[This post is adapted from our Q2 2018 letter, which was shared widely on Twitter and was even excerpted in a book. For more of our letters, please register here.]
Software… coming now to an industry near you
The big picture is that software is eating the world—that is, many of the products and services developed over the past 150 years are transforming into, or being disrupted by software (full credit to venture capitalist Marc Andreessen for articulating this idea years ago). The implications are enormous: software is infinitely replicable and, through the internet, can be delivered at zero marginal cost. When a major input to business—distribution cost—goes to zero, entire industries get disrupted. When one can build a business model from the ground up with entirely new assumptions, one can attack incumbents in a way that is very difficult to defend.
This shift has resulted in the creation of new business models that would have been impossible in the previous era. One example of this is software as a service (SaaS). Previously, software was sold in shrink-wrapped boxes in retail stores once every several years. The consumer had no way of knowing if the vendor would be in business for improvements down the road. The vendor’s sales were lumpy and there was no feedback loop between usage and product improvement. Enterprise sales involved lengthy and expensive installation periods followed by similarly expensive maintenance contracts.
With software that is delivered as a service, all of this changes. The financial incentives between consumer and vendor are better aligned. There is instantaneous feedback in usage and indeed the vendor can update the software yearly, monthly or even by the minute. You can already see this in large consumer businesses: Google, Facebook, and Netflix are all examples of software businesses that live in the cloud (computer servers around the world) and are delivered to you on a continuous basis, with seamless feedback loops providing the vendors with ample data to improve and expand the service.
Cloud computing, the hosting of these services, which is a business led by Amazon, Microsoft and Google, has the potential to be a platform: a surface that itself is very profitable, but that enables an even larger ecosystem of companies to be built on top of it. A good example of a platform is Microsoft’s Windows: it was (is) an incredibly profitable franchise, and allowed an even more profitable ecosystem to evolve around it. I’m optimistic about the future of the partnership not only because of our existing investments in this area, but also because of the deep pipeline of interesting cloud-enabled companies we are evaluating.
As we have explained in our recent Q3, Q4, and Q1 letters, as well as during our investor day, our focus is increasingly in exceptional franchise businesses that can grow and reinvest for many years. We plan on holding these businesses indefinitely, subject to our evaluation of their underlying competitive advantages (we want growing or at least stable moats over time), and valuation (even the world’s best business at too high a price is not an attractive investment).
We have good reasons for our shift in focus. First, I believe that the traditional “value investor” mentality of buying cheap securities, waiting for them to bounce back to “intrinsic value,” selling and moving onto the next opportunity, is flawed.
In today’s world of instant information and fast-paced innovation, cheap securities increasingly appear to be value traps; often they are companies ailing from technological disruption and long-term decline. This rapid recycling of capital also creates an enormous drag on our after-tax returns. In addition, by focusing on these opportunities, we incur enormous opportunity costs by not focusing instead on the tremendous opportunities created by the exceptional innovation S-curves we are currently witnessing.
There is another reason “cheap” is a poor proxy for value: the new business models I described above—SaaS in particular—are not well suited to traditional GAAP accounting. Here’s why: if distribution costs are zero, the optimal strategy is to gain as many customers for your software product, as quickly as possible. In digital businesses, there are increasing advantages to scale, and many of these companies operate in winner-take-all or winner-take-most markets. The name of the game is thus to build, grow, then monetize. Frequently, this means spending a lot of money in sales and marketing, which depresses reported earnings.
Thus, SaaS companies spend to acquire customers upfront, and recognize revenue from those customers over many years. This mismatch burdens the income statement. Some of the most successful—and highest performing stocks—in the SaaS world have spent many years growing despite producing no meaningful accounting profits. They are very profitable in terms of unit economics (as we’ll explain below), and very profitable once they stop reinvesting every dollar generated into further growth. The traditional method of screening for low P/E stocks doesn’t work in this scenario.
We expect the world to have more of these types of businesses in the future; we are not going back to the old world. The diffusion of the internet around the world, and cloud computing on top of it, will see to that. The investment in exceptional businesses riding these trends is our primary goal.
Ditch the cigar butts
We talked about a useful example of investing in secular trends during our investor day: In 1948, the most famous value investor of the “traditional” variety (Benjamin Graham) put 20 percent of his fund in GEICO stock. It wasn’t traditionally cheap; indeed, Graham noted that “Almost from the start the quotation appeared much too high in terms of the partners’ own investment standards,” referring to his and his investment partner’s traditional valuation metrics.
Over the following 24 years, the stock went up 145x, or about 23 percent compounded per year—a fantastic result. Graham later wrote, “Ironically enough, the aggregate of profits accruing from this single investment decision far exceeded the sum of all the others realized through 20 years of wide-ranging operations in the partners’ specialized fields, involving much investigation, endless pondering, and countless individual decisions.” (Emphasis mine.)
