Tuesday, February 26, 2008

Overstock.com and Patrick Byrne: Phony Accounting and False and Misleading Disclosures - Revenues

In a previous blog post entitled, "Overstock.com and Patrick Byrne: Phony Accounting and False and Misleading Disclosures," I detailed how the Securities and Exchange Commission found that Overstock.com (NASDAQ: OSTK) had neglected to report its revenues in compliance with GAAP since its very first filings more than five years ago. In addition, in that blog post and two other previous blog posts (link here and here), I detailed how Overstock.com violated Securities and Exchange Commission Regulation G that governs the disclosure of non-GAAP financial measures. These two observations illustrate how Overstock.com continually cherry picks information it deems most favorable to the company, while omitting material disclosures that are important to investors. That would appear to be the primary reason that Overstock.com can't seem to keep its books in order.

Once again, Overstock has proved to be the story that keeps on giving. Every time I try to find the answer to one simple question, I am confronted by even more violations and inconsistencies. The latest evidence, analyzed in this blog post, indicates that the company has knowingly violated its own revenue recognition policies since at least 1999.

The SEC instructs Overstock.com to report its revenues in compliance with GAAP

In Overstock.com's recent press release disclosing the company's fourth quarter results, management made the following surprise disclosure:

Note regarding our Q4 and 2007 financial results:

From the company's inception through the third quarter of 2007, we have recorded revenue based on product ship date. In the fourth quarter of 2007, in response to an accounting comment from the staff of the SEC, we retrospectively changed our policy to recognize revenue based on estimated product delivery date. We have recorded the cumulative effect of this change in the fourth quarter of 2007.

Note: Bold print and italics added by me.

But Overstock.com's acknowledgment that "From the company's inception through the third quarter of 2007, we have recorded revenue based on product ship date" is only half the story. This misleading disclosure omits the fact that Overstock.com knowingly violated its own stated revenue recognition policies since at least 1999. In addition, contrary to the company’s stated policy, Overstock’s revenue accounting was in violation of Securities and Exchange Commission Staff Accounting Bulletin No. 101.

The latest discovery resulted when I decided to look back to the first publicly available financial reports issued by Overstock.com. In its first S-1 filed with the Securities and Exchange Commission on May 5, 2002, Overstock.com disclosed its revenue recognition policy which states in part:

For sales transactions, we comply with the provisions of Staff Accounting Bulletin 101 "Revenue Recognition" which states that revenue should be recognized when the following four revenue recognition criteria are met: (1) persuasive evidence of an arrangement exists; (2) the product has been shipped and the customer takes ownership and assumes the risk of loss; (3) the selling price is fixed or determinable; and (4) collection of the resulting receivable is reasonably assured. We generally require payment by credit card at the point of sale. Any amounts received prior to when we ship the goods to customers are deferred.

Note: Bold print and italics added by me.

The key takeaway is that Overstock states above that it is in compliance with Securities and Exchange Commission issued Staff Accounting Bulletin No. 101 (which went into effect in 1999). This revenue recognition policy was provided in financial statements reported in the company's S-1 from 1999, and has continued relatively unchanged ever since. Since 2004, the company has also claimed to be in compliance with the latest revenue recognition guidance from the SEC, Staff Accounting Bulletin No. 104. But, take a closer look at the highlighted disclosure:

(2) the product has been shipped and the customer takes ownership and assumes the risk of loss;

Note: Bold print and italics added by me.

Overstock.com asserts that, consistent with Staff Accounting Bulletins 101 and 104. that it recognizes revenue when "the product was shipped and the customer takes ownership and assumes risk of loss." However, as detailed below, the company deliberately failed to follow this stated revenue recognition policy, leading to a material departure from GAAP in every one of its SEC filings to date.

Patrick Byrne uses the “everyone else was doing it” excuse to explain why Overstock.com's revenue recognition policy was not in compliance with GAAP

During the fourth quarter fiscal year 2007 earnings conference call, Patrick Byrne used some more fancy footwork to try to avoid taking responsibility for the company’s latest accounting failure. This time the CEO claimed, more or less, that “everyone else was doing it, so we did too.” Of course, being the subject of an SEC inquiry, Byrne delivered a suitably lawyered up version of the story:

Let me explain real quickly, just when do you recognize a sale? Do you recognize it -- the convention has been and many – has been when you ship the product, in the case of when you ship things by UPS and FedEx, title does not actually change until the consumer receives it. At the beginning of our Company's history, we looked at it, and the difference was immaterial or small enough, and of course, we don't really know exactly when people receive their products in general or we did not eight years ago. Plus, as far as we could tell in the industry, everybody recorded at ship date. Amazon, according to their financial statements, recorded at ship date. So we chose the policy of recording at ship date. And one of the things that has happened in the last seven years is the information has gotten better where you can more feasibly look at when people receive their products. And more significant part of the industry has now migrated to recording at the date that things are received, even if it's an estimated date. And so we have put together, Dave has put together, the systems that can make that estimate, so it will have the effect of shifting two or three days of revenue.

Note: Bold print and italics added by me.

To make a long story short, the lawyered up version of the story is that, while the company was in violation, "we looked at it, and the difference was immaterial or small enough, and of course, we don't really know exactly when people receive their products in general or we did not eight years ago." The problem is, the lawyered up version of the story is still a significant problem according to SEC rules. Even if Overstock.com's departure from GAAP was "immaterial" by conventional percentage rules of thumb—which I don’t think it was, more on that later—under SAB No. 99, public companies are never permitted to make intentional immaterial departures from GAAP under any circumstances. Stated another way, if the GAAP departure was intentional, by definition it is unlawful.

Let's examine Securities and Exchange Commission Staff Accounting Bulletin No. 99

According to SEC Staff Accounting Bulletin 99, Overstock.com’s intentional GAAP violation cannot be ignored as immaterial:

2. Immaterial Misstatements That are Intentional

Facts: A registrant's management intentionally has made adjustments to various financial statement items in a manner inconsistent with GAAP. In each accounting period in which such actions were taken, none of the individual adjustments is by itself material, nor is the aggregate effect on the financial statements taken as a whole material for the period. The registrant's earnings "management" has been effected at the direction or acquiescence of management in the belief that any deviations from GAAP have been immaterial and that accordingly the accounting is permissible.

Question: In the staff's view, may a registrant make intentional immaterial misstatements in its financial statements?

Interpretive Response: No. In certain circumstances, intentional immaterial misstatements are unlawful.

Note: Bold print and italics added by me.

Therefore, the SEC's response to Byrne's explanation above is a flat out "No" and in certain circumstances even "intentional immaterial" departures from GAAP "are unlawful." But there’s more.

Was everyone else really doing it?

Naturally, I was skeptical when Patrick Byrne claimed, "Plus, as far as we could tell in the industry, everybody recorded at ship date. Amazon, according to their financial statements, recorded at ship date." Again, Byrne didn’t disappoint. After looking at Amazon’s financial statements, I found that Patrick Byrne’s comment was just another example of his half-true, materially misleading, doublespeak. According to Amazon's fiscal year 1999 10-K, issued after SAB No. 101 became effective:

The Company recognizes revenue from product sales, net of any promotional gift certificates, when the products are shipped to customers, which is also when title passes to customers.

Note: Bold print and italics added by me.

Note the subtlety. While revenue was recognized upon shipment, in the case of Amazon, title passed at the point of shipment. So Amazon met the standard.

