This past weekend at Berkshire Hathaway's Annual Shareholder Meeting, Warren Buffett, when asked about his own investing skills, referred to "pattern recognition," in certain businesses and scenarios. I have identified several interesting cases that are analogous to Chipotle's predicament, as I think it very helpful to look at prior case studies - not food safety incidents (although we'll get to those) but recent cases of "broken growth stocks." In tracking the long term the performance of similar disruptors that had terrifying collapses, analogies are very interesting.
2007: Starbucks (SBUX) became a tired brand and its expansion had led to a decline in quality after Howard Schultz left. Having traded at 50x earnings in 2007 and a 25 billion dollar market cap, the stock plunged from $18.88 a share to around $4.50 (split-adjusted), though this drop was augmented by the crash of 2008. The P/E fell from 50 to 20, and many thought the best days were over.
But Howard Schultz returned, and, while it took a while, his reinvestments in the business eventually reinvigorated the brand. Since January of 2009, the stock is up 15x, for a 46% annualized return.
2011: Netflix (NFLX) had a self-imposed disaster when it simultaneously raised prices significantly, and decided to separate its streaming and DVD business, putting the DVD business on another site called Quickster, and did this as they were losing rights to certain content. The result was a disaster as subscribers fled. The stock, which had traded as high as 6.5 EV/Revenue and a very high PE, crashed roughly 80%, had a bounce, and stayed at that level for a little over a year.
But, Reed Hastings and co. got streaming right, invested in original content, and the company regained subscribers at a rapid clip. The result: Since 2012, Netflix is up 10x, or a 75% annualized return.
2011: Pandora A/S (P), the Danish jewelry maker, had a hit product, their well-crafted charm bracelets, and grew like crazy even through the recession before going public in 2010. But in 2011, silver prices spiked, and the company decided it wanted to raise prices, along with a general strategy to sell more high-end jewelry. The plan totally backfired and its core value-oriented customer fled. Sales crashed, along with the stock from 350 DKK to 40 DKK. It stayed there for about a year. A senior analyst at Danske Bank at the time said that he knew "in the mid-to long run, it couldn't go on forever."
But the company then corrected its price hikes, and implemented a new ERP system to better monitor inventory, and the company recovered. Since October 2011, the stock is up over 15x, or a 120% annualized.
2013: Ulta Salon (ULTA) was a fast-growing, middle-market beauty retailer. As with Starbucks, its rapid unit growth got a little ahead of itself, the experience became less-optimized, and SSS fell in November. The company announced it was going to tone down its torrid unit growth and invest in the customer experience (though still opening a healthy amount of units). The stock price, which was trading at a 44 P/E, plunged from over $130/share to under $85/ share, or ~40%, and stayed there for roughly eight months.
But in late 2014, the investment showed signs of paying off, SSS grew, again, and the stock took off, more than doubling 140% in less than two years.
Why do I believe these are apt parallels? All of these companies shared many key attributes as Chipotle (NYSE:CMG):
1) All were mid-cap, consumer discretionary stocks.
2) All were "affordable luxuries" that people consumed on a regular basis (coffee, TV, makeup, etc., perhaps not Pandora) and affordable enough to have mass appeal.
3) All were "disruptive" companies that were category killers with terrific business models and/or brands.
4) All had high P/E ratios prior to the crisis, but also had high and growing ROICs.
5) All were derailed by operational mistakes or poor decisions that lost the trust of customers.
6) As "high-flyers," all had their skeptics that were very quantitative/ deep value oriented, all of whom "seemed" right at the time. Many analysts thought these problems were permanent, rather than temporary.
7) But all the companies had talented managers, most of whom were founders that had created the great company to begin with, who made prudent investments to regain trust.
8) The results: Some of the best investing opportunities of the past 10 years.
So now we have Chipotle: A high growth consumer discretionary stock, with a disruptive, "category killer" business model, run by its original founder and highly regarded management team, that was trading at a high P/E, which has now run into operational problems, has lost trust of its customers, and the stock has fallen 40%. The company had invested in food safety measures that are "way ahead" of the industry. Wall Street is worried.
Sound familiar?
While these examples are new, these stories are behind many of Warren Buffett's greatest investments: American Express, during the salad oil scandal, GEICO in 1976, which was thought by some to be going bankrupt when Buffett bought it, and Wells Fargo in the 90s, where the fears regarding a California real estate downturn overshot to the downside.
All of these were fantastic businesses with large TAMs that were going through a severe downturn but were still the same fundamental business underneath. Chipotle's best-in-class unit economics, plus the best management in the business, gives me confidence that this is a golden opportunity for anyone with a time horizon beyond the next 1-2 quarters.
Keep in mind that Chipotle has only 2,000 locations, while McDonald's (MCD) has 36,000+ and YUM has 40,000, and the mission is to "change the way people eat and think about fast food."
Keep in mind that before E.coli, Chipotle's model achieved operating leverage just by adding new restaurants, even without restaurant-level margin expansion. Consider that from 2010 through Q3 2015, restaurant-level margins improved by 85 basis points, but overall company EBIT margins improved by 310 basis points.
