On several occasions I wrote that I have trouble finding new bargains. In the beginning of 2014 however I found something different. Namely a way to gain when the share price of Green Mountain Coffee Roasters Inc. (NASDAQ: GMCR, later in the article: Green Mountain) will have declined: buying Jan 2016 put options. In this article I want to answer the question why I want to ‘short’ Green Mountain.
Green Mountain Coffee Roasters Inc.
Green Mountain produces and sells single-cup coffee makers and produces and sells K-Cups (the cups to put into those coffee makers or brewers), this is comparable to the in the Netherlands more familiair Senseo machine. The K-Cups Green Mountain sells can contain their own coffee brands but they also have several license agreements with for example Starbucks and Dunkin’ Donuts.
The company was introduced to me through the news, after hedgefund manager David Einhorn held the below presentation in 2011. I always like to read the sheets of a presentation like this or something Bill Ackman made about Herbalife. I don’t act on it, first because I should do some work myself and second, because the reaction on the stock market such a presentation creates won’t make it that interesting. But presentations like this are nevertheless very interesting because it can give you an impression on certain companies and on what other investors find interesting. Hedgefund managers also have far more resources than I have, so there can be really valuable information for me in there.
The thing about Green Mountain is that the share price came back to the old levels. So, if I wanted to decide whether the investment thesis of Einhorn was still in tact, I had to do some work. This article is the summary of that work.
Products and Markets
Over the years Green Mountain managed to create a razor / razor blade model comparable to the Gillette model. The coffee brewer (razor) was sold at or near cost and money was made on the K-Cups (razor blade). Unfortunately this business model is something of the past for the company. Although the single-cup serving market has been growing from 23% to 33% of ground coffee sales in Nielsen’s measures for the 14 months through June 8, Green Mountain’s share shrank from 52% to 43%. The reason why this happened and at this speed, is because of the expiration of key patents on the K-Cup in September 2012. Since then, every company can produce the K-Cups that are just as compatible with the brewing machines of Green Mountain as the ones they produce themselves. Companies like private-label supplier Treehouse Foods and North America’s fourth-largest roaster Mother Parker’s Coffee & Tea did do just that and threaten the razor / razor blade model because of the far lower prices they ask for K-Cups.
Although Green Mountain didn’t do much to ease that threat in the years before the expiration of their patents, they did do something. In 2012 the Vue coffee system was introduced in an attempt to sustain the business model. But instead of only selling the Vue system, they kept selling their normal Keurig brewers alongside it. Beside that those brewers weren’t only much cheaper, they also could brew far more coffee brands than the newly introduced Vue system. The Vue system therefore flopped.
So, the market for K-Cups was wide open. Green Mountain has, however, a few large brands licenses with for example Starbucks and Dunkin’ Donuts. Those companies were probably attracted to the large position of Green Mountain, a large majority of the single-cup coffee market was after all in hands of the company. By licensing their brands to Green Mountain, their coffee got available to a lot more potential customers. The problem for Green Mountain is now however that Starbucks and Dunkin’ Donuts don’t need Green Mountain anymore to get into reach of those customers. Private-label producers can now also make their K-Cups and maybe for far less than Green Mountain as well. Even if Green Mountain manages to keep their licensed partners, the margin they make on them might get under a lot of pressure. And if this is not enough, there is also the possibility that a large brand will exit the K-Cup market and will introduce their own brewing system. Starbucks did already introduce their own system but continues to use the K-Cups for now as well.
‘Luckily’, Green Mountain has more plans to spur further growth. First, although the Vue was a flop, Green Mountain plans to introduce Keurig 2.0 during 2014. I can hardly imagine a different fate for it. Second, they want to introduce a cold beverage platform. This platform would also be a single-serve system, which rules customized cold beverages probably out. Further, by going into the cold beverage market you will have to compete with companies like Coca-Cola and Pepsi on the high-end of the market, and private-label brands on the low-end. But what brand does Green Mountain have to compete with them, and ask a price that might be higher, and maybe must be higher because of the costs that accompanies single-serve drinks? They don’t have cold-beverage brands like they have coffee brands, brands that some people probably drink since they started drinking coffee. It would be very hard to win in that market. (Source)
Now that we’ve talked about Green Mountains’ deteriorating market position and their efforts to develop new markets, I want to talk about something more disturbing that came up in a lawsuit called Horowitz vs. Green Mountain Coffee Roasters (GMCR). In that lawsuit are several witnesses cited and I want to cite them again here. According to witnesses:
- GMCR had no warehouse management system, no enterprise resource planning system, no transportation system, no supply chain management system, no inventory controls systems and none of GMCR’s warehouses were linked. A Director of Operations explained once that, instead of implementing these systems, GMCR was going to focus on growing the company through acquisitions. Senior management was not willing to pay for newer computer systems.
- Inventory had been counted multiple times as a result of a virtually non-existent inventory control system at GMCR.
- Production levels were not based upon prior ordering history or inventory levels. Instead, to give the illusion of continued growth to shareholders – and to receive bonuses, which were based upon production – products were knowingly overproduced.
