2018 Annual Letter
By Chamath Palihapitiya Social Capital’s Performance vs. the S&P 500
Social Capital’s Performance vs. the S&P 500
To those who support Social Capital:
This is the first of our annual letters where we discuss our investments and other thoughts on technology, markets, and our mission.
Seven years ago, we started Social Capital to tackle hard problems at a time when few investors were doing so. While many investors fawned over social media networks, photo-sharing apps, and other consumer-oriented investments, we invested in healthcare, education, and frontier technology businesses. Many of the companies we funded were initially unable to raise money from other institutional investors until we provided them with the long-term, patient capital that they needed. We believe we will be rewarded handsomely for our contrarian bets. As we write to you today, we look back at our first seven years and look forward to our next seventy. As the world changes, our tactics will evolve but our goals will remain the same: we want to empower entrepreneurs to solve the world’s hardest problems while generating compounding returns, both economically and socially, for the long term.
The maturation of today’s tech ecosystem
Looking around the technology industry today, there is an indisputable maturity in both front-end products and the back-end infrastructures that support this new internet economy. Cloud infrastructure has become a common foundation for the entire industry. Initially misunderstood by many, it is now unlocking waves of innovative entrepreneurship, frenzied business competition, and newly created value for consumers. The same can also be said for the continuous deployment and improvement of mobile phones and wireless internet. With the majority of the world’s population now connected, and with our phones with us personally at every minute of the day, it is little surprise that every industry is being remade in the mobile Internet’s likeness: media, retail, transportation, and more. Opportunity is everywhere.
This technological maturity is critical to understand as we are no longer in the earliest innings of innovation, and the implications on progress and returns manifest in several important ways.
Consider the systematic processes with which the Googles and Amazons of the world continually reinvest new capital into their own businesses. For instance, when they build a new data center, they know that their capital is being put to work efficiently, and they understand clearly their potential return on investment. An investor with capital to allocate has a strong case to make for simply giving it to Jeff Bezos or Larry Page, and asking, “Please put this into whatever you see fit.” This is important because large internet companies are truly just hitting their stride. They are now competition for every company in every sector, and they are formidable.
Said more simply, Big Tech will get bigger and will leave less room for obvious companies doing obvious things. The demands of innovation are going up, and the quality of the ideas and teams working on those ideas matter now more than ever in this David v. Goliath landscape.
Of course, one would think that investors should become more circumspect about the utility of their capital during times like these. Curiously, the opposite is currently true and is setting up for a massive rude awakening.
Since the great financial crisis, the quantity of capital that has made its way into the tech ecosystem seeking to fund the next generation of successful businesses has steadily increased. We don’t just have big companies anymore. We also have big funds. We hear a lot about the Softbank Vision Fund, which has a minimum check size of $100 million and a target of $50 billion per year of investment. Funds like the Vision Fund are not just great “disruptors of industry dynamics”: these mega-funds exist because there’s a real opportunity to act as a king-maker for growth-stage companies trying to be big companies. As a result, these “mega-funds” are likely here to stay.
However, these mega-funds only tell half the story: there has also been a continuous surge of seed capital flowing into the industry as successful founders, builders, and fund managers reinvest their own money into the earliest stages of technology startups. They invest not only in pursuit of future returns, but also for the social cachet associated with claiming, “I've backed the next big thing.”
Whether small or big, everyone wants into the party.
It’s not surprising, then, that all of this new investor money flowing into the tech asset class might put pressure on both present and future returns. In the early days of venture capital, when pioneers like Arthur Rock and Don Valentine crafted deals by hand, venture capital could hardly be considered an asset class at all: it was simply a small, obscure kind of private placement that dealt in emerging technologies like information technology and biotech. Deal terms were negotiated one at a time.