This same disclosure appeared in Amazon's fiscal year 2002 10-K, the year that Overstock.com became a public company. For 2002, Amazon included the following revenue recognition disclosure:

Product sales, net of promotional discounts, rebates and return allowances, are recorded when the products are shipped and title passes to customers. Retail items sold to customers are made pursuant to a sales contract that provides for transfer of both title and risk of loss upon our delivery to the carrier (commonly referred to as “F.O.B. Shipping Point”).

Note: Bold print and italics added by me.

And again, in Amazon's fiscal year 2006 10-K, the company included a similar revenue recognition disclosure:

Product sales and shipping revenues, net of promotional discounts, rebates, and return allowances, are recorded when the products are shipped and title passes to customers. Retail items sold to customers are made pursuant to sales contracts that generally provide for transfer of both title and risk of loss upon our delivery to the carrier.

Note: Bold print and italics added by me.

So, to summarize, Amazon has consistently reported revenue when title passes—unlike Overstock.com, which reported revenue prior to the passage of title.

My take is that Overstock.com would like to fool investors into believing that it was following Amazon's revenue recognition policy when in fact, a very careful readings shows that there is one very important difference. While both policies claim to record sales when their products are shipped, in the case of Overstock.com, title does not actually pass until customer receipt. That means Overstock.com is in violation of Staff Accounting Bulletins 101 and 104.

SAB No. 99 clearly states that GAAP takes precedence over industry practices

So maybe you think Overstock is arguing that industry practices should take precedence and allow them to record sales at the point of shipment. Well, regardless of the baloney spewing from the mouth of double talking Overstock.com CEO Patrick Byrne, SAB No. 99 also expressly prohibits the use of industry accounting practices when they are in conflict with GAAP:

GAAP Precedence Over Industry Practice

Some have argued to the staff that registrants should be permitted to follow an industry accounting practice even though that practice is inconsistent with authoritative accounting literature. This situation might occur if a practice is developed when there are few transactions and the accounting results are clearly inconsequential, and that practice never changes despite a subsequent growth in the number or materiality of such transactions. The staff disagrees with this argument. Authoritative literature takes precedence over industry practice that is contrary to GAAP.

Note: Bold print and italics added by me.

What about that claim of “immateriality”?

Another problem with Mr. Byrne’s latest round of excuses is that, as far as I can tell, he’s telling another half truth when he says that the practice lead to an immaterial departure from GAAP. The problem as I see it is that, because the company reported a series of small losses (and two small profits) up until the end of 2004, even a very small change in sales is likely to produce a material misstatement in one or more quarters. This problem is even more pronounced in the only two quarters that Overstock reported a small profit. SAB 99 provides several considerations that would "render material a quantitatively small mistatement of a financial statement item" such as revenues:

Among the considerations that may well render material a quantitatively small misstatement of a financial statement item are -

  • whether the misstatement arises from an item capable of precise measurement or whether it arises from an estimate and, if so, the degree of imprecision inherent in the estimate
  • whether the misstatement masks a change in earnings or other trends
  • whether the misstatement hides a failure to meet analysts' consensus expectations for the enterprise
  • whether the misstatement changes a loss into income or vice versa
  • whether the misstatement concerns a segment or other portion of the registrant's business that has been identified as playing a significant role in the registrant's operations or profitability
  • whether the misstatement affects the registrant's compliance with regulatory requirements
  • whether the misstatement affects the registrant's compliance with loan covenants or other contractual requirements
  • whether the misstatement has the effect of increasing management's compensation – for example, by satisfying requirements for the award of bonuses or other forms of incentive compensation
  • whether the misstatement involves concealment of an unlawful transaction.

More misleading disclosures from Byrne and Company

In a previous blog item entitled, "Did Overstock.com's recent press release about Gradient violate Rule 10b-5?," I noted how Overstock.com has continued to make false and misleading disclosures when it suits the company's purposes. Well here we go again. In the case of the latest disclosure about Overstock’s noncompliance with revenue recognition rules, Byrne has once again violated Rule 10b-5.

To review part of my earlier post, recall that according to Securities and Exchange Commission Rule 10b-5:

It shall be unlawful for any person, directly or indirectly, by the use of any means or instrumentality of interstate commerce, or of the mails or of any facility of any national securities exchange,

a. To employ any device, scheme, or artifice to defraud,

b. To make any untrue statement of a material fact or to omit to state a material fact necessary in order to make the statements made, in the light of the circumstances under which they were made, not misleading, or

c. To engage in any act, practice, or course of business which operates or would operate as a fraud or deceit upon any person,

in connection with the purchase or sale of any security.

Note: Bold print and italics added by me.

To summarize, whether a public company violates GAAP, Regulation G governing non-GAAP financial disclosures, or generally any federal securities law, it is the presentation of untrue information or the omission of material information that gives rise to fraud (as opposed to a technical violation, without consequence to the public) under the federal securities law. Therefore, if a public company issues any statement to the public, whether in a press release or an SEC filing, Rule 10b-5 mandates (DEMANDS) that the company include all material information in connection with the subject matter of the document.

According to the Securities and Exchange Commission:

Section 10(b) of the Exchange Act and Rule 10b-5 proscribe, among other things, misstatements of material fact made in connection with the purchase or sale of securities. Information is material if there is a substantial likelihood that its disclosure would be viewed by a reasonable investor as having significantly altered the total mix of information available.

Note: Bold print and italics added by me.

So, once gain, Byrne and Overstock proved to be the gift that keeps on giving. Overstock.com's deliberate noncompliance of its revenue recognition policies and disclosures, the explanations offered by the company in its fourth quarter fiscal year 2007 press release, and Patrick Byrne's double talking explanations during the fourth quarter earning conference call are all clear violations of Rule 10b-5.

To be continued....

Written by:

Sam E. Antar (former Crazy Eddie CFO and a convicted felon)

Wednesday, February 20, 2008

Did Overstock.com's recent press release about Gradient violate Rule 10b-5?

On Tuesday, February 12, 2008 Overstock.com (NASDAQ: OSTK) issued a press release entitled, "Overstock Applauds Trial Court's Ruling in Gradient Case; Gradient Defendant Loses Dash for Exit as Court Sets September 9 Trial Date," hyping its litigation prospects in its ongoing lawsuit against Gradient Analytics, an independent research firm and Copper River Partners (formerly Rocker Partners), a short seller. However, Overstock.com failed to disclose in its press release the fact that Gradient Analytics recently filed a Motion for Leave to File a Cross-complaint against Overstock.com, Patrick Byrne, and others.

On Wednesday, the day after Overstock.com's press release, above, the company's stock rose to $11.58 per share from the previous closing price of $11.07 per share or a 4.6% increase (before the issuance of the press release).

Did Overstock.com violate Securities and Exchange Commission Rule 10b-5?

Let's review Securities and Exchange Commission Rule 10b-5:

It shall be unlawful for any person, directly or indirectly, by the use of any means or instrumentality of interstate commerce, or of the mails or of any facility of any national securities exchange,

a. To employ any device, scheme, or artifice to defraud,

b. To make any untrue statement of a material fact or to omit to state a material fact necessary in order to make the statements made, in the light of the circumstances under which they were made, not misleading, or

c. To engage in any act, practice, or course of business which operates or would operate as a fraud or deceit upon any person,

in connection with the purchase or sale of any security.

Note: Bold print and italics added by me.

As any securities attorney will tell you, Rule 10b-5 is the bedrock upon which federal securities law is built. Whether a public company violates GAAP, Regulation G governing non-GAAP financial disclosures, or generally any federal securities law, it is the presentation of untrue information or the omission of any material information that gives rise to fraud (as opposed to a technical violation, without consequence to the public) under the federal securities law.

If a public company issues any statement to the public, whether in a press release or an SEC filing, Rule 10b-5 mandates (DEMANDS) that the company include all material information in connection with the subject matter of the document.

