Giving GMCR the Thumbs Down
Green Mountain Coffee Roasters (Nasdaq: GMCR) continues to percolate, emitting the sweet smell of stock market returns. So I'd have to be a little crazy to consider giving it a "Thumbs Down" in CAPS, right?
I will make the case for "Thumbs Down." But that's only half the battle. The other part of the equation is picking the right price. After all, there's risk in every investment decision so I'd better be sure to invest at a price that puts the odds squarely in my favor.
So let me walk you through my investment thesis, the risks, and at the end I will give you my estimate of the price where I think Green Mountain would be a solid "Thumbs Down" decision.
But before I do, if you like the analysis, please pass it along to others to see. And please notice I didn't say agree with. You can like the thought process but not agree with the conclusion. Please let me know your thoughts in the comments.
Here are the main points of my short thesis (the supporting data will follow):
1. K-Cup usage per day is declining
2. EBITDA/K-Cup is flat (maybe declining slightly)
3. The market is paying 3 times the EBITDA multiple over a 12-18 months ago for no relative improvement (I'll discuss the absolute rise in EBITDA)
4. Despite over 4.2 billion (yes, with a "b") K-Cups sold and counting, the company hardly generates any cash flow and has almost no cash on its balance sheet.
5. The dramatic rise in price comes at a time when short interest declined by about 50%
And here's the on-coming train that I could be stepping in front of, namely, the risks:
1. The growth rates for sales and profits are unbelievable as the company has done a fantastic job of transforming its business from a roaster to a "coffee standard." Kudos to management!
2. Shorts could continue capitulate, which could drive prices higher.
3. Did I mention the growth rates are unbelievable? And that the market loves growth? If the growth period is longer than expected, the stock could move higher.
4. The market may not have gotten too carried away and discounted too much performance forward. Profits could simply fill in over time, causing the price to remain flat rather than drop.
This isn't a slam-dunk short at prices under $70. As promised, here's a stroll through my analysis in pictures and words, along with a price that should compensate me for taking the risk of perhaps being ground to bits.
I love pictures
Green Mountain has gone through an impressive strategic transformation. Rather than remaining a coffee roaster, it purchased Keurig to become a standard in at-home coffee consumption. I cannot say it enough -- this was a great move by management.
I won’t go into any more details about the company. You can read more about GMCR at its website and in its 10-K. (Give links to both.)
Sufficed to say, the Keurig brewer and the K-Cup are everything to Green Mountain. It does have a roasting segment that feeds the K-Cups, but K-Cups are the most important part of the company
Brewer sales (units)
There is no doubt that the company has been selling lots of brewers, with nearly 90% of them serving the At Home market. Coffee drinkers are buying them and they make great gifts!
With about 3.6 million sold, Keurig brewers continue to find their way into homes and businesses. GMCR has done a great job of selling these and there is room to sell more over time.
K-Cup sales (units)
With all those brewers, is it any wonder that this curve continues to climb skyward, eclipsing 4 billion K-Cups sold with little resistance? Kudos to management for selling lots of caffeine-delivery mechanisms and the caffeine itself. That's a strong business model!
Average price of a brewer
Average price of a brewer: prices seem to be between $90 and $250.
The brewer is the enabler. And GMCR, according to its 2008 10-K and 2009 10-Qs, does not make any gross profits from these.
Range of prices for K-Cups
From the example above, we see:
$125.99/72 = $1.75 per cup for the Columbian Fair Trade K-Cup. This is probably the most extreme example on the expensive side.
For the Hazlenut K-Cup:
$50/72 = $0.69 per K-Cup, or nearly 1/3 the price of the Columbian Fair Trade K-Cup above.
By the way, there's tons of variety -- a very smart move by GMCR was to co-brand and offer more than just coffee.
Let's assume the average cost of the brewer is $150 and the average cup of coffee from the brewer costs $0.75. If the average cup of coffee outside the home (think Starbucks et al) is $3.00 (it may be more), then we can determine the breakeven point between brewing coffee at home and buying it away from home.
$150 = x ($3.00 - $0.75)
x = $150/$2.25 = 67 cups of coffee to breakeven (if the average cost of coffee goes up, the cups to breakeven goes down)
For a serious coffee drinker, this is nothing. This adds to the value proposition of the Keurig system.
