Why Warren Buffett Bought Coca-Cola (KO)

Author's Avatar
Mar 08, 2012
Someone who reads my articles asked me this question: I've read a few of your articles, especially on those about Warren Buffett’s investment style…I would be grateful for your help in clarifying a question about Warren Buffett’s purchase of Coca-Cola (KO, Financial).


Throughout 1988 and 1998 Warren Buffett acquired more than $1 billion of KO stock. At the time Wall Street deduced Buffett paid way too much for the earnings indicating a high P/E.


Some investors believe Warren Buffett used some sort of discounted cash flow analysis to value his purchase in 1988 and 1989, which he projected out ten years. However from what I understand you may not be convinced that Warren Buffett uses discounted cash flow analysis in his valuation of businesses. I would be very interested to know how you think Warren Buffett valued his purchase.


Kind Regards,

Caesar



As best I can, I’ll let Warren Buffett answer the question himself. Let’s start with what Warren Buffett said back in the 1990s at Notre Dame:


“We have 7% of Coke. There are 660 million eight ounce servings of Coca Cola products being served around the world today, so in effect, we’ve got a 45 million soft drink business with our 7%. We think of businesses that way. I say to myself ‘just increase the price a penny and that’s another $450,000 a day for Berkshire.’ I mean, it’s a nice sort of thing. When I go to bed at night I figure that by the time I wake up 200 million Cokes will have been consumed.”


I think there were really only 3 reasons why Warren Buffett bought Coca-Cola stock:


1. Pricing power


2. Consumption per capita growth


3. CEO focused on growth


Here are the top 10 countries where people drink the most 8-ounce servings of Coca-Cola products per person per year:


1990 2000 2010
Mexico 280 456 675
Malta 171 333 606
Chile 148 341 445
U.S. 289 405 394
Australia 218 288 319
Argentina 105 224 318
Spain 148 251 284
Panama 73 165 258
Austria 140 174 255
South Africa 139 173 255



Okay. And here is the number of 8 ounces servings of Coca-Cola products drank per person per year for the entire world:


1990 2000 2010
Worldwide 43 67 89



So, in 1990 and 2000 – which we’ll use as the closest years to when Warren Buffett bought Coke stock – U.S. per person consumption of Coke products was over 6 times higher than the worldwide rate per person. In 2000, U.S. consumption per person was still 6 times the worldwide rate. In 2010, U.S. consumption per person was still more than 4 times the worldwide rate.


Coca-Cola’s brand was not limited to the United States. The history of cola – both Pepsi and Coke – is a history of increasing consumption per capita. There are many reasons for this.


The biggest is the complexity of the flavor. What exactly does Coke taste like? I can describe what orange soda tastes like. I can describe what grapefruit juice tastes like.


Can I really tell you what Coke tastes like?


How many products are there that you can drink and yet have a very hard time describing.


I can think of 3:


1. Water


2. Wine


3. Coffee


They are commodity products. They’re natural. Water is the hardest thing to describe of all – it’s pretty much the easiest thing to drink too much of (as long as it’s served at room temperature).


Wine and coffee are also tough. A ton of past associations has already subtly branded them.


You have regional branding. People will buy French wines that are no better than Spanish wines – and pay 3 times the price – because France ended up with a certain image. So whether growing grapes along the Loire really makes them worth more than grapes grown along the Oja – people will pay more for the French stuff. France imposed a very strict labeling system. They’ve created a system where some winemakers can bask in the glow of what’s essentially a regional name.


So you’ve already got a cluttered brand landscape there.


Wine also has some very serious limitations. Its ability to penetrate deeply into all but a few cultures actually hasn’t been that great. Look at a map of U.S. wine consumption. Wine consumption in the U.S. depends pretty heavily on consumption in coastal states – essentially the West Coast and the Northeast. The situation was even worse when Buffett first bough Coca-Cola. There has been an increase in wine consumption in the South, Midwest, etc. relative to the highest wine drinking states over the last 40 years. But, overall, the pattern is not good.


Wine has been available for a very long time and yet it still is mostly a Western European and urban American product. It has not permeated new cultures very easily. And it’s not at all clear that wine drinking is higher in places that are wealthy because they are wealthy – rather it seems an accident of history that some big wine drinking countries (around the Mediterranean) got rich.


You will always have a percentage of the population – in the U.S. it’s about 35% – who won’t touch wine because they don’t drink alcohol at all. And even those who drink in moderation all the time are unlikely to use wine as a morning, noon, and night drink.


What about coffee?


