Riding the AI Bubble

For some time now, the hype around artificial intelligence (AI) has driven an explosive increase in the market capitalisation of firms operating in the AI space. Late in October, Nvidia, which provides the advanced chips that power AI deve-lopment, became the world’s first $5 trillion company. That is surprising, since its valuation stood at $4 trillion three months earlier and $400 billion three years earlier, when ChatGPT was first announced. The boom does not end with AI, but pervades firms in the tech space with very different connections to the evolution of AI. Firms like Apple and Microsoft that have bought into AI, through investments in market leader OpenAI or in AI infrastructure like cloud computing and data centres, have found their valuations exploding as well, touching $4 trillion or more, as has Oracle, which provides cloud-computing services. As a result, the sharp rise of the S&P 500 equity index is due to the rise in the prices of the shares of a few AI-linked tech firms. According to a recent estimate quoted by the Financial Times, eight of the 10 dominant stocks in the S&P 500 are tech stocks, accounting for 36% of the entire United States (US) market’s value and almost 80% of the S&P 500’s net income growth in the year to November. There is evidence that a substantial part of the investment driving this boom is also borrowed. The frenzy on the part of financial investors has been fuelled by the decision of firms in the AI space to undertake huge investments in pursuit of the AI dream. Alphabet, Amazon, Google and Meta have together spent an estimated $112 billion on chips and data centres in the quarter ending September 2025 alone. Investment is booming, even though current trends in revenues and net earnings do not justify that exposure. The real investment boom is based on the expectations of the revenue growth that a “revolutionary,” generic technology like AI is expected to deliver in the future. Similar expectations riding on the hype surrounding the technology drive financial investors as well. In the event, investment to revenue ratios and (share) price to earnings ratios are at record levels, underlining the dramatic acceleration in revenue and profit growth that the players involved are betting on. There is, however, much cause for scepticism. The record on the benefits of AI use, even where visible, is still way short of revolutionary. And in some instances, those benefits are yet to be seen. High investments mean that depreciation costs to be covered in the future would be high, necessitating high prices to support the required earnings from sales to users, besides related advertising. But competition from much cheaper and equally effective alternatives like Deepseek could prevent such pricing strategies. And the product range required to deliver higher earnings is yet to unfold. These uncertainties matter because of the fragile base on which the boom in share prices and investments of companies in the extended AI space is built. Based on the promise of a “predicted” glorious future for AI, firms like OpenAI, planning to develop large language models and programmes that put them to practical use, have been contracting with chip makers like Nvidia, and investing in subsidiaries (like Amazon Web Services, Microsoft Azure, and Google Cloud Platform) or buying into services of independent data centres and cloud-computing firms like Oracle. These firms and those investing in and lending to them are predicting rapid growth based on this projected demand, hinging on the AI boom. In the event, the unbridled spending by these firms has been encouraged not just by hugely optimistic estimates of the future of AI, but by the large volumes of still cheap liquidity available in the system because of the easy money policies of central banks and the liberalised financial system, which has seen non-bank financial players, such as private equity/credit firms, mobilising that liquidity and deploying it for profit. It is now becoming clear that a disproportionate share of such funds is being directed to AI and AI-related firms. The volume of capital these AI-related firms are absorbing is so large that they cannot only rely on the large sums being drawn through the sale of equity at inflated values. Rather, they are piling up large debts as well. US companies in the AI space have issued bonds valued in excess of $200 billion this year. An estimate from Goldman Sachs suggests that the sale of bonds to the tune of $180 billion by a few firms like Meta, Alphabet and Oracle has accounted for a quarter of the corporate borrowing this year. As a result, these firms are adding interest costs to the large depreciation costs that would show up in their accounts in the years to come. Sometimes, this form of financing is associated with deep entanglement with vendors. Thus, Nvidia, which is cash-rich, following the expansion of sales and boom in its share values after the launch of ChatGPT, invested $100 billion in OpenAI, which in turn has promised to buy $100 billion of Nvidia chips for its ChatGPT development. A significant part of that spending is therefore effectively financed by Nvidia’s investment. Meanwhile, market assessments are that Nvidia chose to invest in OpenAI to prevent it from relying excessively on rival suppliers of chips like Broadcom. Thus, the euphoric rise in substantially leveraged investments rides on the expected performance of a few entangled firms in a single tech space. This concentrated exposure of financial firms and investors based on the mere expectations of dazzling future earnings has raised concerns that, once again, the US is the centre of a bubble that could unravel, as occurred in 2008. Yet the government and regulators are not stepping in to temper, if not end, the euphoria because the investments that the boom is giving rise to and the luxury consumption that the beneficiaries of the financial boom are indulging in are partly responsible for much of the growth the US economy records. But there are signs that the investors are getting nervous. Over the first week of November, there were strong signs that the boom may be unwinding. Firms riding on the AI boom lost market value of around $1 trillion over just that week. Eight of the most valuable AI-related tech firms, such as Nvidia, Meta and Oracle, lost as much as $800 billion in market value. Even Nvidia, which had quite recently broken the $5 trillion valuation record, lost around $350 billion. If the trend continues, the US and the world could be looking at another financial crisis, which would have external effects on the real economy that could be as damaging as in 2008 and after. In fact, if this is the beginning of the next financial collapse, things look worse than they did in 2008. The Federal Reserve’s balance sheet is so bloated that it would be hard-pressed to inject as much cheap liquidity into the system to save financial and non-financial firms as it did last time. And the elasticity of the spending power of the US Treasury is also likely to be limited by the political stand-off and divide that has led to the prolonged shutdown of the US government. With the capacity to bail out firms and the economy thus restricted, stalling the downturn would be difficult. But then, those governing capitalism never learn enough from history to prevent these periodic collapses. This time, that error could precipitate a crisis that is as bad as it was in the 1930s.