Green Mountain Coffee Roasters (GMCR) has provided investors, both long and short, with quite a ride over the past year or so. Emerging from an extended hyper-growth period, investors have been anxiously waiting to see what is next for the company. Will they still manage to produce growth at a more modest, sustainable level, or contract into irrelevance? Based on their positive quarterly/annual report in November 2012 and recent initiatives, it is clear that GMCR is responding to their circumstances, so what can investors expect? In this article, I attempt to gather information from a variety of sources to provide some insight into what we might see in the next year or two.
GMCR 1.0 and 2.0
Green Mountain Coffee Roasters started life as a premium coffee shop in Waterbury, VT in 1981. The coffee that they roasted for their cafe acquired a reputation and following, and the company branched out to providing bagged coffee via a variety of retail and wholesale outlets. The company didn't achieve national exposure during this phase, and was little known outside the northeast US. Even though GMCR 1.0 didn't provide a particularly interesting model, their products were popular and provided enough of a growth rate for their groundbreaking transition. They made their first K-Cups during this period, providing manufacturing and distribution in partnership with Keurig.
In 1996, GMCR acquired a minority stake in Keurig. While this was the catalyst for the single-serve growth to follow, GMCR 2.0 really began when they acquired the rest of Keurig ten years later. GMCR's strategy was to sell the brewers with little or no profit, and make their money on sales of the patent-protected K-Cups. Through a series of partnerships and acquisitions, Green Mountain ended up with over 200 owned or licensed varieties of K-Cups over the next several years. This razor/blade model provided spectacular growth as the Keurig brewers emerged from anonymity to ubiquity, garnering an overwhelming share of the North American single serve coffee/tea market.
The GMCR 2.0 era started to end in October 2011 with David Einhorn's publication of his short thesis, and culminated with the collapse of share price in May 2012 with a poor earnings report and lowered outlook. After that report, share price dropped by half almost immediately and by two thirds within six weeks. The investing press piled on, and the future looked very tentative. The stock bottomed out around $17 per share, almost $100 below the peak less than a year earlier.
The advent of GMCR 3.0 occurred in August 2012. Share price had dropped near book value, and shorts were having a field day. Then, in their FY12 third quarter report, GMCR simultaneously lowered estimates and announced a $500M stock buyback. Remembering prior years when they beat estimates on a regular basis, it appears they set the bar pretty low for themselves, conservatively predicting very modest growth for the remainder of FY12 and the upcoming FY13. Normally, this would have meant another big payday for the shorts, but the announcement of the buyback spurred a spike in share price. The company basically told the shorts that they'd had their fun, but GMCR had decided to put a floor under the company's valuation.
The next quarter passed without much movement in share price, other than normal for this high-beta issue. Then in November 2012, GMCR beat their modest estimates for the fourth quarter of FY12, and slightly raised estimates for 2013. And they reported positive cash flow for FY12, where they had previously estimated a negative cash flow for the year. Share price skyrocketed immediately to the upper 30s on high volume, and has risen to the 40s as of this writing. After wild gyrations in both directions, GMCR ended 2012 about where it began, losing about 8 percent over the course of the year.
Estimating GMCR 3.0's FY13 Numbers
So how can we determine if the positive fiscal 2012 fourth quarter report was just a one-quarter wonder, or if GMCR 3.0 will produce more beats throughout FY13? The key statistic in predicting GMCR earnings is the size of the installed base of brewers. While the sales of those brewers don't have a big impact on the bottom line (being sold at or near cost), sales of the profitable K-Cup portion packs scale with the size of the installed base. The brewers are durable items, and continue to generate revenue for GMCR long after the original sale. Thus even flat (or lower) brewer sales compared to FY12 would still result in an increased installed base, and increased K-Cup sales and profits.
Here are some interesting numbers that GMCR has shared:
- As of March 2012 (around the end of the second fiscal quarter) the estimated installed base of Keurig (and partner) brewers in the US was 10.8M-12.2M brewers. Let's assume the middle of the range, or 11.5M brewers.
- A total of 3.5M brewers were sold in fiscal quarters 3 and 4 of 2012. The installed base would not go up by this number, since some brewers in the existing base were undoubtedly retired due to breakage or other factors. If we assume 2.5M of those brewers went to increasing the installed base, there were around 14M by the end of FY12.
- CEO Larry Blanford stated in the earnings call in November 2012 that they expect an installed base of 17M brewers by the end of FY13.
- Assuming K-Cup sales scale with the installed brewer base, a FY13 increase of 3M brewers (from 14M to 17M) would be just over a 20% increase in the installed base, which is consistent with GMCR's 13-20% estimated growth for FY13.
However, I have a real problem with these numbers. GMCR seems to be estimating that the installed base of brewers will only increase by 3M brewers in all of FY13. However, last year, they increased the installed base by more than 3M brewers in the holiday quarter alone. For GMCR to achieve an installed base of only 17M brewers by the end of FY13, they would basically need brewer sales to drop to roughly half of number they sold in FY12. Does this seem likely? Well consider these other data:
- The installed Keurig and partner brewer base in the US is estimated to grow by 8M brewers in FY14, and an average of 5M brewers in each of FY15 and FY16.
- As reported in their last earnings call, NPD data for October 2012 (the first month of FY13) showed an increase of 90% in brewer unit sales over the same period in FY12. Not the 50% decrease that GMCR seems to be estimating.
Given these numbers, it is hard to imagine the installed base being any less than 20M by the end of FY13, which would represent an increase of about 6M brewers. This would be much more consistent with FY12 sales and company-provided estimates for FY14 through FY16. The 20M level would still represent a decrease in total brewer sales for FY13 compared with FY12, but would increase the installed base by over 40%, rather than the 20% in GMCR's gloomy estimate.
So is competition making GMCR lower their K-Cup prices? Nielsen data from December 2012 indicates the average price of a K-Cup actually rose 0.6% over the previous year. This report also shows a 39% year-over-year increase in brewer sales for four weeks in December, further supporting the thesis that GMCR is grossly underestimating FY13 brewer sales. What about production costs - could green coffee prices cause lower K-Cup profits? Well, coffee prices have pretty much collapsed over the past year, and GMCR says their production costs will drop in FY13 as they utilize the last of the higher priced beans in their inventory.
If these numbers are correct, even a very pessimistic interpretation of data from GMCR and the other cited sources seem to indicate an EPS increase for FY13 that is well beyond the official estimate of 13-18% (even if GMCR doesn't buy any more shares back). All things considered, it appears that GMCR 3.0 is interested in going back to the days when they underestimated and overdelivered.
GMCR 3.0 K-Cup Competition
Recent reports indicate that K-Cup competition isn't materializing very quickly. Patent expiration dates were well known - if the short thesis was correct, competitors would have been tooling up for a year before patent expiration, and flooded the market with low cost alternatives the day the patents expired. Clearly, this didn't happen. The effect is well-illustrated with a tale of two warehouse clubs.
Sam's and Costco (COST) are two huge warehouse club chains. They each sell club-branded portion packs for use in Keurig brewers. Costco is a licensee of GMCR, and their K-Cups benefit from the GMCR connection, as seen by the positive reviews posted on the Costco web site. With full access to GMCR's impressive stable of patents, Costco's Kirkland K-Cups don't seem to have any mechanical problems at all. Sam's, on the other hand, is not a GMCR licensee, and they can't even legally call their portion packs K-Cups. You can read the reviews of their knock-offs here, and it is not a pretty sight. There are page after page of angry reviews regarding malfunctioning Sam's portion packs. Other unlicensed knock-offs seem to be having similar issues.
Instant coffee-filled and open-bottom portion packs (designed to not infringe on the recently expired patents) have been around for a while, and haven't exactly taken over the K-Cup market. Sam's knock-offs are some of the first attempts at "real" K-Cup clones - that is, portion packs that have an internal filter and ground coffee (not instant coffee), and have a bottom that is enclosed for freshness. Some possible reasons for the problems:
- Real K-Cups with filters and enclosed bottoms are not cheap to make. GMCR had already perfected the processes and invested the CapEx to establish the ability to inexpensively make billions of K-Cups per year. It is hard to envision anyone competing with GMCR any time soon on price, without taking manufacturing shortcuts.