As we fast forward through the 80s, 90s and into the dot com bubble, venture capitalists and their funds became hot commodities. Their capital was in high demand, and they were rewarded for being in the right place at the right time. Now, we seem to have arrived in a new, distinct third phase of venture capital where money is no longer scarce at all. Rather, it has become plentifully available. The collective returns reflects the new reality that venture capital does not deliver a premium for its investors. In fact, the VC industry reliably trails the S&P.
So why keep putting good money after bad? How is it being spent? And why does this mediocrity perpetuate?
The Accelerating Treadmill of User Acquisition
There was a time when investing money into an early-stage startup was something of a black box exercise: Intel’s early investors knew they were betting on a new kind of microprocessor, but the technical and go-to-market steps required to get from Point A to Point B required considerable exploration into the unknown. Today, venture investing has a different feel. Most new products and services are built largely on top of rented cloud computing and open source software, and software business models typically fall into a few well-understood buckets (ad supported, SaaS, freemium, etc). The hardest thing for most startups today is the path to market: first finding product-market fit and a way to reach customers, and then building a ruthless machine to acquire, monetize, and retain them. Because of this, when the VC industry invests capital into fast-growing startups today, the plurality, if not the majority, of invested capital will go into user acquisition and ad spending, for better or worse (usually worse).
Nicholas Sibley has a great quote about banking: “Giving capital to a bank is like giving a gallon of beer to a drunk. You know what will come of it; you just don’t know which wall he will choose.” We’ve reached a point today with ad spend in tech that feels metaphorically similar. Startups spend almost 40 cents of every VC dollar on Google, Facebook, and Amazon. We don’t necessarily know which channels they will choose or the particularities of how they will spend money on user acquisition, but we do know more or less what’s going to happen. Advertising spend in tech has become an arms race: fresh tactics go stale in months, and customer acquisition costs keep rising.
We know this game well because we played it for a long time, and quite successfully. At Facebook, my Growth team pioneered the discipline of user acquisition as a science. Our tactics for acquiring and retaining users became the industry standard for a good reason: we were the best at what we did. When Social Capital was founded, we brought over that expertise from Facebook and shared it with our portfolio companies. Our growth and data science teams played key roles in helping some great companies become even better. It’s no accident that due to its importance, user acquisition and growth has become such an entrenched part of the Silicon Valley zeitgeist.
Peter Drucker once said that “The purpose of a business is to create a customer”, and he said so for good reason: your customers are ground truth, and your business will live and die by them. Silicon Valley companies have learned this lesson, and that’s partly why so many successful businesses have emerged out of this culture and this mindset. Unfortunately, today’s massive venture-backed advertising, sales, and user acquisition playbook has morphed into one that champions growth at any cost.
And it is creating a big bill that will soon come due...
One important reason why ‘growth for its own sake’ has come to dominate the tech industry is because of the powerful network effects that come from size (again, the byproduct of living in a world dominated by Big Tech). In an internet-connected world, several kinds of businesses - platforms, marketplaces, aggregators, and social networks, to name a few - stand to become enormously valuable and profitable should they reach a certain critical mass. There’s a reflexivity to these network-based businesses. They reason, “as we become large, our product will become better and our business more valuable. Therefore, we should spend money to become large. We’ll obtain that money by raising equity at a high valuation, which is justified by how large and valuable we will become once we spend the money.”
In a world where only one company thinks this way, or where one business is executing at a level above everyone else - like Facebook in its time - this tactic is extremely effective. However, when everyone is acting this way, the industry collectively becomes an accelerating treadmill. Ad impressions and click-throughs get bid up to outrageous prices by startups flush with venture money, and prospective users demand more and more subsidized products to gain their initial attention.
Warren Buffett once observed that this kind of arms race is not unlike a parade where one spectator, determined to get a better view, stands on their tiptoes. It works well initially until everyone else does the same. Then, the taxing effort of standing on your toes becomes table stakes to be able to see anything at all. Now, not only is any advantage squandered, but we’re all worse off than we were when we first started. Such is the world of user acquisition in tech today: as growth becomes increasingly expensive, somebody must be footing the bill for all of this wasteful spending. But whom?
