A U.S. view of Firmus
Everyone seems to have jumped on the carcass of the Firmus Initial Public Offering (IPO) to dissect what went wrong. Since it was a local company listing on the domestic market, it’s understandable Australian news outlets and financial commentators have shown so much interest.
But while local commentators and media outlets continue to report on what would have become Australia’s largest-ever float, U.S. commentators are also responding. The difference is some are reading the pulled IPO through a far more ominous lens, one that says the artificial intelligence (AI) bubble is ending.
To Wall Street observers and Silicon Valley critics, the wreckage here in Sydney isn’t just a local misfire with implications for future Australian IPOs, but a potential canary in the coal mine for the whole global AI trade. Indeed, they see it as an event, a milestone, a sign that public market investors are finally questioning, if not revolting against, the entire narrative-driven AI boom.
In the United States, prominent anti-monopoly commentator and writer Matt Stoller, writes: “A failed initial public offering from an Nvidia-backed “neocloud” called Firmus, run by a man with a criminal record, suggests the AI bubble is getting ready to burst.”
Framing the failure as an alarming signal for the tech sector, Stoller argues the pulled listing offers an unflattering look into the financial plumbing sustaining chipmaker Nvidia’s US$6 trillion circular domain.
At the heart of Stoller’s critique is the “circular financing” we’ve discussed on our blog, here: The AI bubble – cracks beneath the surface, and here: The artificial intelligence boom sounds the alarm – is this the Dot.com bubble 2.0? – as well as elsewhere.
Circular financing is a ‘pretend-to-extend’ mechanism in which intermediate or downstream companies buy chips from NVIDIA using funds NVIDIA invested in them for an equity stake. By handing cash over to customers for equity, NVIDIA effectively funds their purchases of its chips, inflating its revenue.
This dynamic has given rise to a class of intermediate entities known as “neoclouds.” Essentially, they’re just a collection of permits, server racks, real estate, and massive debt wrapped in a corporate shell. Their existence lets major tech firms keep colossal capital expenditures (capex) off their balance sheets while inflating demand for processing hardware.
Then private equity and venture capital funds invest in these neoclouds at absurd valuations, confident in the belief that a subsequent public market listing will eventually relieve them of the asset and provide an attractive return.
In the case of Firmus – which pivoted to AI infrastructure after starting out as a bitcoin miner – the model hit a wall.
With an IPO valuation of A$43.7 billion and backed by blue-chip financial giants like Blackstone and Jane Street alongside Nvidia, the IPO collapsed when Australian fund managers revolted with reports citing a variety of issues, including:
- A A$43.7 billion valuation that priced in five years of flawless execution for a loss-making business with little track record, yet reportedly missed its first payment to a landlord.
- That any major company could invest A$7 billion in equity and borrow $30 billion to build data centres. Yet that $7 billion of equity was being valued at $43.7 billion. “There is no magic in this business, simply buying and leasing the same assets in competition with hundreds of competitors.”
- That having OpenAI and Meta as tenants was not a gold stamp of approval. Its biggest customers are only leasing capacity (server space and chips) today, but they can hand it all back and walk away tomorrow, and are likely to do so after developing their own more capable and more efficient chips.
- That Firmus is a relatively young business that hasn’t yet been tested by a full cycle.
- That the IPO bookbuild was described by bankers as well covered at $11 per share, but was then cut to $8.25.
- That Firmus was a startup being valued on the assumption that it will execute almost flawlessly and grow revenue by roughly 25,300 per cent in three years.
- That the cancelled partnership with established Australian data centre operator CDC Data Centres, increased execution risk and reduced credibility, and
- That Firmus’s greatest asset is its relationship with NVIDIA, which, perhaps unwittingly, admitted earlier this year that, “If we didn’t support CoreWeave to exist, these neoclouds, these AI clouds, wouldn’t exist.”
Unusually, the float was structured to let some insiders dump shares immediately on day one, bypassing standard escrow arrangements. To both ‘insto’ (institutional investor) and sophisticated investors, this structure signalled a possible lack of confidence from the very people running the company.
And while they might be financially aggressive, U.S. ‘establishment’ investors tend to be socially conservative. Unsurprisingly, many were stunned that major global banks greenlit the Firmus deal given the leadership’s backgrounds. Stoller reports, co-founder Oliver Curtis served prison time a decade ago for a $1.4 million insider trading scheme, while his co-founder and cousin, Tim Rosenfield, previously ran a loss-making lingerie firm called Steadmark. If the investment had gone south, how would these fund managers explain investing in a company with this personnel history?
Elsewhere, and not unlike the wave of companies that added ‘DotCom’ to their names during the 1999 Tech Bubble, Stoller notes that Firmus was among a string of similar neocloud and infrastructure plays – from TeraWulf and Sharon AI to Applied Digital – that have completed rapid pivots from bitcoin mining, plasma technology, or tech staffing into AI data centre operators. And just as many are backed by prestigious tech names despite colourful executive histories or questionable underlying unit economics.
Ultimately, Stoller argues the Firmus IPO flop marks a turning point in the AI thematic. While private markets, venture capitalists, and tech giants have successfully circulated and raised private valuations amid peak hype, the public market has now drawn a line in the sand. As the tide recedes on speculative infrastructure deals, only solid cash flows, bulletproof governance, and sustainable business models will attract support from secondary markets. And even then, it might be a much lower valuations.
Roger Montgomery is the Founder and Chairman of Montgomery Investment Management. Roger has over three decades of experience in funds management and related activities, including equities analysis, equity and derivatives strategy, trading and stockbroking. Prior to establishing Montgomery, Roger held positions at Ord Minnett Jardine Fleming, BT (Australia) Limited and Merrill Lynch.
He is also author of best-selling investment guide-book for the stock market, Value.able – how to value the best stocks and buy them for less than they are worth.
Roger appears regularly on television and radio, and in the press, including ABC radio and TV, The Australian and Ausbiz. View upcoming media appearances.
This post was contributed by a representative of Montgomery Investment Management Pty Limited (AFSL No. 354564). The principal purpose of this post is to provide factual information and not provide financial product advice. Additionally, the information provided is not intended to provide any recommendation or opinion about any financial product. Any commentary and statements of opinion however may contain general advice only that is prepared without taking into account your personal objectives, financial circumstances or needs. Because of this, before acting on any of the information provided, you should always consider its appropriateness in light of your personal objectives, financial circumstances and needs and should consider seeking independent advice from a financial advisor if necessary before making any decisions. This post specifically excludes personal advice.
