James Pethokoukis and Sebastian Mallaby
Venture capitalists are derided as incompetent gamblers whose mistakes prop up worthless startups and whose successes are more the result of luck and branding than any actual skill. But some of Silicon Valley’s most valuable companies were backed with early VC financing. So what market function does venture capital serve? And what can trends in VC financing tell us about the future for startups? To answer those questions and more, I’ve brought on Sebastian Mallaby.
Sebastian is the Paul A. Volcker Senior Fellow for International Economics at the Council on Foreign Relations and author of “The Power Law: Venture Capital and the Making of the New Future.”
What follows is a lightly edited transcript of our conversation. You can download the episode here, and don’t forget to subscribe to my podcast on iTunes or Stitcher. Tell your friends, leave a review.
Pethokoukis: Now, I generally avoid two questions during these podcasts. I don’t like asking the guests about the title of the book or why they wrote the book. But I’m going to make an exception here because I actually really want to know the answer to those questions. We’re going to start with the name of the book, because I think it’s important for how we think about the subject of the book. “The Power Law”: Where does that come from?
Mallaby: The best way to explain the power law is to start with what it is not. It’s not a normal distribution. That’s one where you have a bell curve distribution. Most of the observations in a data set are near the average. The power law distribution refers to a skewed distribution, and so just to illustrate this, let’s just take the height of American men. The average man is five feet, 10 inches tall. Nearly all American men are within three inches of that average. And so if you think about a movie theater and there’s an NBA player right at the back, and he gets bored of the movie and walks out. Although he was seven feet tall, the average height of the residual men doesn’t really change. He’s not so way off the average that it really changes things. Now flip the example and think about the average wealth of American men.
Imagine that movie theater again. This time you’ve got Jeff Bezos at the back, he walks out. The average wealth of the residual men will crash. That’s because the distribution is way more skewed when it comes to wealth, and a power law distribution like that is what you get with venture capital investing. Most startups go to zero because most startups fail. But a handful, maybe one or two out of 10, will deliver exponential growth and returns of 10x or 20x your money. So all of the profits in venture investing come from that tail of sort of exceptional companies.
Most of the books that I see these days about capitalism broadly seem to be about an explanation of why it’s failing — why the key institutions have failed, why we’re in a stage of so-called late capitalism — and maybe there’s some speculation on what the post-capitalist world is going to look like. That is not the book you chose to write. Why did you write this book about something which is a pretty important institution of modern capitalism, but one we maybe don’t know that much about?
Well, a couple of things, actually. First of all, there’s just an intellectual mystery I was fascinated by, which is all investing involves tricky bets on an uncertain future, but in venture capital investing, you are dealing with these startups and there are no data. There are no profits. You can’t work out the price-earnings ratio. There aren’t any earnings. You can’t do the price to book value, because there’s no book value. None of the normal stuff applies. So how do you even start to allocate capital? And I try to unravel that mystery by telling lots of case study stories about how the good venture capitalists do it. But the second thing kind of gets to your question, Jim, which is that I wanted to explain the impact of venture capital on society and on economies. I believe it’s fundamentally positive.
I think this notion that we’re in late capitalism or capitalism is failing misses the fact that actually corporate forms and financial arrangements always evolve. I think there’s sort of an optical illusion created by the fact that Adam Smith came up with the idea of the invisible hand in the late 18th century and market price signals have not changed. So we think of capitalism as almost a static thing, but that’s wrong. The business side of capitalism, the shape of corporations, has changed a lot over the years. Most of those changes are driven by a combination of technological shifts and then on top of that financial shifts. So when you suddenly had an argument for having scale in corporations, because steam and electricity rail had been invented, then you have the invention of the joint-stock limited liability company in the late 19th century.
And that’s why you get big companies. Then equally with the arrival of the personal computer in the 1970s and early 1980s, there’s suddenly a case for de-layering the company, un-bundling, re-engineering. And those things are made possible by a combination of junk bonds and the leverage buyout. I think now we’re in a phase where capitalism is not late stage, but just the new phase of it is that we have lots of intangible capital and to value intangible capital, things like software and brands and patents and business processes and customer relationships. All of those things are valued only by investors if the investors are pretty close to the company, because you can’t see those things by reading the financial disclosure. So you need to be a hands-on investor and that’s where venture capital and private equity come into play. I think that’s just the new optimal type of finance for a new phase of corporate structure because of this rise of intangible goods.
When people think about venture capital, they think about them funding these disruptive companies that will disrupt a particular field in the technology sector, but you’re saying they also disrupt the basic capitalist structure of a country. They’re disrupting what the corporate form looks like and it is not a static thing and that is ongoing today.
