The past year of rampant inflation and energy system chaos is a clear indication that we need paradigmatic change. Any new economic system is going to be anchored by major scientific innovations; historically, spurring these technological transformations has required a mix of initial state action followed by entrepreneurial execution. Sebastian Mallaby’s The Power Law is a cutting, prescient analysis which argues that the individual who currently stands as a symbol of market excess—the venture capitalist—will be required for such a transformation.
Mallaby’s 2022 book recounts the origins and evolution of an industry that aims to upend the economic order while still working within its structural confines. Along the way, it punctures several myths. While the Silicon Valley press heralds founders as iconoclastic saviors, Mallaby illustrates how the top venture capitalists (VCs) mold them, systematically finding ways to get a seat at the table—if not making the table outright.
VCs—through capital allocation, ingenuity, and often straight-up guile—shape the parameters of the future. But the industry is not simply a history of big checkbooks. Mallaby emphasizes the social basis of innovation: the power of social networks in cultivating new ideas, resolving conflicts, and generating outsized economic returns. The network is select, and the gains from the innovation are not evenly distributed. But is that still the price to pay for technological progress?
An interview with Sebastian Mallaby
NIKHIL KALYANPUR: Over the course of writing the book, how did you change your thinking on the value of speculation?
SEBASTIAN MALLABY: If one understands speculation to be a somewhat reckless risk-taking attitude, hedge fund investors are not always speculators, whereas venture capitalists are. Hedge fund investors in public markets often have crunchy statistical bases for the bets they make. When they pile on big, they’re not taking crazy risks. They have very solid logic for making their trades.
A “power law” style of investing—investing under conditions intense uncertainty—is speculation on steroids. When two individuals walk into your office and say they’ve got a vision, you’re making a bet on whether that vision will come true. You always know you’re going to be at least as likely to be wrong as right. But you make the bet anyway, because you’re hoping for that right tail upside. Venture capital is highly speculative, requiring a kind of superhuman risk appetite.
NK: The VC model implies that there’s something wrong with the efficient markets hypothesis. Do you think we need private sector bubbles for big innovations to take place?
SM: Inefficient markets allow skilled investors to generate alpha—better-than-expected risk-adjusted returns. I do think markets are efficient to a first approximation. But they are not perfectly efficient, so active managers—hedge fund managers or venture capitalists—can generate alpha through skill.
Do you need bubbles? I don’t think you need bubbles to generate innovation, because entrepreneurs and VCs would want to develop new technologies even in the absence of market overshoots. But if you turn the question around, when you do have successful technologies, they almost always come with bubbles. There’s almost always a moment when the euphoria overshoots—the railways in the 1850s, bicycles in the 1890s, and the internet in the 1990s.
NK: One of the core tensions in the book is the diverging role of the investor and different market actors. At one level, we really want the venture capitalists to come out swinging and take risks, yet we want the public market investor—the hedge fund type or the institutional investor—to be the disciplinary force. What does that tell us about how our markets are operating today?
SM: A diversity of actors that have slightly different objective functions or analytical processes is ultimately a healthy thing. You get different types of investors with different levels of risk appetite, and different specialties in terms of which part of the economy they understand. That diversity is more likely to yield sound capital allocation than a monolithic system in which all the capital is allocated either by the government or by banks or by some other particular player.
NK: There are so many varying approaches just within the VC industry itself. Historically, you describe the initial venture capitalists as taking such an active role that they’re working with founders to eventually replace them and bring in new CEOs. We’ve also got the other extreme, characterized by Peter Thiel, where VCs just hand the money off and let the founders run. Is there something specific about the venture capital industry that allows for such different models to thrive at the same time?
SM: Venture capitalists invest in people and ideas, and those can be pretty diverse. Different personalities and technologies will come into your office. Peter Thiel is special in his professed reluctance to get involved with companies after he invests, but in many cases, the venture capitalist may need to bond quite deeply with the entrepreneur. They can’t bond with everybody. Maybe the personality types won’t mesh, maybe their skills are not complementary enough, maybe their strengths and weaknesses don’t match up.
We talk about startups as if they make up one bucket of things. But there’s a massive difference between a social media company and “deep tech”—people building advanced batteries or whatever it might be. In my book, I describe how different sorts of technology opportunities will come with different types of entrepreneurial personalities. The kinds of founders that you’ll back when looking to fund routers in the 1980s will be distinct from those getting into web 2.0 in 2004.
NK: Thiel seems to represent thinking in the other extreme, to the point where he argues that capitalism needs to avoid competition. Do you think that we require monopolies to make the VC model work? Would a more credible US antitrust policy harm the VC model?
SM: I think his argument is thought-provoking, but Thiel exaggerates for effect. Sure, every business wants to accumulate as much pricing power as possible. But one business’s power is another business’s problem, and VC-backed startups are often on the losing end of that.