It’s not who you think, and the dynamics we’ve entered is, in many ways, creating a dangerous, high stakes Ponzi scheme.
The Shuffle Game
Over the past decade, a subtle and sophisticated game has emerged between VCs, LPs, founders, and employees. Someone has to pay for the outrageous costs of the growth described above. Will it be VCs? Likely not. They get paid to allocate other people’s (LPs) money, and they are smart enough to transfer the risk. For example, VCs habitually invest in one another’s companies during later rounds, bidding up rounds to valuations that allow for generous markups on their funds' performance. These markups, and the paper returns that they suggest, allow VCs to raise subsequent, larger funds, and to enjoy the management fees that those funds generate.
Picture this scenario: if you’re a VC with a $200 million dollar fund, you’re able to draw $4 million each year in fees. (Typical venture funds pay out 2 percent per year in management fee plus 20 percent of earned profit in carried interest, commonly called “two and twenty”). Most funds, however, never return enough profit for their managers to see a dime of carried interest. Instead, the management fees are how they get paid. If you’re able to show marked up paper returns and then parlay those returns into a newer, larger fund - say, $500 million - you’ll now have a fresh $10 million a year to use as you see fit.
So even if paying or marking up sky-high valuations will make it less likely that a fund manager will ever see their share of earned profit, it makes it more likely they’ll get to raise larger funds - and earn enormous management fees. There’s some deep misalignment here...
There’s an analogy to be made between today’s venture backed startup ecosystem and the American healthcare industry. Part of the reason why American healthcare is so expensive is because insurers, who play a key middleman role in setting prices for medical care, have a fantastic two-sided business model. High prices, which ought to be a cost of doing business for them, are actually a key revenue driver. Why is this? High costs allow them to charge higher premiums, allowing them to pull steadily more and more money out of patients’ and payers’ pockets. As a result, the cost of medicine steadily rises, as do the insurers’ take. In the end, both patients and payers are the ones who end up as bag holders footing the bill.
The same thing is happening in today’s venture world. Highly marked up valuations, which should be a cost for VCs, have in fact become their key revenue driver. It lets them raise new funds and keep drawing fees. And just as insurers’ business model translates into higher costs of patient care, so does the modern venture model translate into higher costs of, well, just about everything. We have higher salaries, higher rents, higher customer acquisition costs, Kind bars, and kombucha on tap!
So if it’s not VCs, who ends up holding the bag? It’s still not who you’d necessarily expect. Later-stage funds, who invest large follow-on rounds into these marked up companies, do indeed pay inflated prices - but they also usually get their money out first upon a liquidity event, and are also happy to exist in “Fee-landia”. In some cases, high prices may even work to their advantage. They’re able to hold certain late-stage companies hostage to their high valuations by demanding aggressive deal structures in return for granting “Unicorn Status” (the billion-dollar valuation that VCs so crave). Unlike in other pass-the-buck schemes, the bill is not getting passed from early investors to later investors.
The real bill ends up getting shuffled out of sight to two other groups. The first, as you might guess, are early stage funds’ limited partners, particularly the future limited partners that invest into the next fund. Their money, after all, is what pays the VC’s newly trumped up management fee: marking up Fund IV in order to raise money for more management fees out of Fund V, and so on, is so effective because fundraising can happen much faster than the long and difficult job of actually building a business and creating real enterprise value. It might take seven to ten years to build a company, but raising the next fund happens in two or three years.
The second group of people left holding the bag is far more tragic: the employees at startups. The trend in Silicon Valley today is for a large percentage of employee compensation to be given out in the form of stock options or restricted stock units. Although originally helpful as a way to incentivize and reward employees for working hard for an uncertain outcome, in a world where startup valuations are massively inflated, employees are granted stock options at similarly inflated strike prices.
