWare restated its earnings for the first nine months. The previously reported $4.46 million profit turned into a $3.6 million loss and the company’s stock price plunged by 45% in one day.
KnowledgeWare’s new travails arose from a decision implemented during fiscal 1994. In an effort to boost revenues, the company began supplementing its own sales force’s efforts with agreements to market its products through resellers. Almost immediately, the company started to encounter difficulties in collecting receivables generated by the resellers. In restating the results, management said that it had booked as revenues certain sales for which collections were not made.
The revelation of financial reporting problems initially threatened to scuttle the acquisition by Sterling Software.11 In the end, though, Sterling agreed to go ahead on revised terms. Instead of paying 0.2893 of its own shares for each of KnowledgeWare’s, Sterling slashed its offer to 0.1653 shares. Furthermore, Sterling announced that it would acquire only 80% of KnowledgeWare’s stock for the time being. The other 20% of the purchase price would go into escrow as a contingency for possible lawsuits by KnowledgeWare shareholders.
Fran Tarkenton succeeded in obtaining approval of the reduced terms, but the shareholders’ meeting was testy.
Sterling’s concerns about possible lawsuits had proven well founded, with the plaintiffs alleging that management had inflated KnowledgeWare’s stock price by misrepresenting the company’s financial condition. An attorney representing the disgruntled shareholders attempted to query Tarkenton about the merits of the Sterling offer, but the wily ex-quarterback refused to take questions.12 His stiff-arm tactics worked. The favorable shareholder vote eliminated the final obstacle to the merger. By the terms of the agreement, Tarkenton gained a seat on Sterling’s board and a three-year consulting agreement at $300,000 per annum.Despite all his nimbleness, Tarkenton failed to emerge from the affair entirely unscathed. In 1999, the Securities and Exchange Commission charged him with directing a plan to overstate KnowledgeWare’s revenues and profits during 1993 and 1994. According to the allegations, the company recorded sales in instances in which distributors were told that they did not have to pay for software unless they succeeded in reselling it. The SEC claimed that the objective was to convince investors that KnowledgeWare was making a comeback from its fiscal 1992 operating loss.13
Tarkenton settled the SEC charges by paying a $100,000 civil penalty and disgorging an amount equivalent to the bonus he received as a consequence of the overstated profits—$54,187, plus interest. His attorney hastened to say that his client had neither admitted nor denied the charges and was “pleased to have this matter behind him.” Continuing in a vein familiar to aficionados of financial reporting scandals, the lawyer added, “The events in question took place about five or six years ago during Mr. Tarkenton’s tenure at KnowledgeWare, and he has long since moved on with his life and to other business ventures.”14
To be fair, KnowledgeWare amended its financial reporting practices once the receivables problem came to light.
The company altered its method of recognizing revenue to bring reported results into line with the collection experience in its reseller program. Under the new, more conservative policy, KnowledgeWare recognized software license revenue from resellers only on receipt of payment. Reasonably enough, the company continued to book sales on shipment when it sent the goods directly to end users, as long as specific credit information was available to form a reasonable basis for estimating the collectibility of the receivables.Fixing the accounting problem after the fact, however, did not help investors who valued KnowledgeWare on the basis of its originally reported profits. Analysts had to rely on their own devices to identify the potential for an earnings restatement before the disclosure devastated KnowledgeWare’s stock price. At least a few analysts, it appears, did just that. Around the time that the Sterling Software deal was announced, some analysts were reportedly worried that liberal revenue recognition would lead to a write- off.15 They could have drawn that inference by careful scrutiny of standard financial ratios.
Exhibit 6.1 shows that revenues and receivables began moving out of sync from the moment KnowledgeWare launched its reseller program. In the first quarter of fiscal 1994 (ending September 30, 1993), revenues dropped by 15.5% from the preceding quarter, yet receivables rose by
EXHIBIT 6.1 Revenues and Accounts Receivable
KnowledgeWare, Inc. ($000 omitted)
|
| Quarter Ending | |||
| 1993 | 1994 | |||
| June 30 | September 30 | December 31 | March 31 | |
| Revenues | $40,426 | $34,144 | $38,178 | $38,928 |
| Accounts receivable | 36,894 | 37,084 | 41,506 | 50,252 |
| Days sales outstanding* | 83 | 98 | 100 | 117 |
size=1 color=black face=Cambria>*As reported by company.