Let's take a different look at the data
Again, the sales of brewers and K-Cups continues on a torrid pace: brewers unit sales have been averaging ~100% annual growth and K-Cup unit sales have been averaging ~50% growth. Hmmm … that's an interesting difference. Let's look at the trend of this ratio and think about what it might mean.
Why would the ratio of K-Cups to brewers be trending down on a quarterly basis as well as on a cumulative basis?
1. Higher bases to work from (stupid math - this joke only works if you can say this is your best Bill Mann voice)
2. Already captured the "early adopters" and passionate coffee drinkers?
3. Easier to open new sales channels for brewers than K-Cups?
4. Word of mouth is driving increased brewer sales?
The trend appears to be leveling off on a cumulative basis (red lines, T is for total). There is no way this trend go to zero, but I do wonder if the this chart is an indication that the company has already reached it heavy coffee drinking audience and is now going after the more marginal coffee drinkers.
How much coffee does 1 person need per day?
The purpose of this chart is to try and get an idea of how much coffee each brewer is making per day. The blue area is cumulative K-Cup unit sales (left axis) and the red line is the K-Cups per Brewer per day (right axis). Implied in the calculation are that every K-Cup gets used and that all brewers sold are still in action. Neither of those two assumptions may be correct and thus the consumption ratio that I calculate is likely higher than I show.
Despite the imprecise calculation, I think the analysis makes an important statement. According to a few of coffee sites, the average coffee drinker consumes about 3 cups per day. (Sources: http://www.coffeeloversusa.com/statistics.html and http://www.buckscountycoffee.com/aboutus/media/trends.html and http://www.specialty-coffee.com/ME2/Apps/PublishingMultiselect/Print.asp?Module=PublishingTitles&id=D8C8179DF2DA478D9FED99AF67566735) Admittedly, some of the data is old. But consumption seems to be hanging around 3 cups per day.
It looks as though K-Cup consumption has dropped to (or may even be below) about 3 cups per day. Could that be an indication K-Cups have some faddish qualities? It's possible. I think this is an important number to watch. If this number continues to drop, it would put a ceiling on the growth period for this company and could impact the recurring revenue part of the equation, too.
BTW, the cynic in me checked to see if the company was stuffing the channels and a quick look at Days Sales Outstanding and Inventory turns shows GMCR is not doing this and the trends for inventory and receivables are pretty constant.
There are two ways to look at profits
Take aways from the chart:
1. Consistent profitability on an EBITDA/K-Cup basis -- maybe declining slightly.
2. Growing profitability on a Net Income/K-Cup basis.
3. Lumpy (declining?) profitability on a Operating Cash Flow to K-Cup basis.
4. FCF/K-Cup has turned negative again.
With over 4.2 billion (with a b) K-Cups sold through 2Q2009, why isn't cash flow rising more evenly and why hasn't free cash flow turned considerably higher? Even though GMCR is still in high growth mode, it seems as though its K-Cup success should have turned into strong cash flow generation.
It is encouraging that the EBITDA/K-Cup is steady as that means management remains focused on costs and not just growth. But it also is indicating that GMCR is not generating any economies of scale over time. GMCR has been seeing its gross margin decrease as it saturates the market with brewers. However, its SG&A costs have declined, too. It seems as though the cost of making a billions of K-Cups should be going down on a per unit basis. Still, consistent performance is good, too.
GMCR must reinvest in inventory and accounts receivable in order to fund its growth. On the one hand, the reason why is simple: the company must fuel its growth as it wants to capture market share. But at some point, the growth should generate excess cash flow. Returns on invested capital continue to rise (CapitalIQ shows ROIC rose from 9.3% for FY2003 to 14.6% over the last twelve months), but maybe not fast enough to generate excess cash flow yet. It would be very bad to see a declining EBITDA/K-Cup performance curve.
Just how valuable are those K-Cups
Let's recap before thinking about this chart. GMCR is growing in leaps and bounds. It continues to open new sales channels for its brewers and K-Cups and customers continue to purchase them. While we've seen performance, normalized by K-Cups, (the foundation of GMCR's business) is flat to slightly declining on an EBITDA basis and rising on a net income basis. However, cash flows are lumpy and free cash flow is non-existent.