According to the Securities and Exchange Commission:

Section 10(b) of the Exchange Act and Rule 10b-5 proscribe, among other things, misstatements of material fact made in connection with the purchase or sale of securities. Information is material if there is a substantial likelihood that its disclosure would be viewed by a reasonable investor as having significantly altered the total mix of information available.

Note: Bold print and italics added by me.

Information is material under the Rule:

(1) If it is more likely than not that such information if known to the reader would have been included his/her decision in connection with trading the securities of the company, or

(2) The information would alter the "mix of information" before the reader; therefore, if known would cause the reader to come to a different conclusion.

Overstock.com's press release cherry-picks one element of its litigation with the Gradient litigation, by pointing out that:

California Marin Superior Court Judge Lynn O'Malley Taylor ruled against Gradient Defendant Matthew Kliber's bid to exit the case. In addition, Judge O'Malley has set the case for trial on September 9, 2008.

In the press release, Jonathan E. Johnson III, Overstock.com's Senior Vice President of Legal went on to hype the company's litigation prospects:

Once again, defendants' efforts to derail this case fall flat. In every instance the trial and appellate courts have ruled that Overstock has a 'reasonable probability' of prevailing on every claim, just as the California Supreme Court declined Defendants' Petition for Review of those decisions. They are out of running room -- and we now have a trial date set. It's time to bring their conduct into the light of day."

And Patrick Byrne continued his non-stop hyping of the company's litigation prospects:

These broad-tossers have given us a two-and-a half year cook's tour of the California appellate process. They engage in such legal posturing because they know this case will expose a widespread, illegal Wall Street practice that harms American companies and makes money at the expense of American investors. I look forward to the day when the scofflaws are no longer in a backroom seeing their orthodoxies supped by biddable regulators and tractable New York financial journalists, but are instead sitting in a California courtroom trying to rationalize their schemes to twelve citizens."

However, Overstock.com's February 12 press release deliberately omits another new key element in its litigation: the fact that Gradient recently filed a motion for leave to file a cross-complaint. Gradient's cross-complaint, like Copper River's earlier cross-complaint covered in this blog, makes serious detailed allegations of willful misconduct and violations of law by Overstock.com, Patrick Byrne, and others. In addition, consistent with Overstock.com's pattern of failing to disclose material information to the public (like Byrne’s failure to disclose his personal subpoena), the company's website page covering the litigation fails to present a link to defendants’ cross complaints/countersuits, and omits any mention of, or even reference to, the existence of these material filings.

According to the "Proof of Service," Overstock.com and other Cross-Defendants (from the Copper River countersuit) were served on January 30, 2008 with the following documents:

1. Defendant Gradient Analytics, Inc.'s Notice of Motion and Motion for Leave to File Cross-Complaint;

2. Defendant Gradient Analytics, Inc.'s Memorandum of Points and Authorities in Support of Motion for Leave to File Cross Complaint;

3. Declaration of Brian P. Maschler, Esq. in Support of Defendant Gradient Analytics, Inc.'s Motion for Leave to File Cross Complaint; [Note: Exhibit A to this document includes Gradient's Cross Complaint];

4. [Proposed] Order Granting Defendant Gradient Analytics, Inc. Leave to File Cross-Complaint

The subject matter of Overstock.com's press release was an update in connection with its ongoing litigation between Overstock and Gradient. Since Overstock.com updated the public about the status of its litigation with Gradient in its press release, under rule 10b and 10b-5, the company cannot cherry pick only so-called favorable information and at the same time exclude new materially negative information.

Overstock cannot present a tenable legal argument to overcome the fact that, more likely than not, some members of the public would have included the omitted information of Gradient's request to file its countersuit in their decisions in connection with trading the securities of the company. The disclosure of Gradient's request to file its countersuit would have clearly altered the mix of information to the readers of its press release.

In short, the standard is not if, more likely then not, all members of the public would have included the omitted information---the standard is whether it is "reasonable to assume" that some members of the public would have included the omitted information in their decisions in connection with the trading of Overstock.com's securities.

Written by:

Sam E. Antar (former Crazy Eddie CFO and a convicted felon)

Monday, February 04, 2008

Overstock.com and Patrick Byrne: Phony Accounting and False and Misleading Disclosures

There is no end in site to the lies, distortions, deceptions, and false and misleading disclosures by Overstock.com's (NASDAQ: OSTK) management team, and in particular, its CEO Patrick Byrne. Demonstrating the level of desperation in Salt Lake City, the Overstock black ops team was set into high gear by mid-January. Early last week, the company's resident cyber stalker, Judd Bagley, attempted to smear me, in an effort to distract attention from Overstock.com's poor performance and accounting irregularities. You can read about that episode in investigative reporter Gary Weiss's blog post entitled, "Anatomy of an Overstock.com Smear" and in noted forensic accountant Tracy Coenen's blog post entitled, "Overstock "executive" playing on the internet again."

I believe that I present a problem for Patrick Byrne and Overstock.com because I began asking difficult questions right about the time that Byrne was touting his phony "turnaround". I say "phony" and "turnaround" (in quotes) because the turnaround never materialized, contrary to Byrne's assertions.

One of the problems facing Overstock.com's management team at the close of 2007 was the possibility of an unwanted participant on the fourth quarter earnings conference call, held on Wednesday January, 20. As I wrote about in an earlier blog item, during the third quarter earnings conference call, Patrick Byrne pretended that I was on the phone line, when I was not. This time I decided to call Patrick Byrne on his bluff. As noted in Gary Weiss's blog item entitled, "A Commissar Will Monitor This Conference Call," I sent Byrne an email asking him if I could ask questions about Overstock.com's accounting during the conference call. Patrick Byrne responded that he would only allow me to participate in Overstock.com's earnings conference call if I submitted three questions in advance, limited to 25 words each. He also imposed a condition that would limit my ability to ask follow up questions, by insisting that compound questions would have to be treated as multiple questions. I declined Patrick Byrne's conditions in an email reply that said:

It is highly irregular for a participant in a conference call to be singled out and limited to three questions posed in advance, each consisting of 25 words or less. I believe that such a condition would be at variance with the principles of free and full disclosure required by the securities laws at conference calls.

You appear to be imposing conditions on me that are not imposed on other conference call participants. I will not accept special conditions that are not imposed on other participants in the conference call.

Byrne's terms were a problem for several reasons. For one, the terms would have limited the time I had to analyze Overstock's fourth quarter results. It would also limit my ability to ask follow up questions. I also would not have the benefit of hearing management's dialogue and analysts' questions during the call to help me refine my questions-as other participants do during these calls.

I called into the earnings conference call anyway, hoping that Patrick Byrne would reconsider the special terms imposed on me. However, I was completely blocked from asking questions.

As detailed below, I can see why Patrick Byrne and his cronies have gone to extraordinary lengths to threaten me, harass me, defame me, and smear me. They don't like the questions I have asked in this blog and they simply didn't want me asking these questions in a live conference call. Overstock.com would rather answer softball questions heralded by gutless analysts craving for access to the unprincipled management team running Overstock.com.

The turnaround that wasn't

Since former Overstock.com President Jason C. Lindsey revealed that 2006 results were "bad on purpose," Patrick Byrne has been touting a turnaround all year. On the fourth quarter conference call, the CEO declared the turnaround a success:

...well, I know I said several times a year, a year and a half ago, that the turnaround prediction, the turnaround was going to look like this. We were going to start at the, as close to the bottom of the income statement as we could, and that meant getting the contribution dollars, the contribution profit turnaround, that's gross profit minus marketing. You'd see that returning a growth or hyper-growth, then you'd see growth in gross profit dollars, and only last would you see a revenue growth accelerate.

