The price of vision: What a century of visionaries taught us about markets, multiples and Musk

VanEck has spent more than seventy years investing across market cycles. In that time, we have observed many founders who spoke the language of science fiction go on to deliver it, and some who didn’t. We have watched respected sell and buy side analysts declare paradigm-shifting companies overvalued at one-tenth of what they eventually became. We have also witnessed the reverse.

The lesson we keep relearning is not that the market is irrational. It is that the market is rational about the wrong time horizon.

What follows is not a recommendation. It is a reckoning.

The visionary premium: a concept that predates the stock ticker

Let’s go back to 1903. Henry Ford had already failed twice. When he incorporated the Ford Motor Company in June of that year, Wall Street was not paying attention. The automobile was a curiosity for the wealthy, not a utility for the masses. Capital was sceptical. Ford was not. He raised US$28,000 from twelve investors, a figure that, adjusted for inflation, is not large by any modern standard. The markets were not pricing the Model T. They were pricing the horseless carriage business.

By 1916, Ford Motor was generating revenues of US$200 million. By 1925, a third of all cars on earth wore the Ford badge. The original investors had earned returns that no valuation framework of 1903 could have predicted, because no valuation framework of 1903 had a category for what Ford was building. It was not a transport company; it was an industrial civilisation.

This was perhaps the original sin of fundamental securities analysis: it prices the thing being sold today, not the system being constructed for tomorrow.

Lesson 2, fast forward 20-odd years to 1926. Radio Corporation of America (RCA), we think, was the Nvidia of its age. It had no earnings to speak of. It traded at price-to-earnings multiples that scandalised conservative investors of the era. Roger Babson, one of the era's most prominent market forecasters, warned that it was speculative excess. By 1929, RCA had risen 9,000% from its post-war lows. Then it crashed 97% in the Depression. The technology was real. The fundamentals eventually became real. Timing was everything, and the timing was devastating.

What RCA teaches us is not that visionary companies are bad investments. It is that the spread between vision and reality can span decades, and that the investor who cannot survive that spread, financially or psychologically, should not be in the business of pricing futures.

The great postwar listings: when IBM and Xerox taught markets to price intelligence

The postwar era produced a different kind of visionary IPO: companies that listed not as raw speculation but as institutional anchors. International Business Machines (IBM), listed since the 1910s as the Computing-Tabulating-Recording Company before taking the IBM moniker in 1924, experienced its transformation into a computing giant. It was repriced gradually across the 1950s and 60s.

Xerox's 1961 New York Stock Exchange debut is perhaps the more instructive. It had been public since 1936, but its 914 copier, the world’s first successful commercial plain-paper copier, had only recently become a phenomenon following its 1959 launch. Xerox traded at multiples that made conservative investors nervous. Within a decade, a US$10,000 investment was worth over US$1 million.

What separated Xerox from RCA was not the quality of the technology. It was the timing of the commercial cycle. Xerox listed when the product had been validated, when enterprises were buying, and when the S-curve of adoption was steepening. The market rewarded that specificity. It priced information, not aspiration.

Intel's 1971 IPO followed a similar template: a company with a product, a customer, a margin and an identifiable, if not yet enormous, market. At US$23.50 per share, raising US$6.8 million, it was conservatively priced. Within fifteen years, that conservatism looked like criminal undervaluation.

This is the paradox the market has never resolved: conservative pricing of revolutionary companies is mispricing.

The internet era: when the market priced the infrastructure before the revenue

Nothing tested the limits of valuation theory more brutally than the period between 1996 and 2002. The Nasdaq rose 400% in four years.

It then fell 78%. The companies that survived, including Amazon, Google and Salesforce, are now considered some of the most valuable companies (though Salesforce has had a tough time of late, falling on the SaaS-pocalypse). The companies that didn't, Pets.com, Webvan, eToys, and so many more, are business school cautionary tales.

But here is the number that never gets cited: Amazon's stock in 2001, at the depth of the dot-com collapse, traded at around US$6. By 2025, it had compounded at over 30% annually for twenty-four years. The analysts who called Amazon overvalued in 1999 were right about the valuation. They were wrong about the company.

Google's 2004 IPO is perhaps the best case study in this mispricing. Google’s shares were priced at US$85 per share via a Dutch auction¹, specifically designed to circumvent the traditional investment banking allocation process and fairly price the company. The banks hated it. The market was confused by it. Within eighteen months, shares were above US$400. Within ten years, adjusted for share splits, those US$85 shares represented a return of more than 1,200%. The market had priced a search engine like AltaVista. Sergey Brin and Larry Page had built an advertising utility that would eventually govern how humanity finds information.

Tesla and the birth of the narrative premium

In June 2010, Tesla Motors listed on the Nasdaq at US$17 per share, raising US$226 million. It was the first American automaker to go public since Ford finally listed in 1956. In 2010, Tesla had delivered fewer than 1,000 vehicles. It had never turned a profit. Its debt load was substantial, its production capacity limited, and its supply chain aspirational. Every traditional automotive analyst who examined the prospectus came to the same conclusion: this company is worth a fraction of its offering price.

They were right about the fundamentals. They were wrong about the future.

By 2012, Tesla had successfully launched the Model S, the car that transformed the company from a curiosity into a contender. The share price had oscillated violently, dropping to around US$23 in mid-2012 before the Model S reviews hit, then climbing toward US$34 by year's end. This was the moment the Tesla trade was made or lost: the investors who understood that they were not buying an automaker but a proof of concept for a post-combustion engine transportation ecosystem held. Those who were marking to traditional automotive multiples had already sold.

