When the US joined World War Two, it set back a number of peacetime R&D projects. A team at Bell Labs had been studying some interesting properties of semiconducting substances, with the hope that they might find a way to replace the then-ubiquitous vacuum tube. But the exigencies of wartime complicated their plans—for one thing, the only supplier of sufficiently pure silicon was a German company—so they took a break to focus on wartime applications like radar. By the time they returned to regular work, the war was won, the US industrial base had dramatically ramped up (now DuPont, not German suppliers, made the world’s purest silicon), and the team had acquired an encyclopedic knowledge of how substances like silicon and germanium behave.
Two years later, the transistor was born.
Crises radically reshape priorities; they cancel some projects, and accelerate others. But another effect they have is to make people more conscious of the future. To struggle through a crisis is, implicitly, to view the future as a point very much worth getting to. The Bell Labs team certainly helped build a better postwar future, and arguably Covid-19 has pushed people in the same direction.
We shouldn’t expect the results of this to be visible just yet. Scientific progress is visible on a lag: while the New York Times did mention the invention of transistors, it was midway through a column called "news of the radio," hardly an accurate assessment of the impact of one of the twentieth century’s great inventions. Fortunately, there are more real-time ways to approximate people’s attitudes towards the future—the stock market is a real-time dollar-weighted poll about what kinds of companies will matter, and how that’s changing.
One of the surprising consequences of the Covid-19 pandemic was that, after a brief and fearsome decline, overall asset prices rose. That’s partly an artifact of how policymakers responded to the pandemic; pumping liquidity into the market does help offset supply shocks, and some of that money finds its way into speculative vehicles. But even within the market, there’s been a striking rise in investor interest in electric vehicles, autonomous cars, AI, and software companies ranging from consumer-facing companies like Facebook and Google to complex enterprise products like Snowflake.
The market’s judgments always have to be taken judiciously. Humans are imperfect judges of the present, not to mention the future. And as the recent GameStop fireworks demonstrate, sometimes prices can be driven more by technical aspects of market structure than by a cold and calculating assessment of the net present value of future cash flows. Even GameStop’s run-up, though, was born out of forward-looking analysis of a business, not gambling. The original thesis behind buying GameStop, before it turned into a social movement devoted to punishing hedge funds, was an argument that the company was fundamentally underpriced—because the market had missed its opportunity to transcend brick-and-mortar retailing and digitize its business!
There are definite precedents for extrapolating about new concerns from stock prices. Defense stocks rise when war is rumored to be imminent, for example, and cyclical stocks’ performance tends to change ahead of macro data on the economic cycle. More narrowly, the economist Armen Alchian deduced that hydrogen bombs use lithium by tracking the stock prices of mining companies.
Today, the world’s most valuable automaker is Tesla, with a market capitalization of $827bn, compared to $232bn for runner-up Toyota. Tesla isn’t valued based on its current production (just under 500,000 vehicles annually, compared to Toyota’s 9.2 million) or high margins (it eked out a net profit margin of just under 2%, less than half of Toyota’s results). Instead, Tesla is valued based on three forward-looking intangibles: it’s a pure-play electric vehicle company, its brand name is synonymous with clean energy rather than the more mixed reaction General Motors or Hyundai might engender, and its charismatic CEO has been able to recruit cult-like adherents to his vision of sustainable transportation and a multiplanetary species.
Calling Tesla is a cult isn’t pejorative, just descriptive: any organization that successfully accomplishes something that is widely believed to be impossible has to have distinctive beliefs, and any group of people who behave in unusual ways because of shared beliefs can be reasonably described as a cult. Some companies use their cultish aspects in harmful ways, but cults are a social design pattern that shows up over and over again, in successful companies and political movements. Cult-like traits are more common with companies that are growing fast—it would be hard for a steel mill or a local bank to engender this kind of behavior in its employees. It’s a way to focus everyone’s energy on the future, and avoid distracting questions about whether or not that future is viable—a way to raise the variance of outcomes, which makes extreme upside scenarios possible while increasing the odds of failure.
Tesla is not the only futuristic vehicle stock to be accorded a high value by the stock market. In fact, it’s arguably among the more mature. There are at least earnings to put a price/earnings multiple on, whereas many of the more recent EV companies are at an earlier stage than that. Luminar Technologies, for example, went public through a reverse merger late last year, and currently has a market value of almost $11bn. The company has minimal sales ($11m over the last nine months) and is still pouring money into R&D. But LiDAR appears to be the most promising way to get cars to full autonomy, so investors are willing to value it based on the chance that a) autonomous vehicles will happen, b) they’ll use LiDAR, c) they’ll use Luminar’s systems to do it, and d) Luminar will be able to earn acceptable margins selling it.
The joint probability of all of those possibilities is low if they’re independent: if there’s a 10% chance of autonomous vehicles, a 10% chance they’ll use LiDAR, a 10% chance that LiDAR-using AVs will use Luminar’s technology, and a 10% chance that Luminar will get good margins, then the odds of Luminar being a good investment work out to 0.01%. But the more ambitious a company is, the more its job is to make those variables conditional instead: the closer a company gets to being the only way a given technology can happen, the more technology risk becomes synonymous with business risk, which compresses the overall range of outcomes. It also solves for the viable-business condition: if there’s just one company that can make AVs possible, then that company will have the pricing power necessary to make them profitable, too.
This dynamic actually works in two directions: first, it means that the odds of Luminar selling LiDAR conditional on LiDAR becoming ubiquitous are higher, because the latter is most probably going to happen if the former is true. And second, it’s a recruiting tool: if there’s one company that has a reasonable shot at accomplishing something, and it’s a desirable goal, then the company has a monopoly on the kinds of employees who can achieve that goal. This is a point Peter Thiel articulates in Zero to One, and it’s one of the reasons technology companies have such a skewed distribution of outcomes. They articulate a variant view, which means they attract people who share that variant view—and since the argument is settled by technology and economics, they don’t have to persuade the rest of the world, just to offer a better product.
