My dad warned me about IPOs. He might still be right

In 2007, when I was just starting my career in finance (talk about timing), my dad was offered the chance to participate in the IPO of Boart Longyear.

At $2.3 billion, it was the second-largest IPO in Australian history at the time. Here was a drilling company with more than 120 years of operating history, arriving at the height of the mining boom and offering everyday Australians a chance to own a slice of the action.

"Dad, this looks great. Are you going to invest?" I asked as a wide-eyed rookie. His response has stuck with me ever since.

"Son, always be wary when someone is trying to sell you something. They know more than you do. Great companies are bought, not sold."

As it turned out, he was right.

Some of Boart Longyear's problems were bad luck. The company listed near the peak of the mining cycle (perhaps that was by design), and the GFC arrived less than two years later - which nobody (save for Michael Burry and co) saw coming. 

Other problems were self-inflicted. Debt piled up, restructurings followed, and what was once one of Australia's most celebrated listings eventually entered Chapter 11 bankruptcy protection before being taken private in 2024, at a tiny fraction of its IPO valuation.

The experience left me with a healthy scepticism of IPOs. Perhaps too much. So with SpaceX, OpenAI and Anthropic all likely to come to market at valuations measured in trillions of dollars, should I still be viewing them through the same lens I inherited from my dad?

To test my own biases, I put some questions to Alan Pullen from Magellan Investment Partners and Dr David Allen from Plato Investment Management. What do they think of the businesses, the valuations, and - most importantly - will they be buying stock?

What emerged was not a simple bull-versus-bear debate. Both managers believe these companies are extraordinary businesses. Both are wary of the expectations embedded in their valuations, yet they arrive at that conclusion from different directions.

These aren't dot-com stocks, but that doesn't make them buys

One of the easiest mistakes investors can make is dismissing these IPOs as another Pets.com moment. Neither manager sees it that way.

"There is no doubt these businesses are remarkable in their own way and are already dominant, or at least leading, in high-growth industries," says Pullen.

SpaceX has transformed the economics of space launches and built a fast-growing satellite internet business in Starlink. OpenAI remains synonymous with consumer AI. Anthropic has emerged as a serious challenger through its Claude platform and is rapidly building traction with enterprise customers.

These are real businesses serving real markets, which is precisely why this debate is so interesting.

Pullen argues that while the businesses themselves are genuine, the broader market environment remains "frothy", with many AI-related stocks exhibiting "dot-com style moves on limited fundamentals."

Allen reaches a similar conclusion through a different lens. He notes that history is littered with examples of large IPO waves arriving near periods of peak market optimism.

"The 1968-69 go-go era, the 1999-2000 dot-com bubble, and the zero-rates SPAC boom of 2021 all followed the same script," he says.

Importantly, Allen is not predicting an imminent collapse. He points out that NVIDIA is still growing revenue at around 85% year-on-year and that demand for AI infrastructure remains extremely strong. Genuine technological revolutions and investor exuberance often arrive together, making it difficult to separate opportunity from speculation.

Allen's challenge: the valuation hurdle is enormous

Of the two managers, Allen is the one who attacks the valuation case most directly. His analysis of SpaceX should make investors pause.

At approximately 100 times revenue, Allen estimates that SpaceX would need to grow revenue by roughly 28% per annum for the next 15 years, assuming a long-term net profit margin of 10% and a mature earnings multiple of 25 times. The comparison to corporate history is illuminating.

"Amazon, one of the greatest growth stories in corporate history, compounded revenue at roughly 20-23% over its last 15 years, well short of that bar."

NVIDIA has delivered extraordinary growth, but over a much shorter period and in highly unusual circumstances. Meanwhile, Microsoft, Alphabet and Meta never sustained anything close to 28% annual revenue growth from a comparable base.

Allen is not arguing that SpaceX cannot achieve it. He is arguing that investors appear to be treating an exceptionally rare outcome as though it were inevitable.

"It's not impossible, but it's an exceptionally rare achievement, and the market is pricing it in as though it's the base case."

That observation sits at the heart of the investment debate. The question is no longer whether SpaceX is a remarkable company. It is whether investors are already assuming a level of success that leaves little room for disappointment.

"When you pay 100 times revenue for a business, you're not buying a company, you're buying a specific version of the future", posits Allen. 

That future requires almost everything to go right.