If the US government were more aggressive about fighting tech monopolies, it would arguably be better for venture capital. VC is funding the challengers to the monopolies, and those challengers would have more of a chance to grow. I’ve talked to VCs who are quite bothered by the tech behemoths and would much rather have an antitrust policy that reigned them in. People talk about a “kill zone” around the tech giants—if you launch a startup that competes with Alphabet or Microsoft, they may clone your product. For example, Yelp has a long-standing campaign to persuade antitrust authorities to get tougher on Alphabet. And of course Microsoft was the target of the Department of Justice back in the 1990s for alleged predatory practices towards the upstart browser company, Netscape.
Now, that is just one side of the argument. There are many VC exits that take the form of mergers and acquisitions (M&A), but in many cases the acquired startup is too small to trigger regulatory scrutiny, and would be too small even under an enhanced M&A regime. Set against VCs’ desire for M&A exits is VCs’ desire to grow a startup as much as possible without a large competitor crushing it.
I think there’s also a bit of a myth that VCs are always pushing for early exits, including through M&A, and that therefore they must be on the side of lax antitrust. Sarah Frier’s book on Instagram documents how it wasn’t the VCs who were pushing for a sale to Facebook, but rather the founders. In my research, I found other cases where an acquisition offer comes in and the founder is initially tempted to take risk off the table, but the VC prefers going it alone for longer in hope of a larger exit.
NK: One of the big themes in your book is this progression towards the founder having more power. How much of the lionization of the founder do you attribute to the changes in the interest rate environment?
SM: Over the period covered in my book, capital had progressively less leverage over founders for two reasons. First, interest rates were falling, eventually leveling off at a very low base. Second, and probably more profoundly, technological changes made creating a company less capital-intensive. Both forces pointed in the same direction. Thanks to low interest rates, capital was cheap and plentiful, so the capital providers had less negotiating power. At the same time, founders needed less capital and therefore had even less reason to defer to capital providers. Between the low interest rate effect and the rise of cloud computing, I would say the latter was more important in the rise of founder power versus the VCs.
NK: You make the point that it’s neither market nor hierarchy, but the network that’s doing the work. There’s an image of Silicon Valley where anyone can come in and be creatively disruptive, but when looking at the figures you highlight, it seems that having an MBA is a prerequisite to joining one of these funds. These networks are quite closed—you even bring up the idea of the white man from Stanford basically being the architect of this world. Can we harness the power of these networks without reinforcing this elitism?
SM: Networks are super important. A key example is the Israeli tech scene, which is remarkably strong for such a small country. The success of tech there is linked to the military: a high proportion of people in the Israeli tech world pass through a special unit of the Israel Defense Forces. That, first of all, trains you to be a technologist, but more importantly, it means that everybody who’s graduated from that unit can check each other out. The network is so tight that only one degree of separation is needed to understand whether a person is to be trusted with capital.
Networks are tight and often narrow, but there’s an incentive to extend them. I would argue that VCs deliberately extend the network to make themselves more productive and profitable. Think about the PayPal Mafia—the former PayPal employees who went on to found companies like SpaceX, LinkedIn, and Yelp, among many others. In the same way, somebody coming out of Google may have a good shot of being a VC, because they will know people leaving Google to become founders. After Moderna’s success with mRNA technology, VC partnerships focusing on biotech will try to hire more PhDs in that field.
All these examples of deliberate network extension tell us that at some point, VCs will put aside that clubby white male Stanford elitism. In recent years, we’ve seen that happen with ethnic and immigrant networks. Chinese American, Indian American, and Persian American communities have all contributed to the tech industry, so of course VCs are hiring from those groups. Still, African American representation remains especially woeful.
NK: Do you think Silicon Valley in particular represents something unique about America?
SM: I think the success of Silicon Valley demonstrates the importance of the US law, particularly around non-competes. In the UK, when you hire somebody from another company, it could take four to six months of gardening leave to get them to join. If you’re a VC-backed startup with a runway of four to six months, that’s a problem. American VCs who come to the UK are often shocked by that barrier. In California, non-competes are non-enforceable, and this has contributed to the flourishing of the startup ecosystem in Silicon Valley.
US law regarding limited partnerships has also had a major influence. US venture partnerships are pass-through entities, meaning that investing partners, called the limited partners, pay tax on their capital gains, but the partnership itself is untaxed. The US partnership structure also allows for the use of different classes of stock: common stock for founders, preferred stock for investors, and stock options for employees. One of the reasons that Europe has been slow in developing its venture ecosystem is that, until recently in several countries, it was unattractive to grant employees stock options because the tax provisions were too harsh.
The attractiveness of the US legal model is illustrated by China. Starting in the late 1990s, when companies such as Alibaba began to take investments from western financiers, Chinese startups borrowed the entire US playbook. With the help of Silicon Valley lawyers, they were set up so that they used New York law for dispute resolution and could issue employee stock options. This would have been impossible for a normal Chinese firm. But Silicon Valley lawyers created Cayman Island parents for the Chinese startups, and these parents could issue stock options, preferred stock, and so forth.