Source: KnowledgeWare Forms 10Q and 10K.
0.5%. Management’s discussion and analysis in the Form 10-Q for the period noted that in terms of days sales outstanding, receivables leaped from 83 to 98 days.
The adverse trend continued, with DSO growing to 100 days in the second fiscal quarter, then soaring to 117 in the third, when receivables increased by 21.1% on nearly flat revenues. By the end of the third fiscal quarter (March 31, 1994), accounts receivable stood roughly one-third higher than at the preceding fiscal year-end, even though revenues had fallen slightly.To seasoned financial analysts, the diverging trends of revenues and receivables looked worrisome even in fiscal 1994’s first quarter. Management’s discussion of the matter, however, was as brief as its analysis was thin. The “Liquidity and Capital Resources” section of the MD&A for September 30, 1993, elaborated on the data only to offer the encouraging news that the percentage of gross trade accounts receivable subject to payment terms beyond 90 days had fallen from 25.0% to 21.7% during the quarter. The change implied that KnowledgeWare would be able to convert assets to cash more swiftly than formerly, thereby bolstering its liquidity. On the face of it, the quality of KnowledgeWare’s receivables was improving, possibly offsetting concerns about their increasing quantity. In the quarter ending December 31, 1993, management proudly reported a further drop in the portion of receivables dated beyond ninety days, to 10.4%.
During the quarter ending March 31, 1994, according to KnowledgeWare’s Form 10-Q, the percentage declined once again, to 9.4%. Curiously, though, the company now indicated that the beyond-90-days ratio on June 30, 1993 had been 17.8%, rather than 25.0%, as previously reported. Management did not explain the discrepancy or even draw attention to it. Analysts had to discover the restated number by comparing the latest 10-Q with its predecessors.
The March 31, 1994, report also contained several entirely new disclosures:
■ On top of the 9.4% of receivables subject to payment terms beyond 90 days, another 16% were subject to government funding provisions and would likely take more than 90 days to collect.
Federal and state government agencies, which are generally slow to pay, represented a growing portion of KnowledgeWare’s business.■ Notwithstanding the relative decline in receivables dated beyond 90 days, the number of transactions with “extended payment terms” (not sp ecifically defined) was growing.
■ Resellers in the nongovernment market accounted for approximately 6.2% of revenues for the three months ending March 31, 1994. Many of the resellers, said KnowledgeWare, were not well capitalized and therefore represented greater credit risk than the company’s typical end-user customers. KnowledgeWare compounded the inherent credit risk of dealing with resellers by offering them payment terms in excess of 90 days. Management hoped that the generous terms would induce resellers to initiate, increase, or accelerate revenues.
■ On March 31, 1994, KnowledgeWare had $3.1 million of accounts receivable (6.1% of the net total) that were past their due date by 90 days or more.
In summary, the receivables-related bombshell that KnowledgeWare dropped on August 30, 1994 was telegraphed by a simple comparison of revenues and receivables in KnowledgeWare’s 10Q for the first fiscal quarter, received by the SEC way back on November 2, 1993. The company blunted the impact of that danger signal by stressing the declining ratio of receivables subject to payment terms beyond 90 days. Days sales outstanding continued to escalate over the following two quarters, however. In the fiscal third-quarter report, KnowledgeWare suddenly began disclosing additional, less upbeat, details about its receivables. Battle-scarred analysts of financial statements would have guessed that management’s increased candor was the product of external prodding. To those alert enough to notice the unexplained change in the previously reported receivables-beyond-90-days ratio for the end of fiscal 1993, the sense of impending disaster should have become inescapable. By the middle of May 1994, more than three months before KnowledgeWare’s restatement, the company’s reported revenues were highly suspect. Not even the great Fran Tarkenton, in his glory days in the National Football League, benefited from so many tip-offs of his opponents’ moves.