Overall, you can understand how this arrangement endures: VCs bid up and mark up each other’s portfolio company valuations today, justifying high prices by pointing to today’s user growth and tomorrow's network effects. Those companies then go spend that money on even more user growth, often in zero-sum competition with one another. Today’s limited partners are fine with the exercise in the short run, as it gives them the markups and projected returns that they need to keep their own bosses happy. Ultimately, the bill gets handed to current and future LPs (many years down the road), and startup employees (who lack the means to do anything about the problem other than leave for a new company, and acquire a 'portfolio’ of options.)
What is the antidote here? The antidote is two-fold. First, we need to return to the roots of venture investing. The real expense in a startup shouldn’t be their bill from Big Tech but, rather, the cost of real innovation and R&D. The second is to break away from the multilevel marketing scheme that the VC-LP-user growth game has become. At Social Capital, we did this by actively shifting away from funds and LPs to rely only on our own permanent capital moving forward. Are we crazy to reject tens of millions of dollars a year in fees? We think not, and we believe it’s time to wait patiently as the air is slowly let out of this bizarre Ponzi balloon created by the venture capital industry. In the meantime, we find comfort in the teaching of Andy Grove that only the paranoid survive.
Newly investible technology trends
In keeping with our history and in looking forward to our future, we never shy away from learning about new and challenging frontiers of technology. We would especially like to discuss two fields we find particularly important going forward, and why now is the time to double down on learning about them:
In traditional software development, developers write code that takes input data and computes an output.
For example, if a bank wants to offer a loan, it determines the interest rate of the loan through a program with the following (simplified) logic:
if the applicant has a good credit score
give them a low interest loan
else if they have a low credit score
give them a high interest loan
The input in this program is the credit score, which is calculated from the applicant's information, such as their bank account, credit card, and other loan data. The output is the interest percentage. The above style of programming makes it difficult to create some information-intensive applications, like self-driving car software. You just wouldn’t be able to specify via if statements all of the possible permutations of the car’s external environment. There's just too much input data (the weather, other cars, pedestrians, other obstacles, etc.) to make a reliable program that produces the correct output of where the vehicle should go.
However, computer scientists have recently made new breakthroughs that allow us to better tackle problems like these. In 2012, the error rate for the ImageNet Challenge, an image recognition competition, dropped precipitously to 15% using an innovative branch of artificial intelligence called deep learning. This achievement showed that machines were quickly approaching human ability to classify objects (by 2017, error rates would reach below 5%). Prior to this new innovation, humans had always been better than machines at image recognition.
Image recognition is an interesting problem because it has an unbounded input space that cannot be solved efficiently using traditional software methods (there’s just too much data and too many possible permutations of that data to analyze). Deep learning’s success proves that image recognition and similar problems, problems that humans traditionally excelled at over computers, are no longer impractical to undertake. So far, we have seen deep learning and its derivatives applied most successfully in computer vision and speech recognition. But we have only seen the beginning.
Because of the lack of distribution of talent (most of the top AI experts and new PhD’s unfortunately join Big Tech due to their professional-athlete level compensation packages), deep learning has barely scratched the surface of its potential applications. Over the next few decades, as more engineers are trained in artificial intelligence and as developer tools and frameworks get easier to use, we should see artificial intelligence successfully applied to problems that were previously too difficult using traditional software methods, such as self-driving cars, robotics, and drug design and discovery. AI can truly transform how we work, how we live, and even how we think.
As a result, our goal at Social Capital will be to own companies that are either directly applying these new advancements in artificial intelligence, or that are building the infrastructure that powers these companies.
Computers, at the end of the day, are machines for turning information into processed information. This is obviously a very powerful and flexible capability, and our use cases for computers have expanded far beyond the “information processing” jobs they were initially tasked with. Yet, they can’t do everything.
