Tuesday, August 02, 2005

The Supercalifragilisticexpialidocious Marketing Experiment and Other “One-Time Items”

It’s the tale-end of earnings season, at least for companies on a calendar fiscal year, and for the most part earnings have been terrific—despite a very mixed reaction among the stocks themselves, especially in the technology sector.

For example, Apple, IBM, Amazon and eBay had nothing particularly spectacular to say—and yet their stocks took off based on little more than a relief that earnings had not disappointed and that guidance was, as Wall Street’s Finest like to say, “upbeat.”

Yahoo, Google and Microsoft, on the other had, reported no great tragedies—yet guidance was not judged “upbeat” enough, making the Street suddenly “downbeat” about their prospects and, therefore, sellers of those stocks.

So it is that these and other well-run companies have reported, and what is left are the stragglers—and one in particular that we monitor rather closely which will report tomorrow (following a one-week delay reportedly owing to unforeseen complications from a headquarters move and technology investments and an acquisition).

The company, of course, is Overstock.com, and in the spirit of this blog, which is to look at facts rather than engage in speculation, I thought it worthwhile to examine Wall Street’s earnings expectations for the company and provide my own preview.

For starters, Wall Street’s Finest are all predicting a fairly tight range of revenues right around $150 million, which their computers all seem to translate into a net loss in the equally tight 22c to 24c per share range.

One independent outlier expects far lower sales and somewhat lower earnings than the Street so we will ignore him for purposes of assessing "the consensus." The most bullish "consensus" analyst, as I can see, is Craig Bibb, who assumed coverage at WR Hambrecht in April, a few months after the previous analyst, Bill Lennan, cut his earnings and revenue numbers and took his price target from $85 to $60, helping spark a decline in the stock.

Bibb took over for Lennan with a buy rating and a decidedly cheery note in which he compared Overstock.com CEO Patrick Byrne to Olympic Ski Champion Bode Miller. I am not making that up.

Bibb further ingratiated himself with Overstock by writing that “Management is taking a dynamic approach to finding the efficient frontier of growth and profitability…,” precisely the kind of Wall Street lingo that makes sense only if you start your day sniffing glue.

But back to our earnings preview: Wall Street expects Overstock.com to report a $4 million-plus operating loss in the second quarter of 2005—about a million worse than the $3.4 million operating loss reported in the first quarter of 2005.

This is based on a modest sequential decline in revenues, a pick-up in gross margins, and far higher selling, marketing, administrative and technology costs than in Q1.

To the naïve observer, this losing-money forecast makes no sense, because in his first quarter 2005 letter to investors Patrick Byrne identified $7.9 million worth of supposedly one-time items that reduced first quarter earnings, without which the company would have been handsomely profitable.

(Byrne had done the same thing in the fourth quarter of 2004, identifying $6.5 million of one-time items that supposedly hurt reported earnings.)

The items enumerated by Byrne this time included a $1.2 million shipping promotion; a $600,000 blown electronics deal; a $1.2 million bonus accrual; $1.8 million to market an auction and jewelry site; $500,000 on technology consultants; and a $2.6 million “binomial marketing experiment” which apparently, like Frankenstein’s Monster, went horribly wrong before it could be killed.

(For the record, even though nobody listening to that conference call had the faintest idea what constituted a “binomial marketing experiment,” Wall Street’s Finest dutifully wrote down “binomial marketing experiment” in their word processors and sprinkled the howler throughout their research reports later that afternoon and in subsequent documents, without ever actually explaining what a “binomial marketing experiment” might be.

I believe Patrick Byrne could have called it a “Supercalifragilisticexpialidocious marketing experiment,” and Wall Street’s Finest would have dutifully written “Supercalifragilisticexpialidocious marketing experiment” in their reports without a second thought.)

So, back to the first quarter numbers, and assuming Byrne was being accurate with those numbers on the conference call—keep in mind this is a man who tosses around numbers on conference calls like bond traders toss around $50 bills at Scores—Overstock would have reported $3.5 million worth of positive operating earnings in the first quarter.

And, therefore, without the $2.6 million Supercalifragilisticexpialidocious marketing experiment, the $1.2 million shipping promotion, the $1.2 million bonus accrual, the $600,000 blown electronics deal, the $500,000 on consultants and the $1.8 million on auction and jewelry sites that are stumbling out of the gate, Overstock should be able to make money in the second quarter—even assuming higher depreciation, higher rent and higher technology costs.