The rise of special purpose acquisition vehicles, or SPACs, is a general testament to a more forward-looking market. In a conventional IPO, an operating company sells shares to the public; with a SPAC, an empty shell company goes public, and then identifies a private company to merge with. Due to a quirk in securities laws, a traditional IPO prospectus only shows a company’s backwards-looking estimates, and makes heavily-qualified statements about the future. A company that goes public through a SPAC is technically engaging in a merger, rather than an IPO, and the rules are different. When a public company buys another company, securities laws allow it to talk about that company’s anticipated growth, or the likely cost savings of the merger. Similarly, SPAC offerings can talk up a company’s long-term prospects, and even make exact estimates of future revenue.
SPACs have existed for years, but they exploded in popularity in 2020. Of the 466 SPAC companies that went public from 2011 through 2020, 248 of them went public in 2020 alone. A number of technical forces drove this—a large number of private companies looking for acquisitions, investors eager to get into any growth company early, some technicalities around SPAC issuance that make them attractive to specialist hedge funds. But the key driver of excitement about SPACs is that they can take companies public based on a future-focused outlook.
When Virgin Galactic went public by merging with Chamath Palihapitiya’s Social Capital Hedosophia Holdings, the company, which has taken deposits but generated minimal revenue, was able to project $590 million in annual revenue in the year 2023, and over a quarter of a billion dollars in pretax, pre-depreciation earnings. While this was optimistic, it’s also demonstrative: a company like Virgin Galactic would have been almost impossible to take public in a conventional way, because backwards-looking financial statements showed only losses. it may not work out as a business, or live up to its projections, but those projections got more plausible once it had access to public markets for funding.
High valuations for money-losing, often pre-revenue companies remind people of the dot-com bubble, and it’s worth putting that bubble in perspective. Someone who bought the market at its peak in March 2000 has earned a 6% compounded return since then. Disappointing, especially for stocks, but not a permanent loss of capital. In fact, some economists continued to argue after 2000 that "bubble" is a meaningless category.
Performance of the Nasdaq 100 index from the peak of the dot-com bubble to the present. Graph via KoyFin.com
And the dot-com bubble was predicated on the notion that the Internet would change the way people ate, shopped, dated, traveled, watched movies, and lived—all of which was completely true. Of the five most valuable companies in the world today, two (Google and Amazon) are Internet companies, one (Microsoft) is increasingly shifting its business to the Internet, and another, Apple, makes most of its money selling devices that people use to stay online all the time. Saudi Aramco is the only company in the top five that is not, in some sense, an Internet business.
So the dot-com bubble missed most of the details, but got the broader picture right. And the people who bought into the bubble at the absolute peak of the frenzy had disappointing investment returns, but still made money overall, a situation that would have struck someone in the depths of the dot-com hangover of 2002 as implausible.
Buying into bubbles is not good for the average person’s financial health, but that’s in part because bubbles end up being a wealth transfer mechanism: money from (relatively rich) shareholders gets turned into R&D, capital investment, and marketing to promote the resulting products. In some cases, that money is unrecoverable; the cost of catered sushi and live entertainment at dot-coms’ post-IPO parties did not add to humanity’s capital stock in any meaningful way. But in other cases, it’s a good investment. Level Three Communications spent itself into bankruptcy laying fiber in the 90s, before there was demand for it. Google bought some of that fiber for pennies on the dollar, and once digital cameras were more widely distributed, streaming video was cheaper because some of the capital costs had already been paid.
This is one of the subtle functions of investor over-enthusiasm: it encourages discrete components of the future to get built. If someone is raising funds for a self-driving car AI, Luminar’s market capitalization provides evidence that there will be cars that can make use of such an AI. A wildly risk-tolerant property developer who plans to build the first lunar resort can use Virgin Galactic’s value as proof that the resort will get visitors. It’s a bit like the phenomenon of business clusters—when engineers left Fairchild to start competing chip companies, they often started them near Fairchild, because that’s where the customers, capital, and potential colleagues were. Bubbles create the same kind of cluster in a totally different dimension, encouraging companies to locate their same plans in the same hypothetical future where adjacent startups are highly valued.
One unique aspect of Covid-19 is that, while most wars and natural disasters take up large amounts of time, the pandemic actually freed up time. Less commuting, in some cases less employment, less time spent at restaurants, on planes, or at the movies—for people who weren’t front-line workers or researchers, Covid added a large amount of unallocated time.
Mostly, this was reflected in the immense popularity of Tiger King.
But not in all cases. Covid also forced people to think about long-duration plans; if the length of the pandemic is uncertain, then any plan predicated on the world getting back to normal at some specific date in the future is probably a bad idea. And there are two ways to respond to this. One is to make extremely long-term plans that don’t require much external input—writing a novel, designing a programming language, developing expertise in an academic discipline, etc. But the other way not to bet on normality is to redefine it. In some sense, the world will never return to normal, because "normal" was a world where mRNA vaccines were only theoretically viable, where remote work was materially less viable and acceptable than in-office work, and where the car companies with the most access to capital were the ones that made internal combustion engines. Free time, abundant capital, and the impossibility of returning to the status quo have created a healthy bubble that’s betting on a better world.
Byrne Hobart is a finance blogger and writer of the newsletter The Diff. You can follow him on Twitter here.
Image thumbnail: "Trading Floor at the New York Stock Exchange during the Zendesk IPO" by Scott Beale is licensed under CC BY-NC-ND 2.0.