Pullen's challenge: today's leaders may not be tomorrow's winners

While Allen focuses on valuation, Pullen is more concerned with the assumptions underpinning it. For him, the key issue is not whether AI will transform industries. It almost certainly will.

The harder question is whether today's leaders can maintain their advantage long enough to justify their valuations.

"A rational assessment of valuations requires an assessment of their economic moat," he says.

Can these businesses build sustainable competitive advantages? Can they maintain pricing power? Can they fend off rivals as the industry matures?

Pullen sees meaningful differences between the three companies.

Anthropic is currently the most compelling opportunity in his eyes because of its strong enterprise adoption and revenue growth. OpenAI, meanwhile, faces growing competition from Claude and Google's Gemini, making it a less attractive proposition at current expectations. His comments on SpaceX are particularly interesting.

"As a space nerd who loves the reusable booster SpaceX developed for its Falcon rockets, I hate to say that SpaceX is arguably the least interesting at its target IPO valuation."

The issue is not the quality of the existing business. It is that a meaningful portion of the valuation appears to rely on opportunities that remain largely unproven, including concepts such as space-based data centres and moonbases.

Investors are not simply betting on SpaceX's launch business, Starlink or Elon Musk's ability to innovate. They are also being asked to place a value on markets that may not exist in commercially meaningful form for years.

Pullen's warning is not that these opportunities are impossible. It is that investors appear to be paying for them before they have been proven.

The fear of missing out may be the biggest risk

The obvious counterargument is that investors who hesitate could miss the next Amazon, Microsoft or NVIDIA. Pullen is remarkably relaxed about that possibility.

"The market will give you another chance", he says. 

He points out that Amazon fell more than 90% after the dot-com crash, while NVIDIA has experienced multiple drawdowns exceeding 50% and two declines of more than 80%.

His message is simple: extraordinary businesses frequently experience extraordinary falls.

Allen approaches the problem differently. Rather than avoiding these businesses altogether, he argues for measured participation. 

At current valuations, he expects to hold positions around benchmark weight, participating if the companies continue to outperform without becoming dangerously exposed to a valuation that leaves little room for error.

It is a position that acknowledges a reality many investors struggle with. Being early and being right are not always the same thing, and neither are owning a great company and earning a great return.

Why the Amazon comparison breaks down

Perhaps the most common defence of these IPO valuations is that Amazon looked expensive too. Allen believes that comparison misses a crucial point.

When Amazon listed in 1997, it generated less than US$16 million in revenue and carried a market capitalisation of around US$440 million. It was far from a household name and described itself as an online bookseller.

Today, SpaceX, OpenAI and Anthropic are already global brands commanding valuations that would rank among the largest companies on earth.

"The only reason we mention Amazon in the same breath as SpaceX today is because of what it went on to achieve over the following three decades."

The numbers become almost absurd when pushed to their logical conclusion.

"Amazon has since grown its revenue more than 45,000 times over. If SpaceX were to replicate that multiple from its current revenue base, it would generate roughly US$840 trillion, something like seven or eight times global GDP."

Allen's point is not that SpaceX cannot be successful. It is that the scale of success required to justify the comparison bears little resemblance to the circumstances surrounding Amazon's IPO.

So, was Dad right?

Coming into this exercise, I expected one of two outcomes. Either these managers would reinforce my scepticism or convince me I was being overly cynical. Instead, they landed somewhere in between.

Neither believes these companies are speculative nonsense. Neither believes they are destined to fail. In fact, both think they are likely to remain important businesses for many years to come.

What they question is whether investors are being adequately compensated for the risks embedded in today's valuations.

That brings me full circle, back to my dad's comment almost 20 years ago. He wasn't telling me that Boart Longyear was a bad business. He wasn't telling me that mining was a bad industry. He wasn't even telling me that the company would fail.

His point was much simpler - when someone decides it's the right time to sell, it pays to ask why.

SpaceX, OpenAI and Anthropic may go on to become defining companies of the next decade. They may even justify today's valuations. But both Pullen and Allen are effectively approaching these opportunities with the same degree of healthy scepticism my dad instilled in me, and they are ultimately posing the same question;  

If these opportunities are as extraordinary as they appear, why are existing shareholders so eager to let the rest of us in?

Over to you

Will you be participating in the SpaceX IPO? What about Anthropic and OpenAI if you get the chance?

Let us know why or why not in the comments section below.