... go to slide five and gross profits grew in 2007 at a healthy 36% clip, and slide seven shows that contribution profits actually tripled to 100% growth. That's exactly as we predicted it would come. We actually feel - I feel great about that.

But, as I have shown all year in my blog, the turnaround has been nothing but a hoax fed by a combination of drawing on excessive reserves setup in the previous fiscal year and overstating EBITDA in violation of Regulation G. Even the cash flow that Byrne now hypes is really a result of accounting slight of hand.

It took some time, but the market seems to have finally seen through Byrne's con. Unfortunately, the investors that believed Byrne lost as much as 75% of their investment in 2007. Now, there are signs that the may also be SEC closing in on Byrne. With a little luck, the investors who were taken advantage of may soon see the truth about Overstock.com's accounting games.

Overstock.com issues earnings release - revenues

In its earnings release, Overstock.com reported an increase in revenues to $302 million in the fourth quarter of 2007 from $294 million in the previous year, or about a 2% increase in sales. That sales number fell well short of what you might have expected based on earlier touting by Byrne. In a December 7, 2007 interview with CNBC, Patrick Byrne sounded optimistic about revenues for Overstock.com's fourth quarter saying:

We are having a pretty nice Christmas," he said during the interview where he mentioned that the company's gross bookings had increased thus far during Q4 by approximately 10% over last year's Q4 due to an increase in the order size of the average transaction.

Note: Bold print and italics added by me.

None of the analysts attending Overstock.com's fourth quarter earnings conference call bothered to ask Patrick Byrne the following question that I would have asked:

Earlier in December, on CNBC, you said gross bookings had increased by about 10% as of December 7. But total sales for the quarter increased only 2%. What happened to sales in the last three weeks? Did the accounting error correction you disclosed in the earnings release account for some of this difference? Were there other contributing factors? (Note the compound question that would have used all three questions allotted to me by Byrne.)

Well, gross bookings (a non GAAP measure) did increase to $342.6 million in the fourth quarter compared to $319.6 million in the previous year's quarter, or about a 7% increase while GAAP revenues increased only 2% as noted above. So I suspect that the accounting error accounts for some of the difference. But, other problems also must have played a role.

Given how bad the results were, I also wonder why the CEO used the non-GAAP measure, "gross bookings," in touting sales growth during his television interview. This leads to another question I would have asked Byrne:

Patrick, why did you use a non-GAAP revenue measure to report sales expectations for the fourth quarter, knowing that GAAP sales were lagging so far behind the non-GAAP measure?

The Overstock management team most certainly did not want me to be able to ask questions like that. What better way to avoid it than to require that I submit my questions ahead of time, in writing, and limiting me to just three inquiries?

Did Patrick Byrne use non-GAAP "gross bookings" to hype fourth quarter sales on TV?

If we take Patrick Byrne at his word, then the last three weeks of the quarter must have been dismal. If the CEO is to be believed, sales must have declined in the last three weeks of the quarter to the extent that most of the revenues gains in the first ten weeks of the quarter were wiped out. But, using a non-GAAP financial measure such as gross bookings allowed the CEO to avoid disclosing the abysmal truth. While the rest of the online retail industry was enjoying double digit sales gains, Overstock.com's sales barely budged.

This is nothing new for the double-talking CEO who has routinely used non-GAAP financial measurements as they suit him, depending on his agenda at the time. In a previous blog item, I detailed how Byrne interchangeably used both GAAP and non-GAAP sales figures prior to Overstock.com's IPO in 2002. The trouble was, during that period he used these two sales measures interchangeably without ever telling the user which measure he was using, leading one to conclude that Overstock's sales were much higher than they really were at the time.

Speaking of revenues, the SEC finds revenue accounting irregularities at Overstock.com

It now appears that the SEC may be zeroing in on Overstock.com's accounting and disclosure issues. A comment in the company's fourth quarter earnings release confirms that the SEC has already had some impact on company accounting. The SEC started a formal investigation of the company on May 9, 2006. For over a year Byrne downplayed the investigation, falsely asserting that the SEC was actually after his imagined enemies, the "blackguards." However, on August 12, 2007, Byrne's version of reality appeared to turn even darker when he let it slip that he is actually the real target of the SEC's nearly two year long probe.

Byrne's admission and the latest Overstock revenue accounting disclosure could be a sign that regulators are beginning to take action. According to Overstock.com's fourth quarter earnings release:

From the company's inception through the third quarter of 2007, we have recorded revenue based on product ship date. In the fourth quarter of 2007, in response to an accounting comment from the staff of the SEC, we retrospectively changed our policy to recognize revenue based on estimated product delivery date. We have recorded the cumulative effect of this change in the fourth quarter of 2007.

Note: Bold print and italics added by me.

What is interesting about this error is how elementary it was. Any sophomore accounting student knows that you record sales when title transfers. In the case of Overstock, that is at product delivery. This makes me wonder just how competent is Overstock.com's CFO David K. Chidester? For that matter, how competent is the recently departed (for a second time) Jason C. Lindsey, the company's CFO until mid-2003.

As CPAs both Chidester and Lindsey should have known that you wait until title passes to book revenues. Worse yet, is that PricewaterhouseCoopers, Overstock.com's auditors, failed to notice the revenue recognition irregularities for nearly 10 years. How can that be? Well, unfortunately it is just more of the same for Overstock.com. As detailed in prior blogs, and again below, Overstock.com simply cannot get its act together on even the simplest of financial reporting issues.

Blame it on Amazon

During the fourth quarter conference call, Byrne downplayed the importance of the revenue accounting error. He also used another common tactic of deceptive CEOs, blame shifting. In the past, Byrne blamed the company's poor results on Sith Lords, cows crossing the road (that purported prevented delivery of goods), dead bodies and cocaine in trunks, and dozens if not hundreds of other unlikely causes. This time, it was Amazon's fault. According to Byrne, everyone else including Amazon was doing it, so Overstock did too.

During the conference call, Byrne claimed:

Let me explain real quickly, just when do you recognize a sale? Do you recognize it -- the convention has been and many -- has been when you ship the product, in the case of when you ship things by UPS and FedEx, title does not actually change until the consumer receives it. At the beginning of our Company's history, we looked at it, and the difference was immaterial or small enough, and of course, we don't really know exactly when people receive their products in general or we did not eight years ago. Plus, as far as we could tell in the industry, everybody recorded at ship date. Amazon, according to their financial statements, recorded at ship date.

So we chose the policy of recording at ship date.

Note: Bold print and italics added by me.

As a former fraudster, I've used that sort of tactic before too. So I knew to look at Amazon's 2006 10-K and see if Byrne's claim was true. I was not surprised to learn that, again, Byrne was deceiving investors. According to Amazon's 2006 10-K, filed almost a year ago, it recognizes revenue upon delivery, just like it is supposed to.

We recognize revenue from product sales or services rendered when the following four revenue recognition criteria are met: persuasive evidence of an arrangement exists, delivery has occurred or services have been rendered, the selling price is fixed or determinable, and collectibility is reasonably assured. Additionally, revenue arrangements with multiple deliverables are divided into separate units of accounting if the deliverables in the arrangement meet the following criteria: the delivered item has value to the customer on a standalone basis; there is objective and reliable evidence of the fair value of undelivered items; and delivery of any undelivered item is probable.

Note: Bold print and italics added by me.

So, contrary to what Patrick Byrne asserted in the fourth quarter call, Amazon records revenue upon delivery.