There are many different kinds of problems we can’t count on computers and software to solve. Many of them have one thing in common, which is that they aren’t problems about information. Instead, they’re problems about the physical world ("atoms vs. bits", as people like to say). Whether we need to build matter up (make materials, design drugs, process fuels), break things down (clean up pollution, treat disease) or identify and interact with things in the physical world (sense lead in water, sequester contaminants), there are a lot of hard problems in the world that computers cannot solve, but that biology can.
For many of these problems, the highest-potential path to fixing them lies in the overlap of computers and biology: computational biology. Computational biology is an emerging discipline that generally refers to two overlapping fields: 1) the practice of taking everything we’ve learned about how to build computers and applying that knowledge to building cells as a programmable, flexible, platform with which we can do useful work, and 2) productizing and automating the tools, processes, and methods we use in the lab to manipulate biology and build living systems. Although we’ve gone through a few waves of “biotech bubbles” over the past twenty years, this time may no longer only be about wildly speculative drug development, but instead about something more concrete and foundational. We'll be able to establish biological systems as engineered, all-purpose platforms that we can put to work the same way we do with computers.
Additionally, within a few years, we’ll reach a convergence point where our recent advances will start to overlap, and eventually blend, into our existing computing frameworks and infrastructure. This will have a profound and disruptive effect on many fields such as drug design and discovery, drug delivery, precision diagnostics and healthcare, engineered materials, ecology, agriculture, and much more. We’ll be able to work with biology in ways that increasingly resemble the way we work with software: as a platform for building tools, applications, and infrastructure. This time, however, we’ll be able to do it using living systems instead of code.
Social Capital as a technology holding company
It suffices for an intransigent minority - a certain type of intransigent minorities - to reach a minutely small level, say three or four percent of the total population, for the entire population to have to submit to their preferences. Further, an optical illusion comes with the dominance of the minority: a naive observer would be under the impression that the choices and preferences are those of the majority. - Nassim Nicholas Taleb
In our first seven years of business, we have repeatedly found that high returns and a positive societal impact often go hand in hand (we are called Social Capital after all). This should not come as a surprise. Companies solving the world’s hardest problems are often those with the fewest competitors. These unique businesses that stand alone in tackling seemingly impossible challenges may face difficult initial climbs and may have trouble raising capital from more traditionally minded investors. However, they will be rewarded tenfold for their patience in the long run, as they will be technologically unparalleled and will not be forced to play the same customer acquisition game on an accelerating treadmill. They are the companies that will truly create, rather than merely participate, in new markets.
This long-term constructive mindset is not just something we look for in our portfolio companies. It is a mindset we required from our limited partners as well. Our most valuable investors were not the ones who cared most about returns, but rather the ones who also cared most about making a positive contribution to society. In time, they will be the ones most rewarded for their patience and their outlook.
Venture Capital firms typically do not divulge the composition of their limited partners. However, for the sake of transparency, we show our legacy LP composition below. VCs typically raise money from family offices, university endowments, non-profit foundations, corporate investors, and fund-of-funds. The best VCs are selective about which LPs they will allow into their funds. At Social Capital, our LPs consisted mostly of family offices, and they represent the majority of dollars invested. My family office was the single largest investor in our funds.
LPs that are obsessive only about numerical performance in the short term can be harmful to entrepreneurs and the broader venture capital ecosystem. These LPs, typically fund-of-funds, are incentivized to push for immediate returns and early liquidations, frequently against the best interests of entrepreneurs. In some ways, we can understand this urgency as a competitive chess move in the “shuffle game” we described earlier. But pushing for immediate returns and earlier liquidity while continuing to encourage the same old game is most certainly not the solution.
A new approach is needed. We have therefore decided that a technology holding company was a structure much more amenable to making the contrarian, long-term investments that we wanted to make. Our new structure will allow us to invest with uncompromising conviction, for the long run.
And speaking of the long run, one more thing…
October 31, 2018