- MBlock (a third party fulfillment company for re-sale of brewers and K-cups to certain retailers) was, in essence, a captive company and would do as GMCR instructed.
- GMCR directed MBlock to hold more product than it could immediately ship and that, due to overstocking at MBlock, product had to be destroyed when it remained in the warehouse past its expiration date. At least one third of MBlock’s warehouse is more than likely expired coffee. Millions of K-Cups worth of products was dumped in land-fills near the Knoxville, Tennessee production plants; the dumped product was then covered with dirt by GMCR employees in order to conceal the destroyed product.
- MBlock used rudimentary spreadsheets to track orders and delivery and that its poor tracking systems allowed GMCR to bury inventory, controlling the process and shuffling items into and out of MBlock’s warehouses.
- GMCR improperly recognized revenue on 150 truck loads of coffee K-Cups that were shipped to MBlock. The requisite paperwork, including purchase orders, material requisition orders, or product shipments authorizations couldn’t be located. A conservative value of this shipment was between $7.5 million and $15 million. No payment was ever made on this shipment but it remained on GMCR’s book as an account receivable.
- GMCR improperly transferred product from one plant to the next for no apparent reason.
- One of the ways to get rid of inventory was to load dated or expired coffee onto trucks and, prior to the arrival of the Company’s auditors, park the trucks a few blocks away from the warehouse. After the auditors left the warehouses, the trucks would return to the warehouse and the products was then returned to inventory. Inventory control employees commented once when inventory was loaded onto trucks: “oh, we must be having an audit”.
- “Product was removed and preloaded on trailer trucks to ship to retailers because we didn’t have room on the floor. Then we’d load more product on trailer trucks to nowhere.”
- A “phantom” shipment of 500,000 brewers that was accounted for as an order purportedly for QVC immediately prior to an audit at MBlock; however, the brewers were never shipped and, following the audit, the inventory was simply restored at MBlock.
In case you like to read the whole document:
Accounting and other issues
The above statements from witnesses really hints at fraud, but there is more. Over the years several accounting issues have been exposed and questioned by a journalist from CNBC and several blogs. They found out that:
- At one of the operations of Green Mountain revenues during their 2010 year was $7,141,000 in Q1, $19,147,000 in Q2 and $2,298,000 in Q3, while income before taxes was $902,000 in Q1, $4,520,000 in Q2 and $4,690,000 in Q3. That means that income before taxes in Q3 is twice the amount of the revenue during that quarter. This would be possible with non-operating income, but non-operating income was for the whole company during those quarters only $27,000. Far short of what it had to be. By the way, this is before a restatement they made in 2010 after an internal investigation. (Source)
- After the internal investigation there was another problem. They had to add for their 2010 year $4.1 million in revenue and $0.0 million of income before income taxes to Q1 and $13 million in revenue and $3.8 million of income before income taxes to Q2. This would be $17.1 million revenue and $3.8 million of income before income taxes in total. In a 10-Q report for the quarter ended March 26, 2011 they report however $2.8 million of income before income taxes. (Source)
- There was also a possibility that Green Mountain was manipulating their earning by having a negative $22.259 million provision for sales. This is rather unusual and could have helped the company beat analysts expectations. Although it is possible that company overstated its sales returns reserve in prior periods and had to adjust that overstatement, the company did not state such thing. (Source)
- After more people found out the above, questions were asked to Green Mountain. The company then disclosed, but not to all (something they should have done with this material non-public information), by saying that the numbers from one period to the next was like comparing “apples to oranges” and that there hadn’t been an adjustment. The thing is, that apart from not giving the consistency that should be given in financial reports, that the numbers still don’t add up. (Source)
- I could continue this for some time but on the White Collar Fraud blog alone are more than 30 articles about the company. They are interesting to read though.
As we’ve seen are there several witnesses stating that the company is a fraud. When confronted by these allegations at the company’s investor day the CEO Brian Kelley told investors that if those allegations were right, then you’d have to believe that he was in on it (Source: David Einhorn’s 2013 Q3 letter to investors). On another occasion, the vice president of corporate communications DuLong was also confronted by the same allegations. She cited that people should look at where the stock price was trading at the point that the lawsuit was filed, and where it’s trading right now, hinting that because the stock price has advanced so much, you should discount the fraud allegations made in the class action complaint. (Source) Making such a non-denial denial as the CEO did and seeing a stock price advance won’t make me believe, though, that the company is not a fraud and the CEO is not involved. Besides this, is the business model of Green Mountain probably history. There are, however, risks involved in the put option investment that I made. It is possible that the company is not a fraud, that the witnesses are lying and that the company will grow further. But it could also be that fraud won’t be discovered or that the company won’t deteriorate until after January 2016. Because of this, I’ve made the position small (2.5% of the portfolio) and is the option not the most aggressive I could have bought (strike price at $50, could have been $30. Break-even is at ±$42.40, while it was ±$27.60 for the $30 put option).