Overstock is one of those rare companies where, whenever I turn uncover one deception, I seem to find at least one more deception I wasn't already aware of. Wouldn't you know it? Looking at Amazon's revenue error disclosure revealed another Byrne deception. Look closely at the rest of Amazon's revenue accounting disclosure:

We evaluate the criteria outlined in Emerging Issues Task Force (EITF) Issue No. 99-19, Reporting Revenue Gross as a Principal Versus Net as an Agent, in determining whether it is appropriate to record the gross amount of product sales and related costs or the net amount earned as commissions. Generally, when we are primarily obligated in a transaction, are subject to inventory risk, have latitude in establishing prices and selecting suppliers, or have several but not all of these indicators, revenue is recorded gross. If we are not primarily obligated and amounts earned are determined using a fixed percentage, a fixed-payment schedule, or a combination of the two, we generally record the net amounts as commissions earned.

Commissions and per-unit fees received from third-party sellers and similar amounts earned through Amazon Enterprise Solutions are recognized when the item is sold by the third-party seller and our collectibility is reasonably assured. We record an allowance for estimated refunds on such commissions using historical experience.

Note: Bold print and italics added by me.

Byrne has always said that his company's mid-2003 revenue accounting change made it comparable to Amazon. But, after reading Amazon's revenue accounting policy, the two don't look comparable to me. Amazon's disclosure makes it sound like it records most of these "third party" sales (the same as "fulfillment partner" sales in Overstock's model) as "commissions" and "per unit fees."

Looking into the situation a little more, I found more information that also seems to conflict Byrne's assertions about Amazon accounting. Earlier in the 10-K, Amazon says that:

Through Amazon Services, we offer Amazon Marketplace and Merchants@ programs that enable third parties to sell their products on our websites, allow customers to shop for products owned by third parties using our features and technologies, and allow individuals to complete transactions that include multiple sellers in a single checkout process. We are not the seller of record in these third-party seller transactions, but instead earn fixed fees, sales commissions, per-unit activity fees, or some combination thereof.

Amazon Marketplace generally serves individuals and small businesses, enabling them to offer their products for sale on our websites alongside our products and products made available by other third-party sellers. Our Merchants @program generally serves larger, branded businesses and is primarily focused on expanding the selection of new products available on our websites. We also offer fulfillment-related services to third-party sellers through Fulfillment by Amazon.

Note: Bold print and italics added by me.

According to the same Amazon 10-K, during 2006, Amazon reported that 28% of sales during 2006 were reported under this "net" method of recognizing revenue:

Sales of products by third-party sellers on our websites represented 28%, 28%, and 26% of unit sales in 2006, 2005, and 2004. Since revenues from these sales are recorded as a net amount, they generally result in lower revenues but higher gross margin per unit. Since we focus on profit dollars rather than margins, we are largely neutral on whether an item is sold by us or by a third party.

So, unlike Overstock, Amazon reports these third party sales at the net amount of commissions earned. At about 28% of revenues last year ($10.7 billion) this equals approximately $3.0 billion in commissions recored as revenues. Taking this a little farther, if Amazon's margins on third party goods are 15% (about what most companies receive on these sales) then gross sales on third party goods would have been about $20.0 billion. That means if Amazon had used the same "gross" method that Overstock uses, its sales would have been $17.0 billion higher than reported. Yes, $17.0 billion higher, or $27.7 billion in total revenues.

So why does Overstock use this method that makes its sales seem higher, while Amazon uses a method that makes its sales seem much lower by comparison? This could make for a good follow up blog.

Was this the final adjustment?

Overstock tried to downplay the significance of the revenue cut-off error in its press release. But, the fine print still bothers me as that's often where management covers it's backside. Was Overstock.com covering its backside when it added the disclaimer copied below in its earnings release?:

Although we have initially concluded that this is the appropriate manner in which this change should be reflected in our results of operations for 2007, we are still in the process of reaching a final conclusion. If we determine it was not appropriate to make the cumulative adjustment in the fourth quarter, our results of operations will be adjusted to reflect the results solely under the revised method.

Note: Bold print and italics added by me.

Management also claimed during the fourth quarter call that, if its auditors and the SEC do not agree with the change, then they might have to allocate the change to other accounting periods. Does this also mean they might have to restate revenues? Maybe it is nothing. But, management rarely leaves the door open for more changes without reason.

And then there's the problem of former Director John A. Fisher's recent declaration in a court filing in response to Copper River's countersuit. In a previous blog item I noted how Overstock.com's directors appeared to be distancing themselves from Patrick Byrne. Now I am wondering about Fisher's conflicting statements regarding Overstock.com's mid-2003 revenue accounting change. In the declaration he claims the accounting change made Overstock.com comparable to Amazon. But as I discussed earlier, that is not true. Did Fisher really understand what he was signing off on when the audit committee approved that change?

Overstock.com again resorts to misstating EBITDA to hype company performance

Patrick Byrne once said, "...I think 'EBITDA' is the stupidest thing I ever heard emanate from Wall Street (no small feat)...." (From a 2006 letter attacking Business Week journalist and critic Tim Mullaney). But in the second quarter of 2007, Patrick Byrne had a sudden change of heart. He now claims that he has always believed in using EBITDA, but only in one extremely rare case. That case is a business turnaround but only when the company can reduce capital spending in the future. In my days as a CFO, I can't recall ever seeing that scenario. But, surprisingly, Byrne claims that's exactly the situation Overstock is in now. But, for some reason, Byrne can't even stick to the definition of EBITDA when hyping his supposed "turnaround."

I'm sure it has nothing to do with the fact that the company has never managed to post an annual GAAP profit in its entire history. It couldn't have anything to do with the fact that, with sales declining in the final weeks of the fourth quarter and for the entire year, the company had virtually no chance of reporting a GAAP profit, or even positive EBITDA, in the year of its much hyped "turnaround" (2007).

Overstock.com plays catch up with Reg G

Overstock lost money on a GAAP basis again in the fourth quarter. Byrne also continued to distract the audience with his own make-believe version of EBITDA. In the fourth quarter earnings release, and later repeated in that day's analyst call, CEO Patrick Byrne claimed that:

For the first time in our history, we were EBITDA positive for two consecutive quarters (Q3 and Q4 2007), generating positive $6.3 million of EBITDA during those six months.

Oh really? Below, I analyze how Byrne played games with the numbers by violating SEC Regulation G to materially overstate EBITDA in the combined third and fourth quarters by about $2 million or 46.5%. In addition, the reversal of excessive reserves booked by Overstock.com in the previous fiscal year inflated its reported EBITDA ever further.

As was the case with the simple issue of revenue cut-off, since Overstock.com began to tout its EBITDA performance in 2007, management can't get that measure right either. As I detailed in earlier blogs, it seems that Byrne and Overstock.com have managed to violate SEC Regulation G in every quarter since. Not content by the earnings inflation already provided by EBITDA, it seems Byrne needs just a bit more juice. That juice comes from an unusual calculation of EBITDA which has further misled investors. Some of those investors, by the way, have lost over 75% of their investment in less than three months as a result of what we now know to have been a con-the phony "turnaround".

In two previous blog posts entitled, "Overstock.com and CEO Patrick Byrne: Securities and Exchange Commission Regulation G Violations," and "Overstock.com and CEO Patrick Byrne: Improper use of EBITDA results in Regulation G Violation," I detailed how Overstock.com appears to have violated Securities and Exchange Commission Regulation G in its second and third quarter 2007 10-Qs. Now it appears that Overstock.com is attempting to play a game of catch up by correcting previous errors in non-GAAP financial reporting in its recent earnings release for the fourth quarter of 2007. Despite the efforts to clean up its disclosures, Overstock.com's EBITDA measure still seems to violate the SEC definition of EBITDA under Regulation G.