Some investors who missed the boat were mystified by Chipotle's success, and underestimated its differentiation. I think the reason it was such a hit is because the fast-casual restaurant scored highest across the categories of "Quality, Value, Speed." CMG's food was the highest quality that was that fast and cheap, and the cheapest and fastest among restaurants that served as high-quality food. While others point to new competitors in the QSR space, I still have yet to identify a restaurant - Panera (PNRA), Shake Shack (SHAK), Sweetgreen, or others, who can match Chipotle across these three categories, and thus feel customers will return once more as time goes by and there are no more safety incidents. This is evidenced by Chipotle's own research that said there was no clear winner among their lost customers; the loyal CMG customer seems to have scattered to various other places. I find this encouraging.
My general assessment was that if you were a believer in Chipotle's potential prior to the E. coli incidents, you agree with me, whereas if you always thought it was an "overpriced burrito joint that doesn't even serve real Mexican food," you think that it is a short even here.
So how long will Chipotle's stock price languish? It is hard to tell. In the previous examples, Starbucks had a 2.5 year downturn but, again, that turnaround was a huge fix and the problem occurred in 2007 which was augmented by the 2008 financial crisis. Netflix stayed down for roughly 1.5 years. Pandora troughed for 1 year before commencing its ascent, and Ulta troughed for roughly 8 months.
With Chipotle, we are roughly 5 months in. Some may wish to wait. However, investors with a long-enough time horizon should at least begin a partial position now, if you agree with the thesis. I will next take a look specifically at prior food safety incidents.
Prior food safety incidents
I believe that the stock will begin to move up when SSS starts to meaningfully jump from the low-mid 20% range. When might this happen? Taking a look at prior incidents (thanks to Credit Suisse analyst for supplying the data) it seems that Chipotle may follow the shallower trough and more gradual bounce-back given that there were multiple incidents that really spilled over into the next quarter.
Given that Chipotle had multiple incidents spread over the "event quarter," and given results to date, it appears that recovery will likely track the same pattern as the YUM China incidents (light blue lines), where the trough is drawn out longer but the recovery steadier 3 quarters after the "event." This means that 2Q is likely to remain challenged but the negative SSS for CMG should get a little less negative in 3Q.
In all of these incidents, sales remained roughly 10% below peak 8 quarters later, but, as we know, all of the stocks eventually recovered fully, more or less.
Chipotle, however, maybe slightly different in that it was such a high-flyer and disruptive force in the industry. I tend to believe superior management will do a better job in recovery than predecessors.
"It was already slowing"
Not really. You could have said the same thing in 2013, when the company was making tough comps, before comps ballooned back up to the mid-teens in 2014.
Also management was undertaking a number of steps to bolster SSS prior to the incidents:
1) Digital ordering (there was no compelling app or online options prior). To this end, they hired Starbucks' CTO in October, prior to the incidents.
2) Chorizo initiative and alcohol (fresh-made margaritas)
3) The lower 2015 SSS were facing both tough comps and a carnitas shortage for much of the year, which has been fixed.
Another thing I took away from the Berkshire Annual Meeting was when Buffett talked about trusting his managers, often buying businesses with a handshake and little due diligence. Chipotle's two CEOs and impressive CFO are exceptional managers and extremely passionate about the business, and you can see this in their conference calls and interviews. I will gladly shake their hands.
Valuation
So where does that leave the valuation? It depends on your time horizon. It should be noted that when analysts give price targets they a) have a 1-year time horizon and b) usually use some form or multiple based on 2 years EPS. However, that is really not appropriate in this case, and it is even more inappropriate to both lower the EPS and also lower the multiple (which many have done), as that is essentially double-counting.
The intrinsic value of a business is the PV of all future cash flows, discounted back to the present. Multiples are used as shortcuts to get to this value, that's all. All are based on 3 things: cash flows, growth, and risk.
When you have a company that was growing as consistently as CMG, it is also easier to model.
I therefore made a "Pre-Coli" DCF with the following assumptions:
Unit growth to 5900 over 15 years, with 2% SSS (4000 domestic Chipotles, 1900 international, Shophouse, and Pizzeria Locale, maybe a Burger initiative).
Operating margins modestly expanding over 15 years to 20% (Q3 2015 was already at 18.7%).
Discount rate of 6.2% but rising to 8% over time as the RF rate possibly rises (the restaurant industry is a low-beta industry, hence low discount rates and the high PEs, in general). 2% terminal growth, which leads to a terminal EV/FCF multiple of 16.7x - for reference, McDonald's today trades at a 30x EV/FCF.
I did not capitalize leases - I could have, but that may have even increased my price target as CMG is extremely efficient in their real estate with a 60% ROIC. (prior to the incidents).
I got a price of $835, pre-ecoli, and I think that these assumptions were on the conservative side. Keep in mind that prior to the incidents several analysts had price targets of $850 or more.
The terminal value (5900 restaurants, 2% SSS, discounted back at 8%) alone was ~ $470/share, which is above the current price. This means you are getting a discount to a conservative terminal value, plus the next 15 years of cash flows for free.
Cash flow this year may be slightly negative (net capex last year was ~128M, so EPS would have to be $4/share this year to "break even." So even in the worst year, it's an even bet whether cash flow will be below zero.
If this year is a "0" and then cash flows return to normal in my model, my price drops all the way to $821/share.
If there are 2 years of "0," before normalization, the intrinsic value in this scenario falls all the way to $806/share.
If there are 3 years of "0" before normalization, the value plummets to $790/share.
You get the idea.
Supporting Documents