Yeah, and then I’d also say something in addition to that, which is that Ronald Coase got the Nobel prize partly for this distinction between price signals on the one hand, that coordinate human economic activity through the arm’s length transactions that generate price signals and tell people what the market wants more of, what the market wants less of. That’s one kind of organization. Then the other kind that Coase pointed to in his Nobel work was the corporation, a way of internalizing that organizational structure. You have top-down management and so forth. I think that venture capital is really interesting because it’s like a middle ground between the two.
On the one hand, a venture capitalist is like a strategic sort of planning department inside a big corporation because the venture capitalist is deciding which experiments get funded, which kinds of research get assets and get talented people, and is shifting money and people and ideas around. That’s a bit like a sort of corporate organization function, but then also there’s a price signal because venture capitalists tend to fund things for six or nine months. And then if the startup doesn’t reach its next milestone, there’ll be a price discovery thing where nobody wants to pay the price for the next tranche of the startup and then it’ll just be closed down. So I think it combines price signals and corporate strategy. And that’s another sense in which venture capital is super interesting as a new kind of capitalist form.
If someone asked me to name the ultimate American capitalist, I think I’d say Warren Buffet. And when I think about Warren Buffet, I think about somebody pouring through lots of annual reports. And then the other day I was following a venture capitalist on Twitter who said, “I’ve never looked at a balance sheet. I’m not sure I could read an income statement.” If that’s not where the skill lies in a venture capitalist, like the Warren Buffet set of skills, then how do they do what they do? Would Warren Buffet make a good venture capitalist?
Well, Jim, that’s a fantastic question. Because, first of all, it illustrates exactly this distinction I’m kind of getting at, where Warren Buffet can sit in Omaha, Nebraska with his financial statements and be very far away from where these companies are that he’s investing in because he can find out what he wants to know by looking at those financial disclosures and by analyzing them. And that’s great if you are a traditional value investor, but if you are doing the opposite kind of investing, in other words, growth and extremely high-charge growth in the form of these early-stage tech startups, it’s the opposite.
You’ve got to be hands on. And this is to get to your question. The way you value the software of a new software startup is that you are on the board of that startup. You yourself as the investor understands computer science. You can evaluate the code. On the disclosure, it might say “We have this project software, we’ve spent 10 million on it,” but it could be worth zero if the code is rubbish, or it could be worth a billion if it’s fantastic. Only the hands-on, venture capital–style investor who understands code and has actually got his hands into the company can tell the difference.
In the book you described the creed, if that’s the right word, of Silicon Valley as, “The belief that most social problems can be ameliorated by technological solutions. If only investors can be goaded to be sufficiently ambitious.” Has that creed held back any success in Silicon Valley? Is that still a valid creed that is useful?
There is a critique of this creed, which is that venture capitalists come along and they are too ambitious and they drive entrepreneurs to take too much risk. Then because the entrepreneur is being pushed to expand their company too fast, it blows up. They make mistakes that they would not have made if they had gone slower. Furthermore, this is not just bad for the entrepreneur, so the critique goes, it’s actually also bad for the economy at large, because what you are getting . . . Let’s say a disrupter like Uber comes along, is getting tons of venture capital dollars to go and disrupt the existing taxi market. So the incumbent taxi guys have to compete against ride-hailing companies that have subsidized capital. Basically, every time you got an Uber in the first 10 years of that company, you were paying for two-thirds of the fare. And the other one third of the cost was being met by some venture capitalist who was just eating the loss. And that’s not a level playing field, so goes the argument.
I actually pushed back against those critiques because it seems to me that incumbent companies have tons of other advantages. They often have the regulators in their pocket. They have existing brand. They have existing customer relationships. The incumbents always have the upper hand. And in some sense, if the challengers are getting subsidized venture capital dollars, that levels the playing field, it doesn’t distort it. So, I’m not persuaded that it’s a bad thing to push for scale. Of course, some entrepreneurs don’t want to do that kind of bit-scaling growth. That’s perfectly within their rights. They can refuse venture capital dollars. And I quote one venture capitalist in my book saying, “I sell rocket fuel. Some people don’t want to build a rocket, that’s up to them. They shouldn’t take VC dollars and that’s fine. But for the really ambitious companies, it’s the best way to go.”
You mentioned the rocket fuel, which has been very important to the success of the American tech sector in Silicon Valley. It’s not just the funding that’s important to the success. They do a lot of other things other than just supply capital, right?