The chart above looks at GMCR's market cap relative to K-Cups sold over the previous 12-month period. The market has given GMCR a wide range of values over time, but the recent rise is very clear -- and breaking through boundaries.
Does this make sense? Why the sudden revaluation of the company on a K-Cup basis when the performance per K-Cup is not really improving? The first place I stopped was the short-interest data.
Can you blame them?
This chart shows a near 50% reduction in the short position followed by a huge run by the stock price. Have the short capitulated and the had to close out their positions (alright I'll say it -- a short squeeze)?
Can you blame them? After all, as I keep repeating, GMCR continues to grow at an incredible pace. But it is hard not to ignore this data. The short trade is considerably less crowded and didn't work for a long time.
The company does not generate any free cash flow, so I'll have to use a different tool in my valuation toolbox.
The market is offering you to buy shares of GMCR at an EV/EBITDA multiple of about 30 (about a 3.3% yield) and a Price/Operating Cash Flow (remember, this would be before reinvestment requirements for growth) of almost 50! That's a 2% yield! That doesn't seem like a very good return for the risks that an investor would have to take.
In addition, we see that the market is offering a price of $0.52 per K-Cup sold. The average is $0.31. There are two ways the current value can moved back towards the average: the price can drop or the performance can fill in. Let's assume performance fills in and see the implications for K-Cup sales required.
Current multiple: $0.52
Average multiple: $0.31
K-Cups sold: 4.25 billion
K-Cups sold required to justify the current price: 0.52 * 4.25 / 0.31 = 7.13 billion
The company has to sell an additional 7.13 - 4.25 = 2.88 billion K-Cups to justify its current price based on historical figures (And yes I know I am looking in the rearview mirror. There's no need to remind me.). GMCR has sold about 1.4 billion K-Cups over the last twelve months. That means two years worth of futures sales has already been discounted into the price of the stock. The odds of GMCR suddenly selling almost 3 billion more K-Cups over the next 12 months seems low. However, if growth continues, it could take about 1.25 years to meet those expectations.
During that time period, what's likely to happen to the stock price? The most likely outcome is to remain flat. If the company does not meet those higher expectations, the value of the company could fall.
As you can see from the data below, GMCR's ratios have all nearly doubled in a rather short period of time.
Has the price risen too high relative to the fundamentals of the company? Maybe, maybe not. But I do want to find the price that says yes and offers a potentially profitable short.
I can't just look at the relative data. I have to consider the absolute data, too. After all, if the company is selling product that people are using less, it is still selling product and generating profit. It may not last as long as people think, but financial performance is still there.
We must remember our Keynes
Every investing decision means taking some risk. The idea is to take as little as possible for the highest return possible. On the short side, we must remember our Keynes: "The market can stay irrational longer than you can stay solvent." That's what I'm up against with GMCR, especially considering the performance of the company.
However, there's a considerable amount of performance already brewed into the stock price. And as the price rises, the incremental K-Cup sales required to justify the price gets higher and higher. And the effect is to pull more and more sales forward into the valuation.
The higher the expectations rise, the hard to meet they become. At $80, the market seems to be pricing in an additional 5900 million K-Cup sales. Based on the current ratio of 1191 K-Cup/Brewer sales, that would imply in additional 4.9 million brewers have to be sold. GMCR has sold 3.57 million brewers as of June 31, 2009. So brewer sales would have to more than double again in order to sell that number of incremental K-Cups. While it is possible to increase market share quickly, that seems like too much, too fast based on the market share data from the company below.
From the data above, I would initial at "Thumbs Down" position at $80 or higher. I think that offers about a 30% return based on a reversion back to the mean and I think the probability of a price decline at that point is about 40%. I give the stock staying flat about a 30% chance, a 10% rise a 20% chance, and a 20% rise a 10% chance. Here's what my expected value table looks like.
A 30% decline from $80 would bring the stock price to $56. Assuming a 50% increase in K-Cups sold, that would bring the MC/K-Cup value down to $0.35 per K-Cup, which is closer to the average. The expected value is positive, but it is not extremely high at 8%. As such, the higher the stock price, the larger the potential return from the "Thumbs Down" position.