Let's review certain selected Overstock.com EBITDA disclosures during fiscal year 2007: This next table shows year-to-date operating loss and non-GAAP measures at the end of the second, third, and fourth quarters.

Overstock.com's EBITDA disclosures:

Source

Second Quarter 10-Q 2007

Third Quarter 10-Q 2007

Fourth Quarter 8-K 2007

Items

Six Months Ended June 30, 2007

Nine Months Ended September 30, 2007

Twelve Months Ended December 31, 2007

Operating Loss

(31,239)

(36,113)

(40,690)

Add:
Depreciation and amortization

15,745

22,825

29,495

Stock-based compensation

2,210

3,386

4,522

Stock-based compensation to consultants for service

140

280

189

Stock-based compensation relating to performance shares

-

350

(550)

Treasury stock issued to employees for compensation

715

928

902

Restructuring costs - asset impairment and depreciation

-

-

2,169

EBITDA

(12,429)

(8,344)

(3,963)

Add: Restructuring

12,283

Not Presented

Not Presented

Adjusted EBITDA

(146)

Not Presented

Not Presented

Overstock.com's restructuring costs

Items

Six Months Ended June 30, 2007

Nine Months Ended September 30, 2007

Twelve Months Ended December 31,2007

Restructuring Costs in Consolidated Statements of Operations

(12,283)

(12,283)

(12,283)

What's interesting about the year to date disclosures is that the story changes every quarter. In the second quarter 10-Q, Overstock.com included the full amount of restructuring costs ($12.283 million) as part of EBITDA. They then computed a non-GAAP financial measure called "Adjusted EBITDA" to eliminate the effects of its restructuring charges. That got them to a positive number (Adjusted EBITDA). But as demonstrated in a previous blog post, Overstock.com violated Regulation G which precludes removing charges that have occurred in the prior two years (or are expected to occur in the following two years) in non-GAAP performance measures such as adjusted EBITDA. In addition, as demonstrated in other previous blog posts, it was the reversal of excessive reserves setup at the end of the previous fiscal year that contributed to the entire amount of amount of Overstock.com’s positive “Adjusted EBITDA” in the second quarter. I will return to this issue later.

In the third quarter 10-Q, Overstock.com, as in its previous second quarter, included the full amount of restructuring charges ($12.283 million) in its computation of year to date EBITDA. However, the company did not disclose "Adjusted EBITDA" this time. Although not shown in my table above, this was probably due to the fact that the company did not have any restructuring charges in the third quarter. So, while the nine month EBITDA was negative, Byrne’s non-compliant EBITDA measure for the quarter was positive but (as illustrated later) overstated by a material amount. In addition, the reversal of excessive inventory reserves from the previous fiscal year accounted for most of the contribution to positive EBITDA for the third quarter. As long as you allowed him to make other improper deductions, like removing stock-based compensation, he could materially overstate EBITDA. I'll return to the issue of stock compensation a little later.

The fourth quarter year-to-date presentation shows yet a different presentation. This one eliminated the effects of just a fraction of the restructuring charges taken earlier in the year. In the fourth quarter 8-K, Overstock.com eliminated the effects of just $2.169 million of restructuring charges from EBITDA. In effect, this time Overstock.com presented "Adjusted EBITDA" as though it were the real thing, EBITDA.

Consistency be damned

The accounting profession has always valued consistency as one of its highest principles-right next to conservatism. But, if Overstock.com did not eliminate any of the $12.283 million of its restructuring charges from EBITDA calculation in the second and third quarter fiscal year 2007 10-Qs, why did it suddenly exclude $2.169 million of the $12.283 million from EBITDA in the fourth quarter? And, if somehow the new figure is correct, why were the old figures in error? How did that slip through the cracks? I doubt Patrick Byrne would have wanted me to ask this compound question during the conference call either. Unfortunately for me, I would have had to use up all three questions Byrne would allow just to get to the bottom of just this one issue.

Which EBITDA calculation is correct? Hint: According to Reg G, none are correct. The "E" in EBITDA means earnings, or net income.

The Securities and Exchange Commission has provided specific guidance regarding the computation of EBITDA under Regulation G. According to the SEC's "Frequently Asked Questions Regarding the Use of Non-GAAP Financial Measures" prepared by Staff Members in the Division of Corporation Finance, the following guidance is provided:

Question 14: Section I of the adopting release describes EBIT as "earnings before interest and taxes" and EBITDA as "earnings before interest, taxes, depreciation and amortization." What GAAP measure is intended by the term "earnings"? May measures other than those intended by the description in the release be characterized as "EBIT" or "EBITDA"? Does the exception for EBIT and EBITDA from the prohibition in Item 10(e) (1) (ii) (A) of Regulation S-K apply to these other measures?

Answer 14: "Earnings" is intended to mean net income as presented in the statement of operations under GAAP. Measures that are calculated differently than those described as EBIT and EBITDA in the adopting release should not be characterized as "EBIT" or "EBIDTA." Instead, the titles of these measures should clearly identify the earnings measure being used and all adjustments. These measures are not exempt from the prohibition in Item 10(e) (1) (ii) (A) of Regulation S-K.

Therefore, the intended meaning of "earnings" for EBITDA under Regulation G is "net income as presented in the statement of operations under GAAP." That is, Overstock should have based its computation of EBITDA on net income, adding back only interest, depreciation, and amortization (there were no taxes to add back).

In what appears to be direct violation of this requirement, Overstock.com instead based its computation of EBITDA on "operating loss" rather than "net income" as the starting point. As detailed in my previous blog post entitled, "Overstock.com and CEO Patrick Byrne: Improper use of EBITDA results in Regulation G Violation" this had the effect of materially overstating EBITDA for the second and third quarters of 2007, as well as the six months ended June 30, 2007 and the nine months ended September 30, 2007. Now Overstock.com is at it again in the fourth quarter 8-K by not reconciling EBITDA to net income, it also materially overstated its fourth quarter and full year 2007 EBITDA disclosure.

Even after starting with the wrong number (operating loss, rather than net loss) and mysteriously adding back some restructuring charges that had never been added back before, Overstock.com still had to resort to adding back a portion of management's compensation, as if somehow that doesn't count either. As NY Times writer Floyd Norris pointed out in a blog he authored while the conference call was still ongoing:

The biggest reason the six-month Overstock EBITDA numbers are positive is all the depreciation expense, which reflects old investments the company made before it cut back on investments to conserve cash. But the EBITDA figure would have been $2.9 million lower for the six months if they had counted costs from stock and options as investments.

Why a company should show higher profits (even EBITDA profits) by handing out shares rather than cash to employees is beyond me.

Let's review Overstock.com's fourth quarter 2007 EBITDA disclosure in its recent 8-K:

EBITDA - EBITDA (a non-GAAP measure) for the three months ended December 31, 2006 and 2007 was $(27.4) million and $2.2 million, respectively. For the twelve months ended December 31, 2006 and 2007, EBITDA was $(55.7) million and ($4.0) million, respectively. We believe that, because our current capital expenditures are significantly lower than our depreciation levels, discussing EBITDA at this stage of our business is useful to us and investors because it approximates actual cash used or cash generated by the operations of the business.