Unless, of course, Doctor Byrne was not being precisely, um, accurate with the $7.9 million worth of so-called one-time costs.

For help on this, I turned to Overstock’s “Management’s Discussion and Analysis of Financial Condition and Results of Operations” (usually referred to as the MDA) in Overstock’s first quarter 10Q filing with the SEC, where companies normally enumerate those items which had significant impacts on earnings.

And in so doing I couldn’t find even one of the $7.9 million worth of items highlighted by Doctor Byrne.

For the record, the MDA is the place in a 10Q and 10K where companies are expected to discuss the guts of their business and report important factors behind changes in sales, margins and expenses.

Here’s what the SEC expects in the MDA:

The Commission has long recognized the need for a narrative explanation of the financial statements, because numerical presentations and brief accompanying footnotes alone may be insufficient for an investor to judge the quality of earnings and the likelihood that past performance is indicative of future performance. MD&A is intended to give the investor an opportunity to look at the company through the eyes of management by providing both a short and long-term analysis of the business of the company.

The discussion and analysis shall focus specifically on material events and uncertainties known to management that would cause reported financial information not to be necessarily indicative of future operating results or of future financial condition.

For the record, Amazon’s latest-quarter MDA runs for 28 pages, with great detail on all manner of sales and expense items.

Google’s first-quarter 2005 MDA runs 36 pages.

Overstock.com’s Q1 MDA runs for a little over 2 pages, offering a bare-bones description of the company and the components of its income statement, and that’s about it. Pull it off the web site yourself: you will find that in place of what normally constitutes the bulk of the MDA is something called an “Executive Commentary” which, for some reason, is excluded from the MDA with the following disclaimer:

This executive commentary is intended as a supplement to, but not a substitute for, the more detailed discussion of our business elsewhere herein.

Yet even here, only the $2.6 million marketing experiment was mentioned, and none of the other cost items Byrne had highlighted.

You’d expect a company that missed a quarter thanks mainly to $7.9 million worth of one-offs would explain this somewhere.

And so we await tomorrow’s report, which will no doubt include a wonderfully descriptive letter by the ever-erudite Doctor Byrne, explaining things to Wall Street’s Finest, and a conference call with Wall Street's Finest asking acute and piercing questions of Doctor Byrne.

In the meantime, anybody with an informed opinion on the nature of those Supercalifragilisticexpialidocious “one-time” costs that seem to crop up time after time in the shareholder letters and conference call discussions is welcome to enlighten our readers here

Tomorrow we will see whether Wall Street’s Finest—who expect a loss; or the naive reader of Overstock’s first quarter earnings report—who might expect a profit if for nothing else than the non-recurrence of $7.9 million worth of one-time items, is right.

Jeff Matthews
I Am Not Making This Up

The content contained in this blog represents the opinions of Mr. Matthews. Mr. Matthews also acts as an advisor and clients advised by Mr. Matthews may hold either long or short positions in securities of various companies discussed in the blog based upon Mr. Matthews' recommendations.


DS said...

Jeff, this is fabulous! Why can't we find more people like yourself on Wall Street? The level of BS & stupidity is astonishing, but so little is done to reduce it. Bravo to you. Keep 'em coming...

Its_strange said...

I second ds. This is good stuff.

coolbreeze_23 said...

Is Byrne the guy who was castigating the short sellers on Motley Fool a few months ago?

Vaibhav Gautam said...

Really useful article. Especially, when I am these days involved in studying the financial statements of the companies as part of MBA!
I am one of your old silent readers from India!
When nothing is mentioned in the 10Q reports about these "one time expenses", do you really think that investors can be fooled into buying the management story of "one time expenses"! Aint there any regulatory requirements which can stop the company managment from spreading these "one time expenses" stories?
~ Gautam
~ http://www.blurty.com/~stonedmind

tom said...


Keep up the good reporting. Earnings speak for them selves. How many years do you have to be a company to produce earnings... perhaps critical mass hasn't been reached yet at OSTK. Yeah right, how about Management, is one guy with a whole bunch of yes men, with minor experience. How come of their VPs left last summer to get an MBA, yes to go get an MBA?

