Roxy Jacenko’s billionaire husband Oliver Curtis rocked by $13billion blow as plans for one of Australia’s biggest-ever floats are thrown into doubt
Firmus Technologies’ planned Australian sharemarket listing is in doubt after reports that investors may value the AI infrastructure company at about $30 billion, well below its initial $43 billion target. The possible repricing, delay or cancellation matters because the float was expected to be the second-largest in Australian history.
Founded by Oliver Curtis and Tom Rosenfield in 2019, Firmus has deals with OpenAI, Nvidia and Meta and plans to build a network of AI data centres across Asia-Pacific. Sources said the proposed share price could fall from about $11 to $8.75, while Nvidia has reportedly been asked for more financial support; the company has yet to demonstrate a hyperscale facility at the scale of its expansion plans. Firmus was valued at $6.9 billion six months ago, and one US hedge fund reportedly declined to participate, citing Curtis’s insider trading conviction.
- Firmus’s planned ASX float may be repriced, delayed or cancelled.
- Sources put its possible valuation at $30 billion, down from $43 billion.
- The company plans a major Asia-Pacific AI data centre network.
New here? Start with this
Firmus Technologies is an Australian artificial intelligence infrastructure company founded in 2019 by Oliver Curtis and Tom Rosenfield. The firm has partnerships with OpenAI, Nvidia and Meta, and is building a network of data centres across Asia and the Pacific to support artificial intelligence applications.
The company had been planning to list on the Australian sharemarket in what would have become one of Australia's largest ever share offerings. This listing is significant because it would represent a major opportunity for investors to back Australian infrastructure development in artificial intelligence, an emerging technology sector of global importance.
Firmus aims to become a major provider of data centre infrastructure for artificial intelligence applications across Asia and the Pacific. The company's success depends on building and operating the facilities it has contracted to provide to its technology partners, whilst managing the substantial capital requirements of such large-scale infrastructure ventures.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
The valuation adjustment reflects prudent investor caution about material risks to Firmus's business model. The company has not yet demonstrated it can build AI data centres at the hyperscale promised in its expansion plans, which is fundamental to its viability. Governance concerns stemming from Curtis's insider trading conviction reasonably worry institutional investors assessing management credibility and risk management. The $43 billion initial target appears optimistic, and the revised $30 billion valuation represents a more realistic assessment in a challenging capital environment.
The case against
The market may be overcorrecting based on temporary setbacks and past personal matters rather than the company's fundamental prospects. Firmus has secured substantial partnerships with OpenAI, Nvidia, and Meta, suggesting genuine strategic confidence in its vision. Technology infrastructure companies routinely take time to demonstrate capabilities at planned scale; this is normal, not a warning sign. The valuation has grown significantly from $6.9 billion six months ago, demonstrating momentum and investor interest. Curtis's historical legal issues, whilst noteworthy, relate to personal conduct and shouldn't eclipse the company's technological potential in the booming AI infrastructure sector.