Items

Three Months Ended December 31, 2006

Three Months Ended December 31, 2007

Twelve Months Ended December 31, 2006

Twelve Months Ended December 31, 2007

Operating Loss

(40,885)

(4,577)

(93,766)

(40,690)

Add:
Depreciation and amortization

11,525

6,670

32,327

29,495

Stock-based compensation

1,032

1,136

4,120

4,522

Stock-based compensation to consultants for service

(8)

(91)

23

189

Stock-based compensation relating to performance shares

-

(900)

-

(550)

Treasury stock issued to employees for compensation

108

(26)

787

902

Restructuring costs - asset impairment and depreciation

791

-

791

2,169

EBITDA

(27,437)

2,212

(55,718)

(3,963)

Net loss - Net loss for the three months ended December 31, 2006, was $45.6 million, or $2.15 loss per share, compared to $4.3 million, or $0.18 loss per share in 2007. For the twelve months ended December 31, 2006 and 2007, net losses totaled $101.8 million and $44.1 million, respectively, or $5.01 and $1.86 loss per share for those respective periods. For the twelve months ended December 31, 2006 net loss included restructuring charges of $5.7 million and $6.9 million loss from discontinued operations. Net loss in 2007 includes restructuring of $12.3 million and loss from discontinued operations of $3.9 million.

Note: Bold print and italics added by me.

Now let's compute EBITDA as defined by SEC Regulation G

Items

Three Months Ended December 31, 2006

Three Months Ended December 31, 2007

Twelve Months Ended December 31, 2006

Twelve Months Ended December 31, 2007

Net Loss attributable to common shares

(45,614)

(4,251)

(101,865)

(44,106)

Add:
Depreciation and amortization

11,525

6,670

32,327

29,495

Interest Income

(577)

(1,429)

(3,566)

(4,788)

Interest Expense

1,127

1,103

4,765

4,188

Other income, net

(88)

-

(81)

92

Deemed dividend related to redeemable common stock

-

-

99

-

EBITDA as defined under SEC Regulation G

(33,627)

2,093

(68,321)

(15,119)

Now let's compute Overstock.com's overstatement of EBITDA as defined by SEC Regulation G:

Items

Three Months Ended December 31, 2006

Three Months Ended December 31, 2007

Twelve Months Ended December 31, 2006

Twelve Months Ended December 31, 2007

EBITDA per Overstock.com's disclosures

(27,437)

2,212

(55,718)

(3,963)

EBITDA as defined under SEC Regulation G

(33,627)

2,093

(68,321)

(15,119)

Overstatement of EBITDA

6,190

119

12,603

11,156

Percentage overstatement of EBITDA

22.56%

5.69%

22.62%

281.5%

As the charts above illustrates, Overstock.com overstated EBITDA as defined by the SEC under Regulation G for all four periods presented in its 8-K and by material amounts (5% or over). In fact, in three of the periods presented in the latest 8-K, EBITDA was misstated by over 20%. Incredibly, in the most important period presented (full year 2007), the period of Byrne's purported turnaround, EBITDA was overstated by nearly 300% according to Regulation G.

Now let's review Overstock.com's EBITDA disclosures for the second, third quarter, and fourth quarter of fiscal year 2007 side by side

Item

Three Months Ended June 30, 2007

Three Months Ended September 30, 2007

Three Months Ended December 31, 2007

Operating loss

(13,519)

(4.874)

(4,577)

Add:
Depreciation and amortization

7,974

7,080

6,670

Stock-based compensation

1,137

1,176

1,136

Stock-based compensation to consultants for service

135

140

(91)

Stock-based compensation relating to performance shares

-

350

(900)

Treasury stock issued to employees for compensation

113

213

(26)

EBITDA

(4,160)

4,085

2,212

Add: Restructuring

6,194

Not Presented

Not Presented

Adjusted EBITDA

2,034

Not Presented

Not Presented

Now let's compute EBITDA as defined by SEC Regulation G for the second, third, and fourth quarter side by side:

Items

Three Months Ended June 30, 2007

Three Months Ended September 30, 2007

Three Months Ended December 31, 2007

Net Loss attributable to common shares

(13,768)

(4,704)

(4,251)

Add:
Depreciation and amortization

7,974

7,080

6,670

Interest Income

(1,078)

(1,291)

(1,429)

Interest Expense

1,027

1,029

1,103

Other income, net

-

92

-

EBITDA as defined under SEC Regulation G

(5,845)

2,206

2,093

Now let's compute Overstock.com's overstatement of EBITDA as defined by SEC Regulation G for the second, third, and fourth quarter side by side:

Items

Three Months Ended June 30, 2007

Three Months Ended September 30, 2007

Three Months Ended December 31, 2007

EBITDA per Overstock.com's disclosures

(4,160)

4,085

2,212

EBITDA as defined under SEC Regulation G

(5,845)

2,206

2.093

Overstatement of EBITDA

1,685

1,879

119

Percentage overstatement of EBITDA

40.50%

85.18%

5.69%

As the charts above illustrates, Overstock.com materially overstated EBITDA as defined by the SEC under Regulation G in each of the last three quarters of fiscal year 2007. The company overstated EBITDA by about $2 million or 46.5% in the last two quarters combined and about $3.683 million in the last three quarters combined (all material amounts).

Overstock's version of consistency?

If there is one area where Overstock has been consistent, it is in the sheer number of violations this past year. This brings me to the reconciliation requirement in Regulation G. According to the Regulation, a company must reconcile its non-GAAP financial measure with the "most directly comparable financial measure or measures calculated and presented in accordance with GAAP." On this the Regulation says:

Regulation G contains a general disclosure requirement and a specific requirement of a reconciliation of the non-GAAP financial measure to the most directly comparable GAAP financial measure.

Note: Bold print and italics added by me.

The SEC goes on to state that:

Whenever a company that is subject to Regulation G, or a person acting on its behalf, publicly discloses any material information that includes a non-GAAP financial measure, Regulation G requires the registrant to provide the following information as part of the disclosure or release of the non-GAAP financial measure:

• a presentation of the most directly comparable financial measure calculated and presented in accordance with GAAP; and

• a reconciliation (by schedule or other clearly understandable method), which shall be quantitative for historic measures and quantitative, to the extent available without unreasonable efforts, for prospective measures, of the differences between the non-GAAP financial measure presented and the most directly comparable financial measure or measures calculated and presented in accordance with GAAP.

Note: Bold print and italics added by me.

What is the most directly comparable financial measure calculated and presented in accordance with GAAP?

Maybe Overstock's argument is that, under the consistency principle, when you begin with an erroneous number in calculating a non-GAAP measure, you should reconcile to the same erroneous number. But, that makes about as much sense as the sophomore accounting teacher changing the exam key to match David Chidester's mid-term answers. No, we should look to the Regulation itself for guidance on the right reconciliation figure, rather than making it up as it suits us.

The SEC did not mandate or define "the most directly comparable financial measure." However, the SEC provides certain guidance depending on whether or not the non-GAAP financial measure is intended to be considered a performance measure or a liquidity measure. According to Regulation G:

We do intend that the definition of non-GAAP financial measure capture all measures that have the effect of depicting either:

• a measure of performance that is different from that presented in the financial statements, such as income or loss before taxes or net income or loss, as calculated in accordance with GAAP; or

• a measure of liquidity that is different from cash flow or cash flow from operations computed in accordance with GAAP. [Bold and italics added by me.]

Therefore, a non-GAAP performance measure must be reconciled to items such as "income or loss before taxes or net income or loss" or a measure of liquidity such as "cash flow from operations."

Further, the SEC guidance states that, if the non-GAAP EBIT or EBITDA financial measure is a performance measure, the registrant should make the following reconciliation:

Question 15: If EBIT or EBITDA is presented as a performance measure, to which GAAP financial measure should it be reconciled?

