The IPO of Firmus, an AI data center company supported by Nvidia (NVDA.US), has come to an end! The $30 billion valuation is being questioned, and there is insufficient subscription demand from US investors

Zhitongcaijing · 1d ago

The Zhitong Finance App learned that Firmus Grid, an Australian artificial intelligence (AI) data center operator supported by Nvidia (NVDA.US), originally planned to go public at a valuation of about 30 billion US dollars and is expected to become one of Australia's largest initial public offerings (IPOs). However, due to overvaluation, insufficient investor subscription demand, and market concerns about huge financing needs for AI infrastructure, this much-publicized listing plan was finally cancelled on Friday.

According to people familiar with the matter, Firmus and its underwriting team originally wanted to attract a large number of US institutional investors to participate in the subscription, but during the promotion process from Tuesday to Wednesday, many US fund managers believed that the $30 billion valuation sought by the company was too expensive.

As subscription demand fell short of expectations, the underwriting team once considered reducing the distribution scale and lowering the valuation to salvage the deal. According to reports, the relevant plan includes reducing the financing scale to about 3 billion US dollars and lowering the company's valuation to between 20 billion and 25 billion US dollars. However, even after the conditions were adjusted, the underwriting team was unable to secure sufficient support from US investors and eventually had to cancel the IPO.

Bank of America (BAC.US), J.P. Morgan Chase (JPM.US), Morgan Stanley (MS.US), and Australian financial services company Morgan Financial acted as co-lead underwriters for this IPO.

In fact, Firmus' listing plan had previously sent positive signals. When deciding on the distribution plan, the company claimed that it had received strong demand from strategic investors and global institutional investors, and the underwriters also indicated that their initial subscription intentions exceeded the distribution scale.

However, as the listing process progressed, market questions about the company's valuation and transaction structure gradually heated up. One important issue is that existing shareholders may not be subject to share lock-up agreements, which means they have the opportunity to quickly sell their shares after the company goes public, putting potential pressure on stock prices.

On Wednesday, news about weak subscription demand and a possible reduction in the issuance price came out one after another, further exacerbating investors' unease. Some fund managers have begun to reduce subscription orders, while others have directly withdrawn their subscription intentions.

Firmus' IPO subscription process ended as planned on Thursday morning, but the market is still uncertain whether the company can complete the offering at the original price of $11 per share. Affected by related concerns, the stock price of Maas Group Holdings, one of Firmus' supporters, once plummeted 30% in the Sydney market, the biggest intraday decline since a record was set.

Jun Bei Liu, co-founder and chief portfolio manager of Australian investment agency Ten Cap Investments, said she had never seen an IPO that caused such serious market differences. At the same time, she pointed out that if the listing fails, Firmus may seek financing from existing investors instead.

The sharp rise in valuation in the short term has become one of the core reasons why Firmus' listing was blocked. In April of this year, Firmus was valued at about 5.5 billion US dollars in a round of financing supported by Coatue Management and Nvidia. By August, the company completed another round of financing of 2 billion US dollars, attracting the participation of well-known investment institutions such as Jane Street and Blackstone Group (BX.US), and the valuation rose to more than 10.5 billion US dollars.

However, just two months later, Firmus tried to enter the open market with a valuation of about 30 billion US dollars, which is nearly double the valuation at the time of the August round of financing. Such a rapid rise in valuation has raised questions from some investors about whether the actual scale of the company's operations can support its market value.

According to disclosed data, Firmus' revenue for the 2026 fiscal year was only 51 million US dollars, while the current IPO is seeking a valuation of more than 30 billion US dollars. At the same time, the total capacity of the company's planned data center reached 912 megawatts, but currently the actual built capacity is only 46 megawatts.

This means that investors need to value the company based on the large number of data center projects that have yet to be built and the profits that may be achieved in the next few years. However, whether these growth goals can be achieved still depends on the smooth implementation of project construction, customer needs, and huge financing.

Firmus used the ratio of corporate value to profit before interest and tax (EV/EBIT) as a valuation index during the IPO promotion process, and used companies such as the US AI cloud computing company Coreweave (CRWV.US) as comparison targets. Although the 13 times valuation multiplier proposed by Firmus is significantly lower than CoreWeave, this calculation is based on expected profit after two years rather than the company's current actual profit performance. In contrast, CoreWeave has a longer operating history, and its revenue scale is much higher than Firmus.

Leonid Mironov, portfolio manager of investment agency Gavekal Capital, said bluntly that this IPO had multiple problems such as poor transaction structure, excessive issuance scale, excessive pricing, and unreasonable valuation indicators. He pointed out that less than a year ago, Firmus' valuation was less than 2 billion US dollars; in August, it had already risen to about 10 billion US dollars, but by October they were trying to go public with a valuation of three times that.

According to the report, some investors could have accepted a valuation of about 25 billion US dollars, but Firmus announced a few days before the IPO price that it had reached a computing capacity supply agreement with its existing customer Meta Platforms (META.US), involving its data center facilities in Southeast Asia. The company then raised the relevant financial forecasts in accordance with this agreement and determined a higher IPO valuation based on this.

This adjustment has further heightened some investors' concerns. They believe that Firmus is still in the early stages of business development, yet investors are required to pay a high premium for high growth in the next few years in advance, and it will also require billions of dollars to achieve these growth goals.

Furthermore, the personal experience of Oliver Curtis, the co-founder of Firmus, has also received attention. Curtis served his prison sentence about ten years ago for insider trading. Although people familiar with the matter said that this was not a decisive factor in the failure of the IPO, the relevant background still increased the cautious sentiment of some investors.

The sudden failure of Firmus' listing plan also reflects that investors' concerns about AI infrastructure financing risks are heating up.

As global technology companies speed up the construction of data centers, related projects require huge investment, but it often takes a long time from completion of construction to generating stable income. For companies that lack mature operating records and still rely on future projects to meet their growth goals, investors are beginning to scrutinize their valuations, capital requirements, and execution capabilities more strictly.

Ultimately, Firmus officially withdrew its IPO plan on Friday because the subscription demand from US institutional investors fell short of expectations, and the demand from Australian investors was insufficient to fill the gap.

According to people familiar with the matter, the company is currently considering switching to smaller private financing, which may raise about 3 billion US dollars from existing investors. Previously, these existing investors were already preparing to subscribe for about half of the shares issued in this IPO.