In my view, GEICO was the only real value in the portfolio; the rest were distracting cigar butts. Imagine if Graham had recognized the characteristics of GEICO that made it so exceptional, and had instead focused his efforts on finding more investments like it?
This begs the question of hindsight bias: would it have been possible to foresee GEICO’s success? Perhaps so; after all, Graham thought it was a good enough business to make it a large position.
In 1948—indeed, decades earlier—it was clear that cars were here to stay. Government data since 1900 shows that by 1947 the registration of cars and trucks in the US had grown at 20 percent compounded over the previous 47 years. The population of the country had compounded at 1.4 percent per year during the same period. There was only one car or truck for every four people in 1947. Looking at this data, it was clear that there was room for continued growth.
Indeed, a New York Times article from February 1947 quotes an auto dealer, “Go to a dozen or more showrooms and try to buy a car. The average terms set for delivery will be four to six months and these will be hedged in with provisions. Some will not even accept orders because of the size of the backlogs, while others will designate them for late 1948 or even 1949.”
It’s fair to say, then, that autos were clearly a growth industry. Here’s how the future unfolded: in 1948, there were about 0.28 cars and trucks per person in the US; this increased to 0.8 in the late 90s, up nearly three-fold. Similarly, there were 41 million cars and trucks in 1948 and 251 million in 2014, up six-fold.
Below are a couple of charts showing these adoption curves. These aren’t unique to automobiles; we can find similar curves for the diffusion of the printing press, electricity, indoor plumbing, radio—all technologies predating 1948. Graham could have plotted the growth in the auto industry through 1948, reasoned by analogy, and observed that with GEICO, he was riding a powerful adoption curve.
What else did GEICO have going for it? It was direct-to-consumer, bypassing the sales agent and therefore offering a product at times up to 30 percent cheaper. It also had superior risk targeting, underwriting lower-risk customers like government employees, teachers and veterans. GEICO was therefore a superior, disruptive product, riding a growing S-curve (car and truck adoption).
![](https://static.wixstatic.com/media/659a41_239f17caeaea41ce8227f71ccd49a544~mv2.png/v1/fill/w_980,h_816,al_c,q_90,usm_0.66_1.00_0.01,enc_auto/659a41_239f17caeaea41ce8227f71ccd49a544~mv2.png)
![](https://static.wixstatic.com/media/659a41_1c11402d4730464dab0bbcd6e26c29d1~mv2.png/v1/fill/w_980,h_840,al_c,q_90,usm_0.66_1.00_0.01,enc_auto/659a41_1c11402d4730464dab0bbcd6e26c29d1~mv2.png)
The logical conclusion by an intelligent investor would be to find more GEICOs and eliminate the cigar butts from the portfolio, and that’s exactly what we have done. GEICO was ultimately acquired by Benjamin Graham’s student, Warren Buffett, and now thrives inside Berkshire Hathaway. It continues to gain share in US auto insurance.
Advertising for customers
GEICO is typically among the top five advertisers in the US. Advertising is GEICO’s expense for acquiring customers—both direct response advertising (you go to Google.com, search for car insurance, see a GEICO ad, click and purchase)—and brand advertising (you need car insurance, and the first brand you recall is GEICO and its friendly gecko). As a large advertiser, GEICO has enormous customer acquisition costs.
Why is Buffett, a savvy investor, willing to pour so much money into advertising? Because the payoff is huge: once you buy car insurance from GEICO—input your credit card, social security and vehicle identification numbers, and print out the insurance card to keep in your glove compartment—it becomes a hassle to switch. The product is sticky, it has high switching costs. It’s essentially a subscription product (that auto-renews) with a very high lifetime value. When a business has a high ratio of lifetime value (LTV) to customer acquisition costs (CAC), it’s rational to invest as much as possible in acquiring new customers, and that’s exactly what GEICO is doing. (Think back to the SaaS example above.)
GEICO worked out for Graham and Buffett because of these superior characteristics: the company had the ability to grow, and to do so while earning high returns on capital.
Grow… but only if you have high returns on capital
In the latest, sixth edition of the book Valuation, Measuring and Managing the Value of Companies by McKinsey & Company, the authors wrote, “We’ve found, empirically, that long-term revenue growth—particularly organic revenue growth—is the most important driver of shareholder returns for companies with high returns on capital. We’ve also found that investments in research and development (R&D) correlate powerfully with positive long-term shareholder returns.”
Growth works for companies with high returns on capital. (In lousy, low-return businesses, growth actually destroys value.) The internet has enabled two very interesting dynamics: huge growth (with zero distribution costs, companies can scale globally), and the use of very little capital (no expensive factories and stores to build). Internet-first businesses therefore score very highly in the high growth, high return on capital metric. And as the authors point out, this metric is highly correlated with attractive shareholder returns.