Yeah. They get involved in the company. They advise it on how to scale up. Typically, a startup founder might have a fantastic product idea, but often even if it’s a good idea, it’s going to have to be adjusted and adapted before it really fits the market. But on top of that, once you get the product market fit you’ve got to then scale up your sales operation, and many founders haven’t done that before. Whereas venture capitalists have seen countless other examples of startups that have gone through the growth process and they know how to help that to happen. So, they act as mentors and coaches. They often get involved in making the first four or five hires.
They would source people from their own network of useful engineers or sales executives who could be brought into the startup. They would help to interview those people. And importantly, the brand of the venture capitalist is reassuring to those early-stage hires because it’s risky to join a startup. Startups fail all the time. I’m always struck by the story that Eric Schmidt told me about why he agreed to be the chief executive of Google. It was a risky move for him. He’d been the chief executive of Intuit, another company, before. And he was joining this startup called Google where the two founders were these punks. Larry and Sergey, PhD students who were incredibly arrogant and had very low regard for anybody over the age of 30.
In fact, Larry and Sergey had admitted that the only CEO they would accept would be Steve Jobs and he was not available. So for Eric Schmidt to take that job was risky, and the only reason he agreed to do is that the venture capitalist involved, John Doerr said, “Look, Eric. Yeah, it’s true. It’s risky. They might fire you. But if they do, I’ve got your back and I’ll slot you in as a CEO to some other company because my venture capital partnership, Kleiner Perkins, is funding new startups all the time.” So the venture capitalists de-risk the choice of joining a startup both for the entrepreneur who starts the company because capital is provided, so you take the risk with somebody else’s capital, and for the early hires. I call venture capital a machine for manufacturing courage.
Well, I think what that story also highlights is how it creates the connective tissue between the company and potential executives. Everywhere, they help create those linkages that make it all work. Every city in America, every state would love to have their own Silicon Valley, and even having lot of the ingredients doesn’t quite work unless you can all connect them together in the right recipe. And venture capital is pretty important for making sure that happens.
Yeah, and you could see this as actually quite a sort of leveling the mechanism because it’s a cliche to say of somebody who does well in life. “Oh, well this person did well because they had money and connections.” Venture capital is the machine for delivering money and connections. The job of the venture capitalist is to find somebody who’s really smart. Who’s got a good idea for a new product and bring those money and connections to that person. So venture capital is often accused of perpetuating inequality and one can debate that, but in some sense it’s a super democratic, super leveling, super egalitarian thing.
Is there a potential real competitor for the Bay Area? Especially since there are governance problems in San Francisco. We have remote work. We’ve heard a lot about Miami becoming the new Silicon Valley. Do you see a legitimate competitor to the Bay Area?
Yes, I do. I think what’s happened in the last 10 years is that the formula for creating Silicon Valley, which in my view is really about the venture capitalist coming along and de-risking entrepreneurship, has been understood. And the process of spreading it out to other regions has been tried and it’s worked. The biggest example is China where Silicon Valley venture capitalists went around 15, 20 years ago and they successfully built up an amazing digital economy in China. It’s not about the government’s long-term planning. Don’t believe that story. It’s about the fact that Baidu, Alibaba, Tencent, Ctrip — all of these early Chinese digital companies received American venture-style investments and were structured by American Silicon Valley lawyers with dispute settlement under New York law, a Cayman Islands parent, and the ability to issue Silicon Valley–style stock options to the early employees.
It was just basically taking the Silicon Valley formula and taking it to China. Now it’s now happening all over the place. Silicon Valley is taking its model to India. Since my book has come out, I’ve been flooded with emails from venture capitalists in India. It’s coming to Europe. It’s coming to Latin America, and it’s spreading within the United States. I absolutely do think that Austin, Miami, New York, Boston, maybe Chicago can do this. And what you need is, you need to have some flow of technical people coming out of preferably a good local university. Texas has a bunch of those: Rice as well as Texas A&M. I think there’s enough of a flow of people there. I don’t see why it shouldn’t be true of Pittsburgh with Carnegie Mellon. And I think once you’ve got that flow of technical people, if the venture capitalists move to them and set up shop locally, you can build an ecosystem that’ll flourish.
You can have a successful ecosystem without it being Silicon Valley. Just because you cannot duplicate something as wildly successful, does not mean you have not created a successful tech ecosystem. But what about more of the left-behind areas in the United States? There have been efforts to create more innovation hubs there. Can’t we have successful clusters in the middle of Ohio or pick whatever state you want?