Can't run a company with a full team of also rans (hungry for lunch instead of success). Perhaps they have six sigma black belts or some other crap instead of
hiring folks that have real operational experience at a true company.


Jeff Matthews said...

For the record, 126 diamonds vanished from the Overstock site over the weekend, with an average price estimated to exceed $3,000.

This was almost certainly another wholesale-type diamond sale, not 126 individuals ordering engagement rings, because aside from wholesale deals, only about 23 diamonds have been sold one-at-a-time in the last six months--just about 1 diamond per week.

The sale of 126 diamonds is similar to but much larger than the sale of 38 diamonds back in April--which we broke in this space well before Patrick Byrne acknowledged it on the conference call.

Last weekend's sale would amount to roughly $400,000 worth of diamonds, a big number that Byrne could use on tomorrow's call when he needs to say something positive, such as, "our build-your-own-ring sales really took off in July."

As I said, only about 23 diamonds have been sold one at a time since Byrne established Overstock's Blue Nile-knockoff site. This amounts to perhaps $100,000 in sales in the last six months.

Blue Nile, on the other hand, just reported $43.8 million sales for the last quarter.

As for the Yahoo message board individual attempting to hijack this blog, we do not allow vindictive message board style chat here.

Especially from individuals who resort to posting the names of children and wives of their perceived enemies on message boards, while hiding behind an assumed identity.

Jeff Matthews
I Am Not Making This Up

bsilly said...


What's the point of your article? - Ostk disclosed info in the wrong place. I don't really get what your after. Are you suggesting Patrick is cooking the books? As you know I'm long Ostk. Do you have any comments on the competitive landscape or business model that would suggest ostk is grossly mispriced.


Its_strange said...

I'm wondering if others had thoughts these wholesale diamond deals are really barter deals ?

Jeff Matthews said...

"bsilly": I have never commented on the valuation of OSTK and am not going to start doing so now, in response to your question about whether it is "overpriced." That's for each investor to determine for themselves.

As to the rest of your question, which was really excellent, I am raising questions about material issues and looking for informed answers. I have none myself.

Jeff Matthews said...

"bsilly": I have never commented on the valuation of OSTK and am not going to start doing so now, in response to your question about whether it is "overpriced." That's for each investor to determine for themselves.

As to the rest of your question, which was really excellent, I am raising questions about material issues and looking for informed answers. I have none myself.

Michael Parekh said...

Very good post, Jeff...I had to brush up my knowledge on "binomial probabilities" to figure out that a $2.6 million binomial anything would need a series of samples to determine the "binomial probability". I don't know OSTK that well, but am pretty sure they didn't have a series of these $2.6 million binomial probabilities".

Hilarious stuff you're not making up.

bsilly said...


I wasn't really talking about valuation per se. If you go long or short you've got to believe that the security is mispriced in some form, unless your strictly a mo mo guy.


nukeydog said...

Jeff, "sniffing glue" is such the "yesterday" am drug of choice...youre showing your age...the new thing is computer dusting spray or "dusting" (I'm not making this up). Don't try it!

k9thunder said...


Good job!

You're brilliant in reading thru pile of deCRAPitating rant into bring out oh so many holes and inconsistencies. I at times think Patrick is on the verge of lunacy. Too bad CA state institution in Camarillo CA turned into Cal State campus.

As for Bobo O'Brien Davidson, please ignore and better yet delete his ranting nonsense.


Edge Trader said...

Jeff, thanks, got a great laugh at this post.

Forensic Doc said...

That "binomial experiment" comment cracked me up too. The binomial test is just ONE of many tests that you might run on a set of data. The fact that the experiment use a binomial test means nothing. In fact, if that's the only test they used, that's a really bad way to spend $2.6 million.

How do you spend $2.6 million on a marketing "experiment" anyway? Computer time is cheap. The only real incremental cost of a "true" experiment is labor. How many Stat PhD's can you get for $2.6 million in one quarter? It's more than 100, that's for sure.

Obviously there must be something else in the cost of that "experiment" gone wrong. My guess is that it was simply an attempt to classify some portion of actual marketing costs for the quarter as a "nonrecurring" item. That way analysts would get the impression that ongoing marketing costs are lower than what was spent during Q1.

So what drove the cost per aquired customer even higher this quarter? Are failed experiments really nonrecurring? Or are they just an ordinary cost of doing business?