Here we have a much clearer picture. Rich countries drink a lot of coffee. Culture is less important. For example, wine consumption is very uneven across Europe – basically if I showed you where Greek and Roman ports were 2,000 years ago those are still the countries where folks drink the most wine.


This is not true of coffee. Look at a map of coffee consumption. Pick a region and you’ll pretty much find the richest countries in the region drink a lot of coffee. A G-8 summit is a summit of coffee drinkers. The U.N. general assembly – not so much. But all of Europe – not just the west – the U.S., Canada, Australia, Japan, etc. drink plenty of coffee. And on each continent you find a tendency for richer countries to drink more coffee than poorer countries. Of course, other things being equal, the coldest places do drink more coffee than the hottest places.


Which brings us to the first of coffee’s hindrances as a worldwide blockbuster drink. It’s best served hot. Some places are already hot. Some seasons are already hot. This limits consumption.


More importantly, there’s a ton of work that goes into the proper way to make coffee. You literally have to have a machine in your house dedicated to the production of coffee.


How much soda do you think would be sold if you needed to buy, fill, and clean a soda machine kept on your kitchen counter? Not much. People want to drink a lot of coffee. But there are some real issues with drinking lots of it. It’s an inconvenient drink.


Cola solves these problems. Like wine, it can be ready to drink even while it’s shipped long distances and stored for long times.


Unlike wine, overconsumption doesn’t cause drunkenness. It’s socially acceptable to drink at most times of the day. You can give it to kids (this is critical). And you can openly carry it around with you.


If you read Poor Charlie’s Almanack you’ll get Charlie Munger’s take on cola. If you’re skeptical about the points he’s making – about taste memory – try this experiment:


· Monday: drink 10 servings of cola in one day


· Tuesday: drink 10 servings of milk in one day


· Wednesday: drink 10 servings of orange soda in one day


· Thursday: Pick one of those drinks and have another 10 servings


You’ll pick cola. And this will work with other products too. As much as you think you like juice – you won’t be able to last a week drinking as much juice everyday as some people drink water, cola, etc.


So you have a handful of drinks that have the potential for huge consumption – water, wine, coffee, cola – and in 1988 one of them (cola) had the best economic prospects.


Warren Buffett thought the cola company with the best prospects was Coca-Cola. As it turned out, he would’ve done just fine buying Pepsi (PEP).


Of course, today, there are a lot of negative associations with soda. Almost as much as with wine. The pendulum has swung in this regard. But that happens. One hundred years ago attitudes toward alcohol consumption were extremely negative in developed countries (for good reasons). Today, our social problems are different and a lot of people are more concerned about obesity and all-natural foods than about people showing up at work drunk.


So, I’m not even sure Warren Buffett’s thesis for Coca-Cola’s flagship product worked out as well as he expected in the developed world. It probably didn’t exceed his wildest dreams in the U.S.


The key to understanding Buffett’s investment is that he knew Coke had pricing power. He knew Coke had a beachhead on a lot of different countries. And he knew that once a culture was introduced to Coca-Cola they would drink more and more and more of it.


Coke is oddly similar to Warren Buffett’s GEICO investment. In fact, Coke is basically a blend of See’s Candy and GEICO.


See’s Candy had pricing power. But its brand doesn’t travel well. In fact, See’s has really never crossed the Mississippi River – much less the Atlantic Ocean.


What about GEICO?


GEICO has a huge runway. It can keep growing and growing and growing. There are millions of drivers in America who need car insurance. They literally are required to have a car insurance policy if they want to drive. You can’t – like with Coke – sell them 3 or 4 or 13 policies. But you can keep growing your share of the market. GEICO was the low cost competitor. It had a better way of selling car insurance. A cheaper way. Progressive is a similar company. Buffett even mentioned that GEICO and Progressive had once been in merger talks but it fell apart – for non-financial reasons. I get the impression Buffett wouldn’t mind owning Progressive.


Coca-Cola was a combination of See’s and GEICO. It had the pricing power of See’s. And it had the nearly infinite growth prospects of GEICO.


The two best ways to make more money in business are:


1. Raise prices


2. Increase sales of existing products to existing customers


Coke could do both. They could always raise prices at least as fast as inflation. And they could increase consumption per customer. The reason for this is easy to see – why should people in the U.S. always drink 6 times more Coke than people in the rest of the world?


There’s really no reason. That pattern wasn’t going to be sustained.


And, in fact, this is proved by Malta being on the highest consumption countries list. What do the U.S. and Malta have in common?


Eating habits. That’s all it is. They eat and drink a lot per person. The Maltese obesity rate is higher than the U.S. rate. As long as other countries move in the direction of the U.S., Malta, etc. they’ll increase consumption of Coca-Cola products.


Basically, we are talking about huge gaps between countries that are going to trend closer and closer to parity over time. This is not new. There was once a huge gap in heights between Europeans and Americans. Look at the ratios on that number and how they closed. Why? Because Europe increased nutrition over time while a country already getting enough nutrition can’t keep adding to it forever – in fact, at a certain point, more food will just make people fatter not taller. Basically, you hit a ceiling.


There’s always a limit to growth. There’s a kind of carrying capacity to markets, customers, etc. It’s not a hard limit. It’s not very efficient to press beyond a certain point.


Where was that point with GEICO?


Far, far away. Buffett knew GEICO’s market share in car insurance was tiny. They had decades of growth ahead of them.


And what about Coca-Cola?


That’s simple. It was the U.S. consumption per person. You aren’t going to increase the consumption of Coca-Cola products in the U.S. forever – in fact, it’s dropped from 2000 to 2010.


But there’s no reason that Coca-Cola can’t reach the same consumption levels per person everywhere else. It’s probably never going to hit that limit. But it’ll trend towards that limit.


When the U.S. was consuming 6 times more Coke than the rest of the world, all you had to do was look at these other countries and ask yourself: “What if instead of consuming 1/6th as much Coke as the U.S., the people in this country someday consume half as much Coke?”


That’s all. It’s not a very aggressive assumption. 3 countries – Mexico, Malta, and Chile – have already passed the U.S. in consumption of Coca-Cola products. The idea that the potential for growth in any country is at least half the per person consumption level in the U.S. – that’s a perfectly reasonable assumption.


And that’s the kind of assumption Buffett made. He saw that Coca-Cola consumption was rising around the world. He saw the management team of Roberto Goizueta and Don Keough were focused on growing earnings per share through the core Coca-Cola business.


Now, just saying you are targeting growth is not going to get Warren Buffett to buy your company’s stock. But when a company with the potential to increase physical units sales around the world for a very long time to come and the ability to increase prices for a very long time to come says it’s focused on growth – that’s when Warren Buffett will buy your company’s shares.


He knew Coke could do it. In fact, looking at past reports – and the company’s very long history – he knew the potential had been there and been untapped for a long time.


This is not a new formula. If you look at Gillette you see some of the same thing. Gillette had been a mediocre steward of a great brand for a while before Buffett invested in the company. Or if you look at Disney in the first half of Eisner’s tenure there – it was mostly about tapping a brand that already existed. He didn’t create the ability to raise prices, sell videos, etc. It was already there. Eisner just got a return on the intangible assets he was handed – which the company hadn’t been doing before.


That’s the story with Goizueta. You had someone who was very ambitious at achieving certain goals in a position where it was possible for him to do that for many, many years. That’s unusual.


But, ultimately, it’s about value. Even though he was paying a high P/E ratio, Buffett thought he was buying Coca-Cola for less than it was worth. He didn’t just mean the stock would one day earn him a great return because of all this future growth he saw coming. Buffett actually thought the Coca-Cola company – on the day he bought shares in it – was worth more than its market cap:


“At the time we bought Coca Cola just two years ago, [we ended up buying] 7% of the company. We paid a billion dollars, so we were paying $14 billion, essentially, for the whole thing. You can sit down in five minutes – I mean, everybody here understands Coca Cola. If Philip Morris were to buy Coca Cola that day, they would have paid $30 billion.”


As far as my point about untapped potential, Charlie Munger said it best:


“Within the growth stock model, there's a sub-position: There are actually businesses, that you will find a few times in a lifetime, where any manager could raise the return enormously just by raising prices—and yet they haven't done it. So they have huge untapped pricing power that they're not using. That is the ultimate no-brainer.


That existed in Disney. It's such a unique experience to take your grandchild to Disneyland. You're not doing it that often. And there are lots of people in the country. And Disney found that it could raise those prices a lot and the attendance stayed right up.


So a lot of the great record of Eisner and Wells was utter brilliance but the rest came from just raising prices at Disneyland and Disneyworld and through video cassette sales of classic animated movies.


At Berkshire Hathaway, Warren and I raised the prices of See's Candy a little faster than others might have. And, of course, we invested in Coca-Cola—which had some untapped pricing power. And it also had brilliant management. So a Goizueta and Keough could do much more than raise prices. It was perfect.”


And that’s why Warren Buffett bought Coca-Cola.


Ask Geoff a Question about Warren Buffett’s Investment in Coca-Cola

Check out the Buffett/Munger Newsletter

Check out the Ben Graham: Net-Net Newsletter