Bitget App
Trade smarter
Buy cryptoMarketsTradeFuturesStocksEarnInstitutionAI & More
Firmus backed by Nvidia delays Australian listing, casting a shadow over data center IPO candidates

Firmus backed by Nvidia delays Australian listing, casting a shadow over data center IPO candidates

智通财经智通财经2026/10/09 00:56
Show original
By:智通财经

The Australian IPO of data center company Firmus Grid Ltd., supported by Nvidia (NVDA.US), is expected to be delayed, which is an ominous sign for other companies planning to go public.

According to Zhitong Finance APP, the Australia IPO of data center company Firmus Grid Ltd., which is supported by Nvidia (NVDA.US), is expected to be delayed, sending an ominous signal to other firms planning to go public. Previously, AI infrastructure companies have experienced a series of disappointing debuts.

Reports indicate that Firmus is about to decide to pause its IPO, which was originally estimated to reach up to $5.5 billion, and is instead considering raising more funds through the private market. Some investors previously believed the pricing strategy for the deal was too aggressive.

In the past month, the yield on the 10-year US Treasury has increased sharply, putting pressure on interest rate-sensitive sectors. Dave Mazza, CEO of Roundhill Investments, said that the so-called "new cloud" companies which investors use for comparison—including CoreWeave Inc. and Nebius Group NV—are among the most highly leveraged and volatile AI-related stocks.

"The short thesis is having its moment, as the cost of capital has spiraled out of control and long-duration stocks are under pressure, making the short-term outlook quite challenging," Mazza commented.

Several companies have slowed their pace after publicly filing documents with the US Securities and Exchange Commission. By regulation, firms must wait 15 days before they can officially begin their IPO roadshows. Although AI cloud computing provider Nscale Ltd. and AI data center and power infrastructure developer SB Energy Inc., which is backed by SoftBank Group, applied for US listings last month, they have yet to start marketing their respective deals.

Singapore-based DayOne Data Centers Ltd. filed for a US IPO earlier this week, and based on its schedule, it could start formal investor presentations on October 21.

Sources familiar with the matter say these companies are all hoping to find an IPO window while investors await the anticipated large-scale listing from cutting-edge AI lab Anthropic PBC, potentially as early as November.

"Our bullish view on these companies is based on their ability to deploy computing power faster than traditional cloud-service providers, but their price-to-sales ratios are extremely high, and they need to issue debt to fund their operations," Mazza said. His actively managed ETF includes the Roundhill Neocloud ETF.

"If you're Nscale and you see what's happening with Firmus, it's not a signal that gives you confidence to price your own IPO," he added.

An Nscale representative declined to comment. The Roundhill Neocloud ETF, which holds positions in Nebius, CoreWeave, and IREN Ltd., dropped 7.2% on Thursday.

AI Spending Surges

According to compiled market data, companies seeking to capture or participate in surging AI-related spending have driven two major deals, pushing this year’s weighted average IPO return in the US to 15%. This data excludes blank-check companies and other financial instruments.

SpaceX, with its record-breaking $86.2 billion IPO that highlighted its future AI-related business ambitions during its roadshow, has seen its share price rise 19% since the IPO; Korean memory chip manufacturer SK Hynix’s US Depository Receipts have climbed 14% since listing.

In contrast, data center companies Csquare Inc. and Blackstone Digital Infrastructure Trust Inc. have each fallen more than 15% since their recent debuts.

Several so-called "picks and shovels" companies, which provide exposure to AI-related industry activities—such as heating, cooling, and ventilation systems for data centers, or power equipment that keeps them running—have also performed poorly since going public this year.

Nonetheless, market researchers project that generative AI spending will reach $2.3 trillion by 2032, with AI agent deployments alone accounting for about $286 billion. As AI usage increases correspondingly, many investors still believe demand for data center computing power will continue to grow.

"There is indeed a lot of equity and debt financing in the market, but we think it’s worth it," said Tim Ghriskey, Senior Portfolio Strategist at Ingalls & Snyder.

"We believe there is a lot of pent-up demand for data centers, and I don’t think this will end. It won’t be smooth sailing, but there’s a lot of growth ahead."

0
0

Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.

Understand the market, then trade.
Bitget offers one-stop trading for cryptocurrencies, stocks, and gold.
Trade now!

You may also like

Trump Entered Before Muse AI's Surge! Accurate Meta and SpaceX Rebound Trades in August, Bought SpaceX Bonds Ahead of New Aerospace Policies

In August, Trump's account conducted 517 securities transactions with a total transaction amount of approximately $74.3 million to $273.3 million, including the purchase of $5 million to $25 million worth of Meta stock.

智通财经•2026/10/09 07:21

JPMorgan analyzes Applied (APLD.US): AI computing power business advancing rapidly, facing real-world challenges with billion-dollar financing

JPMorgan's Credit Research team has released a report providing a comprehensive analysis of company operations, project progress, financing pressures, and risks.

智通财经•2026/10/09 06:42
JPMorgan analyzes Applied (APLD.US): AI computing power business advancing rapidly, facing real-world challenges with billion-dollar financing

Societe Generale's chief bear: The AI boom is replicating the Asian financial crisis, with the deadly “debt time bomb”

Société Générale Chief Strategist Edwards warns that the true trigger for the AI boom is not inadequate productivity, but whether cheap money can continue to flow. Overseas long-term government bond yields continue to rise, while AI giants are issuing large amounts of debt, with hundreds of billions of dollars in new bonds needing to be absorbed by the same group of buyers. This is reminiscent of the 1990s, when Asia maintained prosperity through short-term dollar borrowing.

华尔街见闻•2026/10/09 06:36