NK: You’ve documented the connections between US and Chinese VCs with great depth—the links between the early founders, the different models adopted, and the laws eventually adopted. How different would the Chinese venture capital industry look if it were more state-directed as opposed to American VC-inspired?
SM: Until the late 1990s, venture-backed tech startups that made a big impact were highly concentrated not just in America, but in Northern California. Absent this special formula of risk capital, you don’t get world class tech companies. Before the recent crop of unicorns in Europe, the biggest tech company and almost the only large-scale software company in the region was SAP. A continent with rich consumers and lots of trained engineers simply didn’t generate that kind of startup because of a lack of risk capital. If there had been the same lack of venture capital in China, it wouldn’t have mattered that China was becoming richer. Lots of talented Chinese engineering graduates would not have translated into the formation of companies like Tencent and Alibaba without American input.
In the late 1990s, Shirley Lin of Goldman Sachs, the first investor to back Alibaba, also advised on government-backed attempts to build technology in semiconductors. SMIC was going to be the rival to TSMC. The strategy didn’t work, and twenty-five years later, China is still trying to build a strong semiconductor sector.
But the other half of Lin’s work—backing startups like Alibaba—clearly did work. I argue in the book that western investment and western legal structuring made all the difference to that success. Jack Ma built a world class company by attracting world class talent. He hired Joe Tsai to be the chief operating officer and chief financial officer, which would have been impossible without the ability to offer Tsai equity options. Likewise, he hired John Wu, the chief technology officer at Yahoo. Wu explained to me how his decision to quit a prestigious Valley firm and join the upstart Alibaba boiled down to the options package that Ma was able to offer.
NK: How should we then conceive of the role of the state in the innovation process? The book outlines a few different models—the Mazzucato view of needing the state, or a competing view that the state is only required for large fundamental innovations, and the market can function after that. Yet the conclusion of your book focuses on the state’s role in tax policy.
SM: Investing in basic research, science, and education is a government role. Connected to that is the training of the scientists who can become professors or entrepreneurs. Certain kinds of big tech, like building a semiconductor fab, are too capital-intensive for normal venture capital to back. The government must also provide good intellectual property (IP) rules to govern the technology transfer out of universities. In the United States, the Bayh-Dole Act of 1980 allowed inventions generated with the help of federal grants to become the property of the inventors, who could then form startups to commercialize the inventions or license the IP to entrepreneurs.
I argue in my book that putting government money into VC funds is less effective on the whole than giving tax assistance to VC funds. I favor tax rules that facilitate the use of employee stock options and allow venture partnerships to be pass-through entities, so that capital gains aren’t taxed twice. I am not a fan of the carried interest loophole—this takes tax subsidies too far.
However, I think the idea that venture capital’s role is unimportant and indeed suspect, because it’s reaping subsidies from government investment in science, is too extreme. This is a case of reacting against the mistaken view that venture capitalists are the sole agents of innovation and embracing the opposite mistake of saying that the state is the sole agent of innovation.
NK: Do you think there’s more room for collaboration between VCs and government? There’s also a redistribution question in that relationship: at what point should the government be trying to equalize the gains out of these technologies?
These companies see huge gains, which are some way are needed for venture capitalistic risk, but there’s also a large societal cost. For example, Silicon Valley has been a substantial driving force behind the rise of the billionaire class. What billionaires do to democracies is an open question, but there’s a compelling case that more billionaires leads to more capture and less democratic representation.
SM: I don’t think the government’s record as a limited partner is particularly good. When EU funds dominate capital allocation for startups and they don’t have to generate a market rate of return, the incentives are messed up. The same is true for the state “guidance funds” in China. I’m comfortable with a commonsensical recognition that government is very important in some things, and risk capital in the private sector is very important for other things. At the same time, I agree with your point on redistribution. I think the right remedy for that is a higher personal tax, and much higher wealth and inheritance taxes. I’m not keen on more billionaires.
NK: At what point do you think inheritance tax or a potential higher income tax is going to change the innovation landscape in the US? One argument against higher taxes is that when it comes to billionaires, people who are really good at picking winners have the money to pick winners. We could have Bill Gates pushing innovation in the climate tech space.
SM: I think that the winners could win a fraction of what they’re getting, and they would still be highly incentivized to do what they do. I don’t think higher taxes will dampen innovation. But we can’t have it both ways when it comes to billionaire philanthropy. If we’re going to argue that billionaires threaten state capture, we can’t make an exception for the billionaire we like. Bill Gates is doing great stuff for climate tech, but we’re going to have to tax him too.
I don’t think taxing the billionaire philanthropists will shut down an irreplaceable driver of innovation. We already have savings institutions that allocate to venture capital, including university endowments and pension funds, which arguably have their own useful public purposes. I think one can be anti-individual billionaire, but pro-innovation.