IMF: Global economy relies on AI, but investors start refusing financing

IMF: Global economy relies on AI, but investors start refusing financing
IMF: Global economy relies on AI, but investors start refusing financing

The global financial system has found itself in a situation where artificial intelligence technology has become the main engine of growth. On 13 October, the International Monetary Fund officially recorded this dependency in its "World Economic Outlook" report. Despite oil prices exceeding $100 and borrowing costs remaining high against the backdrop of ongoing military conflicts, the projected global growth rate for 2026 stands at 3.1%. This level is supported by investments in data centres, microchips, and energy infrastructure for AI, which offset other risks. Moreover, analysts have even revised their forecasts upwards for the Eurozone and the UK specifically thanks to this sector.

The credit nature of the tech boom

However, this model has a critical vulnerability: it requires a constant inflow of external funds. According to Morgan Stanley estimates, AI infrastructure will need approximately $1.5 trillion in external financing by 2028. This refers not to companies' own profits, but to loans and invested capital. This turns the technological story into a credit one. In an environment where US Treasury yields are reaching their highest levels since 2002 and interest rates are high, each new tranche of financing becomes increasingly expensive.

The market expects a perfect scenario

Financial markets have already priced in a near-flawless outlook. Analysts predict that from the third quarter, profits of companies in the S&P 500 index will grow by approximately 31%. About two-thirds of this increase is expected to be driven by the technology sector and AI. Prices already assume that everything will go according to plan.

The first crack: Firmus Grid's failure

The first serious alarm signal was the withdrawal of the IPO by Australian data centre operator Firmus Grid, backed by Nvidia. The company was preparing one of the largest public stock offerings in Australian history, with a potential raise of up to $5.5 billion. The share price was set at A$11, and the company's total valuation reached approximately $30 billion.

The dynamics of Firmus's valuations raise questions. In August, less than three months ago, the company raised $2 billion at a valuation of $10.5 billion. Over the year, Firmus raised over $3 billion in equity. Going public at a valuation three times higher was explained not by a tripling of the business, but by a general trend: assets in the AI sector were appreciating simply by virtue of their belonging to this industry.

Major funds said "no"

Just a few days before the planned offering, bankers assured of healthy demand. Major US funds were supposed to be the primary buyers. However, after studying the valuation, deal terms, and Treasury yields that exceeded 5%, these institutional investors declined to participate. Their response was categorical: not "make it cheaper," but simply "no."

On 9 October, Firmus Grid officially cancelled its IPO, citing market volatility. However, the deal fell apart in just 48 hours, and the cause was professional managers' refusal to buy assets at an inflated price. These were not retail investors panicking, nor skeptics who had been warning about a bubble for years. These were the same funds on which the entire AI infrastructure relies.

Consequences for the sector

This case is not isolated. Other projects are in the queue for listing, including British Nscale, also supported by Nvidia. Reports indicate that Nscale filed documents with the NYSE at a valuation of approximately $30 billion. Now, each such offering will be carefully scrutinised by investors, and bankers will take the Firmus precedent into account.

The era when anything with the right name was bought in AI is ending. Now investors are starting to count. The chain of logic is simple: global growth is held up by AI, AI is held up by new money, and new money is getting more expensive, with investors for the first time saying "too expensive." The dependency on the technology becomes noticeable precisely when the source of financing begins to dry up.

Comments

No comments yet. Be the first!

Leave a comment

Links are not allowed in comments. The editors may remove comments without explanation.
Back to top ↑