Tesla, at its 2021 peak, was valued at over US$1.2 trillion, more than Toyota, Volkswagen, Ford, GM and every other automaker on earth combined. Price-to-earnings multiples in the hundreds. Price-to-sales multiples that would have made dot-com-era companies blush. But the company was producing 900,000 vehicles a year, was cash-flow positive and had built a global charging infrastructure that no competitor could replicate overnight. The premium was not fantasy. It was, in part, an option on energy storage, autonomous driving and grid services that hadn't yet been priced into the automotive industry.

Elon Musk did something no automotive executive had done since the war: he persuaded the market to value a car company like a technology company. Whether that was genius or manipulation or both, the result was one of the most consequential capital formations in modern financial history.

SpaceX and the US$1.77 trillion question

When SpaceX listed on Nasdaq last week at an IPO valuation of US$1.77 trillion (and closed with a US$2.1 trillion valuation), raising US$75 billion in what is now the largest IPO in history. It dwarfed Saudi Aramco's US$29.4 billion in 2019, and the financial community reacted with its usual combination of awe and discomfort. Awe at the audacity. Discomfort at the math.

We think it is worth trying to do the math. Because we think some of it has been lost to the hype.

SpaceX's Starlink division, its satellite internet service, had roughly 9 million subscribers globally entering the year, climbing to 10.3 million by the end of the first quarter. Its user base spans people in remote rural communities, private enterprises and government agencies, with major markets in the United States, Brazil, and Australia. Starlink generated approximately US$11.4 billion in revenue in 2025, up nearly 50% year over year, and now accounts for more than 60% of SpaceX's total revenue. This is all from a single product line. The company's launch business, which has pioneered Falcon 9 and Falcon Heavy, and the rapidly scaling Starship, represents a fundamentally different cost structure for orbital access than any of Space X’s competitors have yet achieved. The price to orbit per kilogram has dropped more than 95% since SpaceX launched. That cost compression is not incremental, and it could potentially prove civilisational.

At US$1.77 trillion, or now US$2.1 trillion, SpaceX trades at roughly 60 times projected 2026 revenue, a multiple that is aggressive by any conventional measure. But if the last one hundred and twenty odd years have taught us, the comparables being Boeing, Lockheed Martin, and Airbus may be the wrong ones. A closer analogue could well be Amazon Web Services in 2006: an infrastructure business with near-zero marginal cost at scale, serving a market with a potential ceiling, on an optimistic hindsight reading is vast.

It’s important to be mindful that the bear case is real and deserves respect. Regulatory risk for a company dependent on launch licenses and spectrum rights is real. Musk's political engagements have introduced reputational volatility that is difficult to model. The Starship development program has a failure rate that would bankrupt any publicly traded aerospace company operating under conventional board governance. Finally, the transition from private to public markets introduces accountability structures that Musk has historically found constraining.

The bull case is equally real: if Starlink reaches 30 million subscribers, a number that we think is plausible within five years at current growth rates, the recurring revenue justifies a large part of the current valuation. If Starship achieves the launch cadence that the SpaceX team projects, the margin structure of the launch business transforms into something no aerospace analyst has a model for, because nothing like it has existed before.

If our experience offers any lesson, it is that when something has no precedent, the closest available comparable can mislead.

The pattern that markets keep refusing to learn

Looking across a century and a quarter of visionary companies, from Ford to RCA to Xerox to Intel to Amazon to Google to Tesla to SpaceX, several patterns we think are observable.

First: the market always prices the current product, not the future platform. Ford's investors were pricing transport, but they were inadvertently funding the industrial supply chain. Google's IPO investors were pricing search, but they were inadvertently funding digital advertising, cloud computing and autonomous vehicles.

Second: the discount rate applied to visionary companies is almost always too high. The standard models penalise uncertainty. But in the case of companies with platform potential, uncertainty is not symmetric; the upside tail is longer and fatter than the downside tail, because the downside is bounded by the balance sheet and the upside is bounded by the size of the market being disrupted, or potentially unbounded by a new market.

Third: the investors who win the visionary trade are not those who correctly model the fundamentals. They are those who correctly identify the category they are watching and that it could become infrastructure that will become invisible and indispensable. The investors who analysed Amazon as a bookstore lost. The investors who saw it as logistics and cloud won. Do you see SpaceX as a rocket company or the potential backbone of low-orbit connectivity infrastructure?

Fourth: the visionary premium compresses, eventually. Every company noted above, if it survived, eventually traded at multiples that reflected earned revenues rather than projected ones. The question for any investor entering a visionary position, we think, should not be whether the valuation is justified by today's fundamentals; it is whether the company will live long enough and execute well enough to grow into its multiple.

Most don't.

The market always catches up. The question is whether you can wait

Henry Ford was worth US$188 billion in today's dollars at his peak. John D. Rockefeller, whose Standard Oil grew from its 1870 founding into the dominant force in American oil, was worth nearly US$400 billion in today's terms. The visionaries of the twentieth century made their investors rich not because the market correctly priced them at inception, but because the market eventually had no choice but to acknowledge what they had built.

Elon Musk, irrespective of what you think of him, may be doing what Ford and Rockefeller did: building infrastructure that could become load bearing for civilisation. Tesla's charging network is the most extensive EV infrastructure in the developed world. Starlink is the primary internet provider for maritime, aviation and remote-area users across six continents. In recent years, SpaceX has launched more mass to orbit than the rest of the world's providers combined.

In time, the market will reach its own verdict on all of this.

The question, the only question that matters for an investor, is whether you have the conviction, the capital and the patience to be present when it does.

1 - A Dutch auction is so-named because it is used to sell cut flowers in Holland, in the enormous flower auctions. The strategy in a Dutch auction is that the price starts high. Each bidder watches the price decline until it reaches the point that either the bidder bids or a rival bids, and the auction ends.