An Income versus Cash Disparity
The financial statements of a computer manufacturer likewise telegraphed future problems in the area of revenue recognition. Shortly before Kendall Square Research’s October 1993 revision of its previously reported earnings, a research service known as Financial Statement Alert warned that the company was recognizing revenues too early.
Kendall Square reported $45.4 million in revenue in the first six quarters after it went public in March 1992. Loren Kellogg, copublisher of Financial Statement Alert, compared this income statement information with a figure from the company’s statement of cash flows. Over the same 18month period, Kendall Square’s “cash received from customers” was just $25.7 million. Kellogg viewed the $19.7 million disparity between the two numbers as evidence that a large proportion of sales being booked by Kendall were dubious.
The warning proved prescient. Less than a month after Kellogg’s analysis was reported in The Wall Street Journal, Kendall Square disclosed that its third-quarter 1993 revenues would be “substantially below” securities analysts’ expectations. In lieu of earnings per share of 11 cents (the consensus forecast according to the forecast-tracking firm of Zacks Investment Research), the company said that it would report a loss. Additionally, Kendall Square delayed the release of its third-quarter earnings and announced the resignation of its senior vice president/treasurer, who had joined the company only a month earlier. All these developments, by the way, were classic indications of serious corporate problems.
Revenue recognition controversies were central to Kendall Square’s difficulties. The company indicated that although third-quarter shipments were “generally in line with expectations,” there was some question about the proper amount of revenue to recognize from the shipments. Jeffry Canin, an analyst at Salomon Brothers, speculated about a possible area of disagreement within the company. Some officials, he suggested, may have objected to counting as revenue rebates that might have been given to customers who agreed to upgrade to Kendall’s next generation of computers. Smith Barney Shearson analyst Barry Bosak proposed the possibility that Kendall Square had been hurt by its reliance on sales to universities. A number of these institutions, which were in turn dependent on diminishing government funding, proved unable to pay. Indeed, some critics insinuated that Kendall Square had made research grants to educational institutions as quid pro quos for orders, a charge that management denied.
At any rate, Kendall Square’s troubles continued, as auditor Price Waterhouse withdrew its clean opinion from the company’s 1992 financial statements. Management revealed that the year’s sales figure, originally reported as $20.5 million, included $4.2 million of “improperly recognized” revenue. Unaudited numbers for the first half of 1993 would also require restatement, the company added.
In the wake of these announcements, Kendall Square demoted and then fired its president, its chief financial officer, and the head of its technical products group. The company’s acting chief executive officer announced that henceforth, Kendall Square would concentrate on building computers to order, instead of creating inventories in anticipation of orders. That reform was likely to reduce problems associated with revenue recognition, but by the time it was introduced, the damage to users of financial statements was substantial. At 7½, the company’s stock price was down by about 70% from its peak three months earlier.16
More on the topic CALLING THE SIGNALS:
- Signals, Transparency and Forward Guidance
- Searching for the Right Signals in a Market Economy to Prevent Crises and to Manage Them
- TCR SIGNALS THAT STIMULATE ANTI- AND Proapoptotic pathways
- Confusing Signals from Above
- NEF SIGNALS THAT STIMULATE ANTI- AND PROAPOPTOTIC PATHWAYS
- In this chapter we first discuss some policy solutions necessary to ensure proper working of a capitalist system.
- HMGB1 AS A MAJOR “ALARMIN”
- Different Conceptions of Public Service
- Conclusion
- FROM DOGMATIC ISLAMOPHILIA TO INTELLECTUAL AND MORAL CONFUSION
- DEATH ASSOCIATED WITH GP120 BINDING TO CD4
- DEATH ASSOCIATED WITH GP120 BINDING FOR CCR5
- Developing a Pro-Russian Population Inside Ukraine
- Criticisms of May's View
- CONTENTS