- Real K-Cups are hard to make correctly. Manufacturing a nitrogen and coffee-filled capsule that works correctly in the tight tolerances of the Keurig brewers is not easy. GMCR has a pretty serious home field advantage here.
- As I previously noted here, the patents that expired are only a small part of GMCR's K-Cup patent portfolio. Although it hasn't received much notice, US patent 6,440,256 protects some of GMCR's manufacturing processes until 2020. It even covers how GMCR seals the lid and filter to the plastic cup, which may explain some of the problems Sam's is experiencing.
It is also worth noting that GMCR 3.0 has established a powerful distribution channel. K-Cups and Keurig brewers are everywhere. I see them in pretty much every grocery store, department store, electronics store, hardware store, home improvement center, wholesale club, and retail superstore. GMCR estimated in their last earnings call that unlicensed brands won't exceed 5% of the market in FY13, and will top out around 15% in the next few years.
GMCR 3.0 Brewer Competition
The Keurig brewers have achieved overwhelming market share over the past several years. There were a lot of horses in that race over the last decade, and GMCR won the battle for consumers' counters and grocers' shelves. Single serve brewer competition isn't anything new to GMCR. Existing and past competitors include:
- Kraft (KRFT) Tassimo - formerly in a deal with Starbucks to distribute coffee pods.
- Nestle (NSRGY.PK) Nespresso - very popular in Europe, but haven't had as much success in the US.
- Mars Flavia - been around for many years, but not making a big splash with their home brewers.
- Sara Lee Senseo - once a major factor, they announced they have pretty much given up on the US single serve market.
- Starbucks (SBUX) Verismo - a recent entry into the single serve marketplace.
The relationship between GMCR and Starbucks warrants discussion. Starbucks abandoned the Tassimo platform (which earned them a "billion dollar brouhaha" with Kraft), and a few months later, made a deal with GMCR to become a K-Cup licensee. This deal has been very profitable for both companies, and finally provided Starbucks with the market-leading platform they had been looking for. Starbucks K-Cups remain very popular, and are sold both in Starbucks cafes, and a variety of other outlets in GMCR's retail channel. Exactly how much of the K-Cup market share Starbucks controls is an interesting question. Starbucks CEO Howard Schultz claimed recently that Starbucks controls about a quarter of the K-Cup market. Apparently, GMCR CEO Larry Blanford didn't agree with this pronouncement, and he stated in the November 2012 earnings call (linked above):
In addition, in our fiscal fourth quarter and for all of fiscal 2012, no single partner or non-owned brand represented more than roughly 6% of our net sales. In fact, our largest contributing partner brand Newman's Own Organics is one with whom we've had a longstanding relationship and has been a part of the Keurig system for six years.
I can't reconcile their comments, as Blanford seems to state that Starbucks isn't even the most popular non-GMCR-owned brand, and controls far less of the K-Cup market than Schultz claims. It seems logical that GMCR's Blanford is in a much better position to know how many K-Cups of each brand are produced than Schultz. In the same report in November 2012, Blanford also dismissed claims by some analysts that Starbucks K-Cup license with GMCR was coming to an end soon:
All of our most significant partner brand contracts are multiyear in nature, even from this point in time.
Which brings us to Starbuck's new single-serve brewer, the Verismo. German company Krueger GmbH does the heavy lifting for this brewer, which is visually similar to a variety of other re-branded Krueger "K-Fee" brewers. These include the Paulig Cupsolo in Finland, the Krueger Preferenza in Germany, the Aldi Expressi in Australia, and the Tesco Podpronto in England. With its wide variety of pods for espresso-based beverages, the Verismo competes most directly with GMCR's new Rivo brewer. GMCR developed the Rivo jointly with Italy's Lavazza, who must have been pretty happy with the finished product, as they increased their ownership of GMCR shortly before the Rivo was released.
The jury is still out on the potential success of Starbucks' Verismo. They dropped the brewer prices significantly during the holiday shopping season, and some writers are calling the Verismo a "flop" and a "blunder." Starbucks reported last week that they sold only 150,000 Verismos in the last four months. We won't know how many Keurig brewers were sold during the holiday quarter until next week, but GMCR sold 4.2M of them during the holiday quarter a year ago.
