Jesse Eisinger, one of the best investigative journalists around, raises some very good questions about Green Mountain Coffee (NASDAQ:GMCR) and its accounting in his column that ran in the New York Times earlier this week, Seeking Answers From Green Mountain Coffee, which begins:
Green Mountain Coffee Roasters' first-ever investor day is Tuesday, and the company is flying high.
The stock price of the company, which sells coffee machines under the Keurig brand and the little K-Cups that go in them, has soared more than 260 percent in the last year.
Despite persistent questions, most of Wall Street remains resolutely bullish on Green Mountain, which has a market value of $12 billion.
In 2010, the company disclosed it was being investigated by the Securities and Exchange Commission. In 2011, the hedge fund manager David Einhorn, who is betting against Green Mountain's stock price, delivered a highly critical 110-slide speech at an investor conference, raising questions about the company's future prospects and, more seriously, its bookkeeping. He followed up a year later with another one.
A class-action lawsuit, which was dismissed, quoted anonymous former employees about suspicious activities. Green Mountain has said it conducted an internal investigation that cleared the company.
Green Mountain operates on a razor/razor blade model - selling brewing machines but making its real money on the K-Cups. It used to disclose exactly how many K-Cups it sold but stopped doing so in 2010. Instead, it tells investors the year-over-year percentage growth. Wall Street has dutifully plugged numbers in to estimate the unit sales.
Last year, Green Mountain faced expirations of the patents that covered its brewing system. Wall Street has been monitoring whether Green Mountain will lose market share to new private-label knockoffs. And indeed, a recent Barron's article suggested that it was losing share faster than expected.
A recent disclosure from the company's new chief executive, Brian Kelley, has revived the questions about sales, as do on-the-ground accounts I have received from former factory and warehouse workers.
Because Green Mountain's investor day will give analysts and shareholders unusual access to company executives, it seems like an opportunity to ask them some hard questions.
Here are a few from me.
*Just how many K-Cups has Green Mountain sold year-to-date and is it less than the Street understands?...
*How wide is the gap between how many K-Cups the company says it has sold and how many have ended up in customer's hands? And why?...
*What explains the unusual movements of Green Mountain inventory described by some former company workers and associates?...
*What is happening with the SEC.'s investigation of Green Mountain, which the company has said involves its accounting practices?...
Let's take a closer look at K-Cups, where the math just doesn't make sense - and the company isn't helping with an explanation. At the analyst day earlier this week (see webcast and 188-slide presentation here), the company was asked to reconcile this estimate of K-Cups (since, as Eisinger notes, the company stopped disclosing K-Cup sales in 2010): there are 16 million brewers, GMCR claims usage (an "attachment rate") of 1.4 K-Cups per day x 365 days/year, which results in sales of 8.2 billion K-Cups per year (which doesn't even count maybe 15-20% additional consumption away from home).
But GMCR isn't selling anything close to this number of K-Cups, per both analysts and the company (see Eisinger's article), so what gives? My answer: usage is declining. It makes sense that the people who bought brewers first are likely to be the heaviest users, so the company and analysts should be modeling declining attachment rates - but of course they're not.
When Mark Astrachan, the Stifel Nicolaus' analyst (and the only one with a “sell” rating on the stock), asked about this at the analyst day, Green Mountain’s CEO Brian Kelley said: "We don't multiply it by 365. (inaudible) You're applying straight math that we don't do." (Yes, he really said that!) When Astrachan tried to follow up, Kelley said: "We're not going to get into that here. That's not the intent and we're not going to get into the model in that kind of detail here." (The "detail" he's referring to is the attachment rate of the brewers - one of the most important metrics there is!)
An even greater concern is that 700-900 million K-Cups can't be accounted for. Eisinger writes:
That's a far cry from 5.6 billion. There seems to be a gap in the United States of about 900 million K-Cups.
What's going on?
Mr. Brandt said the company declined to give its overall sales volume, but said the IRI number that I was furnished with was too low. He said a company analysis indicated that this portion of Green Mountain's sales should be about 2.7 billion, not 2.6 billion.
Still, even if we use the company's figure of 2.7 billion, total sales in the United States would be 4.9 billion, or about 700 million K-Cups short of what the company has said. That's a lot of extra K-Cups sitting in the channel.
Maybe I'm just being paranoid, but I've seen this kind of thing before: in many of the China frauds, companies were booking fake sales, resulting in fake profits. But that leads to a big problem for the companies: it's hard to fake all the cash that should be in the bank as a result of the supposed profits. The solution? Make overpriced/fraudulent acquisitions and/or fake excessive cap ex to reduce the cash (that was, of course, never there).
Now go back and read David Einhorn’s 110-slide presentation on GMCR at the Value Investing Congress on Oct. 17, 2011 (posted here) and look at the high-priced acquisitions on pages 50-53 and pages 68-72 on cap ex. Einhorn estimated that $431 million (58%) of GMCR's 2012 cap ex was "unexplained" and concluded:
- Capital spending is growing much faster than the business
- Capital intensity should be getting more efficient as the company achieves scale
- The gap is so large and insufficiently explained that it raises questions about what is being capitalized and casts doubt on the business model
Einhorn gave an update on GMCR in his presentation at the Value Investing Congress on Oct. 2, 2012. He didn't release the slides, but here are my notes:
GMCR's cap ex as a percentage of sales was 11.0% in 2011, 13.1% (est.) in 2012, and 9.2% (guidance) in 2013. Compare this to the 3.3% average in the food products industry, with a range of 1.0% to 6.3%.
GMCR's cumulative cap ex from 2007-2012 was $1.043 billion and K-Cup shipments in 2012 were 7.1 billion. Divide these two and you get 14.7 cents of cap ex over six years for each K-Cup produced in 2012. Einhorn compared this to a competitor, which spent 3.8 cents for each K-Cup produced (buying the same production equipment as GMCR). Again, massive unexplained cap ex.
Einhorn then turned to the production capacity that GMCR's competitors were bringing online and estimated that they would have enough capacity to take 19% market share by the end of 2012 and 26% by the end of 2013.
Lastly, Einhorn showed that competitors were already selling K-Cups for 22-39% less than GMCR was, and highlighted price cuts GMCR had taken that would wipe out nearly all of its profit.
(Obviously these last two things haven't occurred yet - but that doesn't mean they won't…)
Is GMCR committing massive accounting fraud? I don't know - and I certainly can't prove it - but there are a number of warning flags, so I sure can't rule it out. The company could easily put a lot of these concerns to rest by providing some obvious disclosure - like number of K-Cups sold - but refuses to (despite providing highly granular disclosure on most other matters), which makes me all the more suspicious…
The nice thing about GMCR as a short is that I think it's a good one even if its accounting is clean because of its very high valuation (29x trailing EPS and 22x FYE 9/14 estimates (if you believe them)) combined with its patent loss a year ago, which is resulting in a ton of low-cost competition entering the market (see page 44-48 of Einhorn's 2011 presentation and my notes from his 2012 presentation above).
It's almost never pretty when a company with a monopoly market share and monopolistic pricing begins to face competition from low-cost generic producers (think what happens when a drug goes off patent) - but it can take some time for the competition to emerge and impact the monopolist's financials, during which time the monopolist can give whatever guidance it wants (and you can be sure that Wall St. "analysts" won't question the pie-in-the-sky guidance). Witness this week’s analyst day…