Answer 15: Because EBIT and EBITDA exclude [four specific] recurring charges [interest, taxes, depreciation and amortization], companies should consider the answer to Question 8 if they intend to use EBIT or EBITDA as a performance measure. If a company is able to justify such use, EBIT or EBITDA should be reconciled to net income as presented in the statement of operations under GAAP. Operating income would not be considered the most directly comparable GAAP financial measure because EBIT and EBITDA make adjustments for items that are not included in operating income.

Note: Bold print and italics added by me.

In short, if EBITDA is presented as a performance measure, it must be reconciled to net income (or loss, in the case of Overstock).

On the other hand, if the non-GAAP number is intended as a liquidity measure, the SEC provides the following guidance:

Question 12: Are the requirements in Item 10(e) (1) (i) of Regulation S-K for the prominent presentation of, and reconciliation to, the most directly comparable GAAP financial measure or measures intended to change the staff's historical practice of requiring the prominent presentation of amounts for the three major categories of the statement of cash flows when a non-GAAP liquidity measure is presented?

Answer 12: No. The requirements in Item 10(e) (1) (i) are consistent with the staff's historical practice. The three major categories of the statement of cash flows should be presented when a non-GAAP liquidity measure is presented.

Note: Bold print and italics added by me.

This means that the only other alternative available to Overstock was to reconcile EBITDA to one of the three measures of cash flow reported on the cash flow statement.

Well, it's pretty clear from Overstock's reconciliation which I presented earlier, the company did not follow Regulation G. Why? Overstock.com did not reconcile EBITDA to net income nor did it reconcile to one of the three major categories of cash flow. But, I find it is still helpful to examine the nature of the error to better understand management's actions. That might help us figure out how the error was made.

So which measure should Overstock have reconciled to?

Was it a Performance Measure or a Liquidity Measure?

Based on Overstock.com's disclosures in the fourth quarter fiscal 2007 8-K, it appears as though management intends to use EBITDA as a liquidity measure. Specifically, Overstock.com disclosed:

We believe that, because our current capital expenditures are significantly lower than our depreciation levels, discussing EBITDA at this stage of our business is useful to us and investors because it approximates actual cash used or cash generated by the operations of the business.

Note: Bold print and italics added by me.

Therefore, Overstock.com it should have reconciled EBITDA to one of the three major categories of cash flows.

So why didn't Overstock make the required reconciliation? Was it ignorance of the standard? Or was it trying to hide the fact that, in the case of Overstock.com, EBITDA is not a "reasonable measure of actual cash used or generated by the operations of the business"?

Impact of reversal of excessive inventory reserves on Overstock.com's EBITDA

Now let's consider comments made by former Overstock.com President Jason C. Lindsey at the fourth quarter 2006 earnings call before Overstock.com disclosed its inventory reserves in its 10-K:

We took all that to heart in the fourth quarter and although the fourth quarter results are very bad, and I admit they are very bad, they were bad on purpose. In other words, we used the fourth quarter to get rid of all the slow-moving inventory. I am quite pleased with the inventory balances we have now....I am pleased that the fourth quarter is now over. We have sold it. Our inventory turns are much higher. Our margins are much higher and it really does feel like we have made a lot of progress there.

Note: Bold print and italics added by me.

The key words are “I admit they are very bad, they were bad on purpose…. we used the fourth quarter to get rid of all the slow moving inventory..... We have sold it.” But, the bad inventory wasn’t gone if you are to believe Overstock.com's 10-K disclosures. This is clear because inventory reserves rose from about $4.5 million at the end of the third quarter to about $6.6 million at the end of the fouth quarter. On a relative basis, inventory reserves went up a staggering 299% from about 6.14% of gross inventory at the end of the third quarter to about 24.52% of inventory at the end of the fourth quarter. If Overstock.com had sold its entire fourth quarter inventory at its original cost basis during the following year, the company would have realized gross margins of 24.54%, far higher than it has ever reported in its operating history.

How could Overstock.com have “sold it” and gotten “rid of all the slow moving inventory” if the company actually estimated that it had more slow moving inventory in both absolute dollars and relative amounts? Was Overstock.com booking excessive inventory reserves at the end of the fourth quarter of 2006 to help boster future earnings? After all, Lindsey had claimed that Overstock.com made its fourth quarter results "very bad on purpose" and that the company had sold "all of the slow-moving inventory" in the fourth quarter.

As discussed above and in previous blog posts, the reversal of excessive inventory reserves set up at the end of the previous fiscal year had a significant material impact on Overstock.com’s reporting of positive EBITDA in the third quarter of 2007. Inventory reserves dropped about $1.6 million during the third quarter to $2.2 million from $3.8 million at the end of the second quarter, and therefore contributed about 73% of the recomputed positive EBITDA as defined under SEC Regulation G. Therefore, after factoring out the reversal of excessive inventory reserves, EBITDA as computed under SEC Regulation G would have been a mere $606 thousand, far less than the $4.085 million EBITDA improperly reported by Overstock.com.

Going back to the second quarter of fiscal year 2007, Overstock.com reported “Adjusted EBITDA” (which eliminated the effects of restructuring charges as described above) of $2.034 million. However, inventory reserves dropped about $2.7 million during the second quarter to $3.8 million from $6.5 million at the end of the first quarter. Even without considering that Overstock.com had improperly eliminated restructuring costs from "Adjusted EBITDA" in violation of SEC Regulation G, Overstock.com’s reported positive “Adjusted EBITDA” squarely resulted from the reversal of excessive inventory reserves. If we factor out the reversal of excessive inventory reserves ($2.7 million) from Overstock.com's improperly reported "Adjusted EBITDA" of $2.034 million, Overstock.com's reported "Adjusted EBITDA" would have been negative $666 thousand.

To make matters worse, as described in a previous blog post, Overstock.com reversed $1 million of previously recorded restructuring expenses during the second quarter of fiscal year 2007. The reversal of restructuring expense of approximately $1 million resulted in increasing Overstock.com's improperly reported "Adjusted EBITDA" by yet another million dollars. Therefore, if we factor out both the reversal of excessive inventory reserves and restructuring costs, Overstock.com's improperly reported "Adjusted EBITDA" would have been negative $1.666 million. If we compute EBITDA under SEC Regulation G and factor out the combined contributions to EBITDA resulting from the reversal of excessive inventory reserves and restructuing costs, Overstock.com's properly reported EBITDA would have been negative $9.545 million in the second quarter.

The cash flow that wasn't

Patrick Byrne has also been hyping Overstock's cash flows as part of the phony "turnaround". For example, in the fourth quarter earnings release, Byrne claimed:

Free cash flow for the year was positive $7.4 million and operating cash flow was positive $10.1 million.

Byrne makes similar comments in the conference call later that day. But what Byrne doesn't tell you, the most important part of the story, is that Overstock's operating cash flow was positive because he took longer to pay his suppliers this year. If you take out the $13.0 million increase in accounts payable, operating cash flow falls to negative $2.9 million. Subtracting out the increase in accounts payable, free cash flow plummets to negative $5.6 million. But that doesn't tell the full story either.

Overstock spent just $2.6 million for capital expenditures in 2007. That number is clearly not repeatable for an online retailer like Overstock. If the company had spent the same amount in 2007 as it spent in 2006, free cash flow would have been at least $25 million in the red.

In the future, Overstock.com will need to spend considerably more if they are going to stay competitive. Overstock's competition includes extremely well funded online retailers like eBay, Amazon, and Blue Nile and established brick and mortar stores who also have strong online operations, including Walmart, Target, and Circuit City. Do you think these guys will be cutting back on technology spending in the future? Not a chance.

To be continued.....

Written by:

Sam E. Antar (former Crazy Eddie CFO and a convicted felon)