SEOUL — Has the era of investors blindly pouring money into any company with an "AI" label finally come to an end? The market debut of a South Korean artificial intelligence firm has sparked major discussion all the way across the Pacific in Wall Street. What drew attention, however, was not a astronomical fundraising record, but rather a shocking 32% plunge in its share price on its very first day of trading on the Kosdaq market. Major U.S. media outlets analyzed that investors, who have poured massive amounts of capital into AI, are now beginning to carefully scrutinize the actual returns on their investments.

Elice Group, which listed on the Kosdaq market on the 20th, closed at 61,500 won, down 32.04% from its initial public offering (IPO) price of 90,500 won. The pre-listing atmosphere was entirely different. The institutional demand forecast recorded a competition ratio of 388.84 to 1, while the general public subscription reached 408 to 1. More than 10 trillion won poured in as subscription deposits, and the IPO price was finalized at the very top of its targeted range (70,400 won to 90,500 won).

Elice Group’s financial performance was also solid. Backed by its AI cloud infrastructure and corporate AI transition support businesses, the company posted 34.4 billion won in revenue and 10.3 billion won in operating profit in the first half of this year. Yet, once trading began, investors turned their backs, sending the stock tumbling.

Wall Street Takes Notice of 'Elice'

What makes this development fascinating is that the news captured attention in the United States. In an article titled "AI IPO Flop Sparks Fears Over End of Investment Boom," Bloomberg reported, "Elice, a company largely unknown outside of South Korea as of this week, has found itself at the center of an intense debate over AI company valuations and whether they can deliver earnings to match."

The outlet introduced Elice Group as a symbolic case highlighting the waning confidence in so-called "neocloud" companies. Neocloud refers to emerging cloud service providers that secure GPUs and rent out computing power to AI enterprises.

Reuters also covered the plunge in-depth, citing analysis from Ke Yan, head of research at Singapore-based Shentong Research. Yan noted, "Elice’s stock market debut shows that the evaluation criteria for AI infrastructure investment have shifted from 'build at all costs' to 'who is paying, and for how long?'" The remark points out that rather than simply focusing on how many AI data centers can be built, the market now prioritizes whether customers can stably afford sufficient fees to use those facilities.

U.S. financial publication MarketWatch covered Elice Group’s slump under the headline that South Korea's IPO failure is the latest in a string of disappointing cases in global AI investments.

Indeed, aside from Elice Group, other companies have recently faced difficulties attracting investors. Australia-based data center firm Firmus Grid, backed by Nvidia, canceled its planned large-scale IPO this month. High valuations, borrowing costs, and concerns over the recoverability of investments held investors back. AirTrunk, a data center operator backed by Blackstone, also delayed its planned Real Estate Investment Trust (REIT) listing in Singapore, while SoftBank-backed SB Energy shelved its U.S. stock market listing plans.

AI Investments: When Will Capital Be Recovered?

Until now, the AI investment frenzy has been driven heavily by expectations of future demand. The underlying logic was that operating AI services requires large-scale data centers and high-performance GPUs, and companies that secured them first would preempt the future market.

However, AI infrastructure demands massive initial capital. Significant costs go into everything from building data centers and purchasing GPUs to securing power supply and cooling facilities. For companies that borrow funds to build infrastructure, persistent high interest rates amplify financial burdens. If they fail to secure as many customers as expected or if their facilities do not generate sufficient revenue, capital recovery can be delayed, escalating debt repayment pressures.

U.S. AI data center and cloud companies such as CoreWeave and Nebius have also recently experienced heightened stock price volatility amid similar concerns. In particular, doubts are growing over the business models of neocloud companies that rely heavily on external funding to grow rapidly.

From this perspective, Wall Street suggests that Elice Group's first-day market crash may not be an issue isolated to South Korea alone. Investors, who previously assigned high valuations based on future growth potential, are now starting to meticulously examine actual sales, profits, and the possibility of recouping their investments. Is the era when a company could command a hefty price tag simply for being labeled an "AI" firm coming to an end? That precisely explains why Wall Street paid close attention to the fall of Elice Group.

[Copyright (c) Global Economic Times. All Rights Reserved.]