They also mention investments in R&D. Innovation and investment in the business is key. What’s the point of attracting the world’s best, brightest employees, showering them with stock options, only to deprive them of growth opportunities? Not only that, but competition—capitalism—will ensure that a stagnant company that isn’t constantly learning, growing and improving will no longer have a business in the future.
Yet most corporations are incapable of innovating because of misaligned incentives. I saw this first hand at this year’s Consumer Analyst’s Group of New York (CAGNY) conference, which hosts presentations by most of the world’s leading packaged goods companies, like Unilever, General Mills and Johnson & Johnson. The entire conference had a funereal atmosphere, with all executives complaining about “disruption” to their businesses, guiding to tepid revenue growth, and promising more cost cuts (anything to juice earnings).
While some of these companies pay lip service to innovation, overall they are run by management teams focused on keeping margins consistent and earnings per share growing or at least stable.
True innovation requires pain; the reason is that innovation requires attacking new markets, and by definition, new markets cannot be analyzed (because they don’t yet exist). Most companies cannot stomach the short-term hit to earnings required for innovation because management incentives are not aligned.
These managers are mercenaries; they’re in it for the money. Most management teams’ goals are to grow earnings per share over a certain number of years, collect their yearly bonuses (which are predicated on these metrics), and then retire with a golden parachute. As Charlie Munger said, show me the incentives, and I’ll show you the outcome.
Incentives really matter, and when it comes to the majority of corporations, the incentives are at odds with those of long-term shareholders. The McKinsey book cited above notes that “in a survey of 400 chief financial officers, two Duke University professors found that fully 80 percent of the CFOs said they would reduce discretionary spending on potentially value-creating activities such as marketing and R&D in order to meet their short-term earnings targets.”
What advantages do truly innovative companies have? They tend to be run by missionaries, not mercenaries. Missionaries will do whatever it takes to reach their mission and are willing to experiment, disrupt their core businesses, invest for growth, and seek out new markets for expansion.
Alphabet’s mission is to organize the world’s information and make it universally accessible and useful. Amazon’s is to be the world’s most consumer-centric company. At any given point in time, Amazon is running dozens of experiments with two-pizza teams (teams small enough to be fed by two pizzas), in search of innovation and growth. These experiments have resulted in countless businesses and three main pillars: Prime (Amazon’s membership program), Marketplace (opening up the platform to other sellers, now around half of Amazon’s unit sales), and AWS (cloud computing). Ongoing experiments are aiming to develop more such pillars.
Jeff Bezos has said that he’s willing to lose money for several years on any experiment before harvesting a return. Indeed, Bezos and his management team have studied the seminal work on disruption, The Innovator’s Dilemma, by Clayton Christensen. Reading the book, one has the impression that Bezos systematically built Amazon as a disruption antidote, doing the exact opposite of what Christensen explains is the problem with companies that get disrupted. This is why it’s so hard to compete with Amazon, and why the company is growing into a powerful competitor in general goods and an oligopoly on the compute economy (with Amazon Web Services). Their incentives are aligned in a way that other companies can’t even imagine.
Growth is the product of return on invested capital (ROIC) and the investment rate. In other words, companies that have high returns on capital and can reinvest a lot of their generated cash flows will grow faster. And because shareholder return is closely correlated with growth and ROIC, owning companies with these characteristics will over time result in very attractive returns for the fund.
S-curves of the future
If Benjamin Graham was—unwittingly—riding the adoption curve of the automobile in the 20th century, which technology adoption curves are we riding with our investments?
There are several of them. Overarchingly, we are riding the adoption of the internet and particularly broadband around the world. It’s hard to believe, but we’re only halfway through that curve (only 50 percent of the world by population has internet access). On top of that, we have the shift in computing platforms from desktop to mobile; every new user accessing the internet is likely doing so on a smartphone, not a PC.
Hooi Ling Tan, the co-founder of Southeast Asia’s Grab (Uber-like competitor), shared some interesting statistics at a recent conference I attended. She noted that there are 650 million people in Southeast Asia (compared with 326 million in the US), with a total GDP of $3 trillion ($18 trillion in the US). Over the next five years, that region will add 124,000 new internet users per day, every day. That’s 226 million new customers for internet-enabled businesses, albeit with a very low—but growing—spending power.
On top of the smartphone, then, there are several S-curves unfolding: the rise of ecommerce, which even today is only 6.5 percent of U.S. retail sales; the rise of social media (only 33 percent penetrated globally); video and music streaming; gaming and esports; travel and leisure; the secular shift in advertising from traditional media to online; ride-sharing and autonomous vehicles. Our investments are advantaged by each of these secular trends.
Comments