Not a "Price Bubble," but a "Profit Bubble"
AI Ecosystem Growing Increasingly Similar to the Financial Crisis
We Must See Beyond the Dazzling Numbers and Face Harsh Realities

[The View]Beware Not of 2000, But of 2008 View original image

Recently, investors watching the artificial intelligence (AI) boom have been focusing on traditional indicators such as the price-to-earnings ratio (PER). Based solely on this metric, it is difficult to find any signs of a bubble in the current market. Nvidia's forward PER is only in the low 30s, while Microsoft (MS), Meta, and other major technology companies are trading at around 20 times earnings, typical for blue-chip stocks. Compared to the dot-com bubble of 2000, when Cisco traded at a PER of 150 and MS at 70, today's market appears relatively rational and composed.


However, the real and potentially fatal risk lies deep within this very optical illusion created by these metrics. The current AI boom is not a simple 'price bubble' like the dot-com period of 2000, but instead takes the form of a 'profit bubble,' similar to what triggered the 2008 global financial crisis. In 2000, anyone could tell there was a bubble as the stock prices of companies with neither earnings nor revenues surged with no limit. In contrast, today, companies are reporting significantly inflated book profits, leading to the mistaken impression that stock prices are cheap.


The core of this artificially created profit bubble lies in ingenious accounting mechanisms and circular revenue structures. Major companies are extending the service lives of expensive graphic processing unit (GPU) servers beyond traditional periods, reducing accounting depreciation expenses, and thus greatly overstating their book profits. Furthermore, Nvidia not only sells chips to cloud companies but also invests billions of dollars in their equity. In addition, major tech firms are funding AI startups, which in turn spend enormous sums on the tech companies’ own cloud services. This flow does not represent independent demand generated from external end customers; rather, it is simply capital circulating within the ecosystem. Nevertheless, the market interprets it as normal growth demand and cheers it on.


Recent excitement in the market over the SpaceX initial public offering (IPO) and associated concerns must also be viewed within the broader context of this macro-level blind faith and capital concentration. SpaceX may be wrapped in the grand vision of humanity’s space frontier and overwhelming technological prowess, but as an extreme infrastructure industry requiring astronomical amounts of upfront capital investment, it follows the same path as the current AI frenzy. The market pours huge sums into SpaceX, excited by the prospect of future monopoly, yet the nature of the space industry means it will take considerable time to produce meaningful cash flows. Businesses dependent on massive capital expenditures are particularly vulnerable if liquidity shrinks due to events such as interest rate hikes or credit crunches. The hidden capital consumption and long-term profitability uncertainties behind the SpaceX IPO resemble the vulnerabilities of the AI ecosystem, which is supported only by current book profits and rosy predictions.


What is even more concerning is that these exaggerated profits and demand are thoroughly financed by debt. Today’s AI computing supply chain eerily mirrors the securitization structure seen in the 2008 subprime mortgage crisis. Back then, lenders originated mortgages and securitized them to pass on the risk; now, fledgling cloud companies take on enormous debt backed by rental income from GPUs, which have an extremely short economic lifespan. Big tech companies present this demand as if it were guaranteed future profit, but the ultimate risk is passed on—almost unchanged—to pension recipients and ordinary investors who believe in the future of innovative industries and invest in bonds.


It is important to remember that in such cyclical businesses, the lowest PER often coincides ironically with the greatest risk. At the peak of the cycle, profits are abnormally maximized, causing the price-to-earnings multiple to appear at its lowest. In 2008, even firms such as Moody’s and Citigroup posted record results and extremely low valuations at the top of the cycle, but as soon as the cycle turned, hidden insolvencies were exposed all at once, leading to an irreversible disaster.


We are now overlooking the looming collapse of credit because we are deceived by surface-level figures and low multiples. The sweet trap of seemingly cheap valuations, as well as the blind enthusiasm for infrastructure investment witnessed in the SpaceX IPO, may ultimately be a prelude to an impending systemic crisis. Before the spectacularly inflated profit bubble leads to a chain reaction of credit collapses, it is time to view the market with a sober and discerning eye.



Sungkyu Park, Professor at Willamette University, USA


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing