IBM took a record hit due to expanded artificial intelligence (AI) infrastructure investments.

According to the Wall Street Journal (WSJ) on the 14th (local time), IBM's stock price plummeted more than 25% that day, losing $69 billion (approximately 102.87 trillion won) in market value in a single day. This is the largest single-day decline in IBM's history.

This sharp decline was triggered by the company's unusual profit warning announcement, with the main cause being customers shifting spending from legacy software to AI hardware and memory chips.

IBM is preparing to announce its second-quarter earnings next week, and this incident is interpreted as a sign of how companies' AI-related spending will impact software consumption.

▲ New Aspects of 'The End of Software'

The stock price declines experienced by software companies like Adobe and Salesforce in early this year were due to concerns that AI companies like Anthropic would easily create cheap replicas of existing services.

However, IBM's stock crash shows a different pattern.

The main cause is identified as a so-called 'budget crowding out' phenomenon, where as companies purchase AI-related products, they end up eating into existing IT budgets.

▲ Rising Hardware Costs and IT Budget Constraints

The rapid growth of AI has driven up chip prices, which in turn has raised costs across all devices from laptops to gaming consoles to AI data center servers. These cost increases are pressuring IT budgets at large institutions, including banks, which are IBM's core customer base.

These institutions find themselves in a situation where they must purchase massive computing power from cloud companies to power in-house AI tools.

▲ IBM CEO Krishna: "Unexpected Changes in Capital Expenditure Priorities"

Arvind Krishna, IBM's chief executive officer (CEO), stated in June that customers shifted their quarterly capital expenditures toward securing infrastructure with limited supply, such as servers, storage, and memory, due to concerns about price increases.

He said, "Perfect execution is required in this environment, but we faltered this quarter," adding that while he had anticipated some supply chain impacts, he did not expect capital expenditure priorities to be reallocated so dramatically.

▲ Realization of Cutting Existing Spending for AI Adoption

Gil Luria, head of technology research at D.A. Davidson, analyzed that companies are reducing existing software and hardware spending to cover the massive computing costs incurred as they increase the use of AI tools for coding, marketing, and data analysis.

He said such types of companies would proliferate throughout this earnings season, and he is hearing from customers, "We need to cut elsewhere to create room in the AI budget."

▲ Mainframe Slump and Market Environment Changes

IBM's crisis is not limited to software. Sales of the 'z17,' an enterprise mainframe launched for the AI era, also fell short of expectations.

IBM originally expected infrastructure revenue to decline in the single-digit range at the beginning, but now projects a 7% decline.

According to the Wall Street Journal (WSJ) on the 14th (local time), unlike companies that sell hardware or provide cloud computing such as Nvidia, Google, and Oracle, IBM's weakness is that it sticks to a model of directly installing systems at customer sites, making it overly dependent on financial services companies.

IBM[Reuters/Yonhapnews]

▲ Intensified Competition and Response Strategies

IBM's mainframe computing and consulting business now finds itself in direct competition with AI models like Claude Code.

Additionally, with the proliferation of data center clusters, enterprise customers can now access the computing resources they need at more competitive prices.

In response, IBM has been making desperate efforts, recently announcing a $5 billion cybersecurity strategy through 'Project Lightwell' with Red Hat to strengthen security in the financial sector.

▲ Future Investment and Global Memory War

Nevertheless, experts agree that companies are prioritizing AI adoption to improve efficiency and productivity.

IBM is also continuing massive investments in quantum computing infrastructure, which is regarded as next-generation processing technology. In June, it launched the 'Anderon' unit for quantum computing chip manufacturing and announced plans to invest an additional $9 billion over the next five years.

Meanwhile, memory chip manufacturers such as SK Hynix, Micron, and Samsung Electronics are facing supply shortages due to surging demand for DRAM and NAND flash memory, essential for AI model training and execution, which is also leading to increased consumer electronics costs.

▲ Market Reaction and Future Outlook

Compared to IBM's plunge on the 14th, the declines of other software companies such as Workday, Adobe, and ServiceNow were relatively modest.

Gil Luria assessed that after investors carefully analyzed IBM's warning, they identified that the core cause of the weakness was insufficient demand for the 'z17 mainframe,' and fears about software budget erosion were somewhat alleviated. Most software companies do not sell mainframe computers.