GMCR had a limited roll-out of their Rivo system during the holidays, with the brewer available only from Bloomingdales. Allocations sold out as quickly as they arrived at the $229 MSRP, and the Rivo has been on and off back order since its release. Entrepreneurs take advantage of the scarcity, re-selling Rivos on eBay (EBAY) for $50 over that price. A wider roll-out, that should end the price gouging, is planned in the near future.
With their ever-expanding variety of coffee and espresso brewers, protected by dozens of company-owned patents, GMCR 3.0 is still the 800 pound gorilla of the single serve brewer market in the US.
GMCR 3.0 International
The Keurig brewers that most of us are familiar with are only available in the US and Canada. GMCR does have a long-standing joint venture with Ueshima Coffee Co. (UCC) to market the Keurig system in Japan, China, Taiwan, Korea, Singapore, Malaysia, Thailand, Indonesia, Philippines, and Vietnam. The Keurig Trevie brewer (enjoy the translation) and a variety of UCC K-Cups are currently available in Japan. You can see here that GMCR has booked income from this joint venture. The push for GMCR outside of North America has been pretty limited to date; however there are a few recent signs that indicate this could be changing soon.
First, GMCR moved Gerard Geoffrion to GMCR President, International Business Development, to "lead the exploration of business opportunities outside North America." Geoffrion previously served as the lead for GMCR Canada Holding, Inc. and President, Canada Business Unit. He seems a logical choice, as GMCR is very strong in Canada and acquired long-time Canadian licensee Van Houtte a couple years ago. One interesting line from the GMCR's 8-K filing on Geoffrion's new assignment is that his compensation package includes the potential for "bonus compensation based upon the Company achieving certain operational and financial goals." In other words, something seems to be brewing in the not too distant future.
A couple of recent partnerships with strong international partners may provide clues for the directions that GMCR 3.0 is considering. The company recently started selling Eight O'Clock Coffee K-Cups, and has plans to introduce Tetley and Good Earth Tea K-Cups this year. These brands are owned by global conglomerate Tata (based in India), with over $100B in annual sales. The Tata Global Beverages subsidiary is the second-largest tea company in the world, operating in over 40 countries. With this new partnership, Tata seems a logical partner with whom to pursue international expansion, especially with the tea business.
As mentioned above, GMCR also established a partnership with Lavazza, with the Italian coffee giant acquiring a large ownership position in GMCR. When they announced this partnership in 2010, they estimated that a jointly-developed espresso machine would be marketed in North America in fiscal 2013, and they hit that deadline perfectly. The Rivo machine looks like it was designed to compete squarely with European single serve espresso machines, and it wouldn't surprise me a bit to see a Rivo-based machine showing up on store shelves in Europe. Lavazza CEO Gaetano Mele stated "We are confident that this investment is but the first step in a wider-ranging collaboration with GMCR in R&D, innovative technology and international expansion for both companies."
It appears that GMCR 3.0 is poised to expand overseas, and the Gerard Geoffrion, head of GMCR's new International Business Development unit, already has some chips on the table as the company gears up for expanding into new markets.
GMCR 3.0's Carbonation Connection
Over the past few years, GMCR has continued to expand their product choices beyond their coffee origins. Tea, Hot Cocoa, Lemonade, and "Wellness Brewed" Vitamin Burst beverages are already sold in K-Cups. Many of these beverages bear GMCR's "Brew Over Ice" logo. A recent patent filing points to an entirely new direction for GMCR 3.0 - countertop carbonated beverages. This filing has no fewer than 90 individual claims included, and would revolutionize the home carbonated beverage market. SodaStream (SODA) is currently the market leader in this domain, but their US market penetration has been modest in comparison to Keurig brewers. For example, SodaStream analyst Seth Golden from Capital Ladders, recently stated:
Based on 2,619 stores reported to Capital Ladder Advisory Group, during the month of September, Wal-Mart (WMT) has sold an average of 1.8 Soda Jet kits per week/door (Wal-Mart sells SodaStream products in roughly 3,300 stores in North America).
Less than two SodaStream machines leaving the average Wal-Mart every week seems pretty modest market penetration. Though I am pretty sure the recent holiday gadget gift season resulted in much higher sell-through than Mr. Golden's earlier estimate. While the SodaStream machines have been pretty well received (with an average 4-star rating on Amazon (AMZN)), the negative comments tend to focus on the need to exchange the CO2 cartridges either on-line or by going to a local retailer. This inconvenience is likely one of the barriers to wider adoption, especially with convenience-oriented US consumers.
The machine described in GMCR's new patent filing is fundamentally different from SodaStream, and follows the same paradigm as the existing Keurig brewers. The revolutionary aspect (when compared with SodaStream) is that, like the current Keurig brewers, you "just add water." Both the carbonation source and beverage flavoring are in disposable K-Cup-like portion packs, so there is no need to return canisters to the vendor for CO2 recharging. The drawings in the filing show a variety of possible arrangements and orientations, some with separate cartridges for the flavoring and carbonation, others with everything combined into a single cartridge.
It is impossible to ascertain from this patent filing just how close to a carbonated beverage product GMCR is at this point. One would imagine that they have prototyped some or all of the concepts, in preparation for filing for the patent. But with GMCR's existing US distribution channel, and history of taking products from the niche into the mainstream, I would imagine that SodaStream is very concerned about GMCR 3.0 doing the same thing to the soda aisle that they have already done to the coffee aisle.
Branding is another interesting aspect of the current SodaStream lineup. If you look at their beverage flavor options, there are a few familiar brands, namely Country Time Lemonade, Crystal Light, and Kool-Aid. This is a somewhat curious selection though, as these are not normally carbonated beverages. Their traditional soda flavors (cola, root beer, etc) are all house brands, some with clever names like "Dr. Pete." GMCR K-Cups, on the other hand, are pretty much a "Who's Who" of coffee brands, a result of broad-ranging partnerships, license agreements, and acquisitions. It is notable that GMCR already has an agreement in place with the Dr Pepper Snapple Group (DPS). And perhaps there is more than meets the eye to luring rising star Brian Kelley away from Coca-Cola (KO) to be GMCR's new CEO.
If I may be so bold as to enter into a moment of pure speculation - no links for verification as with the preceding paragraphs - I would like to hypothesize a new product for GMCR 3.0. Given the obvious similarities between the current Keurig brewers and the machine described in GMCR's new filing, why not build a combined machine to minimize both consumer buy-in costs and the kitchen counter footprint required for beverage machines? This hypothetical machine would have a single tap water reservoir to brew coffee and create your favorite carbonated beverages. It would have a single "head" to produce both beverages, maybe leaving one slot empty for brewing coffee if the engineers need to go to separate cartridges for CO2 and soda flavoring. I would further speculate that this new machine would brew Keurig Vue Packs, rather than K-Cups, to drive further adoption of their better patent protected Vue technology. I have no reason to believe such a machine is in the works, but it makes so much sense, I would be surprised to learn it is not.
After what has been a harrowing last year for the company, GMCR's reboot a few months ago appears to be paying off. They moderated investor growth expectations and implemented a buyback to stabilize share price. They managed to get their capital expenditures under control, and their over-spending in 2011 and 2012 will serve to enhance cash flow as their continued growth absorbs excess capacity. And the stinging response of the market to their misses and lowered estimates in 2012 seems to have driven them to ultra-conservative growth estimates for FY13.
Based on the information aggregated in this article, the GMCR 3.0 reboot seems targeted in three directions:
- Continuing to grow their cash-cow single-serve US coffee business;
- Expand overseas to provide additional growth; and
- Expand their portfolio of products beyond the coffee/tea business.
Competition, both in brewers and portion packs, has been much slower to emerge than critics have predicted. Factor in the international expansion that GMCR clearly plans, along with a potential push into the home carbonated beverage domain, and the company's longer-term growth appears solid. With raw commodity coffee prices down about half from the highs of the past couple years, GMCR 3.0 should provide some highly caffeinated earnings throughout FY13.
If you make investing decisions based solely on the advice of strangers, you will probably go broke. Research the information presented above, spend some quality time with Google, and draw your own conclusions.
Disclosure: I am long GMCR.