I think this is where there is some bad news, because although what you were saying earlier, that you don’t have to be as successful as Silicon Valley to be successful, is sort of true, there’s an important saying in venture capital, “If it’s not 10x different, it’s not different.” If you are trying to persuade people to switch from some product they understand, or some service they’re familiar with, to use a completely different product, which is going to annoy them initially because they don’t know how to work it and it’s going to be expensive, they’re going to buy it and maybe it’ll just be a waste of their money because it doesn’t really work — to get over those hurdles, you’ve got to deliver something that’s really quite seriously new. Then if you succeeded in delivering that thing, it’s very difficult to do it, but if you succeed, you’re normally going to get exponential growth. And it comes back to what we began with, the power law idea.
Stuff either really works big time and goes 10x, 20x your investment or it fails. This is a skewed distribution. And so it’s difficult to say, “I’m trying to do Silicon Valley light. I’m going to do some startups, but I’m not going to be too ambitious about it.” It’s kind of an all-or-nothing game. You’ve got to really shoot for the big ambitious outcomes if you’re going to break through and do that 10x different and achieve success. I think the bad news is that the more humble vision for a cluster is probably not going to work, but what can work is that in surprising places, you do get ambitious people who break out. And so a great example is a company called UiPath, which is the global leader in robotic process automation software, which is a big category.
And it was begun in Romania. The guy who began it actually taught himself to code without a computer because there weren’t enough in Romania at the time. He couldn’t get one, so he used some secondhand computer manual, which he borrowed from the library or something, and taught himself how to code before he had a computer. It can happen in the most crazy places. There is Drive Capital, which is a Sequoia spin-out in Ohio. But I think the point is, it’s not looking for sort of reasonably good startups, even if it’s in Ohio. It’s looking for the breakouts.
Given what you’ve said though, and given the title of the book, we should not expect to have lots of successful tech hubs evenly distributed across the United States. We should expect there to be a few of them, maybe more than we have now. And they may arise, even though there are other things government can do, for sort of happenstance reasons.
Yeah, I think the power law idea is useful to understanding this issue. The power law operates on multiple levels, which is why it’s such a cool concept and it’s kind of the secret sauce of the whole sector. All startups, whether they have anything to do with tech or not, just the fact is that many startups fail and then a few do well. It operates on the level of technologies. So, when a technology really works, it tends to grow at this exponential rate. Think about Moore’s law for the way that semiconductors grow in power, doubling in power every couple of years.
Then it’s true also of things like the internet networks where you get these network effects, where the value of the network rises as the square of the number of users. And Metcalfe’s law, which is that network law on top of Moore’s law. So you can have both. If you think about an internet router, it’s got Moore’s law going on because it’s using semiconductors inside the router. It’s reaping the benefit of the router becoming more powerful because the semiconductors are more powerful. On top of that is getting the benefit of Metcalfe’s law because the internet is multiplying the value or squaring the value, as more users join. It’s got that startup effect anyway of a power law effect.
Then on top of that, you get investors who invest in companies building routers. And those investors are going to experience power law effects in their portfolio. Because of that, some venture capitalists are going to reap most of the gains in the entire venture capital area. So, there’s a power law effect if you distinguish the outlier of venture capital firms from the rest of them. And there is this final effect, which you’ve just alluded to, which is that there will be some clusters that dominate. It doesn’t mean there has to be one cluster in America. China has some activity in Beijing, some activity in Shanghai, some activity in Hong Kong, and some in Hangzhou and Xinjiang. So I think it can be distributed, but it’s not going to evenly distributed.
What role do you want government to play here? Do you want subsidies? Do you want tax breaks? Is it just very basic kinds of public good stuff? Investment, lots of R&D research to create that feedstock of ideas? What is the role of government?
I think the best illustration of how to set up this answer is the internet. The people who really believe in industrial policies say the government created the internet. They say DARPA created the internet. This is both true and sort of misleading. True in the sense that, yes, DARPA funded the early internet and built it to a point where there were 50,000 users or something. Who created the internet we know today, which is part of everybody’s life and has basically changed the way we live? Venture capital did. Venture capital discovered a sort of hobbyistic little company called UUNET, which was building routers that connected engineers who wanted to hack into the government internet. Venture capital got behind that and scaled it to the point where it just basically connected up America. That’s the story.
And then venture capital, by the way, created Netscape, which was the thing that allowed the internet to be useful to people who didn’t want to type in complicated digits every time they wanted to access a website. I think the answer to your question about the role of government is that early-stage basic science is the role of government. I’m extremely in favor of more for the National Science Foundation. I think it’s great to have more funding for scientific education in K–12. For post-docs, for people who are doing engineering and physics and math: I think we fund life sciences very well with NIH, but we don’t fund the NSF enough. But I’m not in favor of the government trying to pick winners when you get to commercializing this stuff. So I would focus on the early stage.
Courtesy: (AEI.org)
