TradingKey - On Tuesday, IBM ( IBM )'s preliminary second-quarter earnings preview sent shockwaves through the market. The company's disclosed revenue of $17.2 billion was not only far below the $17.86 billion consensus estimate of Wall Street analysts, but more concerningly, highlighted the declining trend of its legacy businesses amid the AI infrastructure boom.
The news sent IBM's pre-market share price plunging by more than 20%, which, if sustained into regular trading, could mark its largest single-day decline since 'Black Monday' in 1987.

Source: Google Finance
According to analyst consensus estimates compiled by the London Stock Exchange Group, the market had originally expected IBM's second-quarter revenue to be $17.86 billion, while the company's preliminary estimate for revenue was only $17.2 billion. Adjusted earnings per share is expected to be $2.93, falling below the market expectation of $3.02.
By segment, IBM's software revenue grew by 5%, a decent performance, but infrastructure revenue fell 7% year-over-year, significantly weaker than the market's previous expectation of a full-year, low-single-digit decline.
IBM stated that it expects infrastructure revenue to experience a single-digit decline starting this quarter. The company said the revenue decline was mainly due to the performance of the Z series and a lack of related software stacks.
In addition, in the final weeks of June, customers shifted their quarterly capital expenditures toward server, storage, and memory purchases to secure tight infrastructure capacity ahead of anticipated price increases, a shift that directly impacted IBM's legacy business.
In a public letter to investors, IBM Chief Executive Officer Arvind Krishna admitted that the company failed to quickly keep pace with the constantly changing market environment this quarter, business execution lagged behind, and multiple large contracts were not signed as scheduled.
He stated, "We had expected the supply chain to be affected, but we did not anticipate such a massive realignment of customer capital expenditures."
Currently, global AI infrastructure development has entered a high-spending phase, with enterprises shifting more IT budgets from mature software products to computing power equipment, storage, and supply chain resources. This trend not only impacts traditional tech giants like IBM, but also sparks market concerns over the prospects of the entire software industry.
For a long time, software investors have worried that AI automation tools would replace traditional software functions. However, IBM's latest disclosure reveals a new risk: even hardware investments centered around AI are squeezing software procurement budgets.
Chris Beauchamp, chief market analyst at IG Group, noted: "This is the darkest hour for IBM and the entire software industry. The market's biggest concern is how long the trend of enterprises shifting from software to infrastructure will persist."
Over the past few years, IBM has been trying to shed its traditional hardware and mainframe labels, accelerating its transformation into an enterprise software and AI infrastructure provider through acquisitions of companies such as Red Hat, HashiCorp, and Confluent.
The company had previously emphasized that artificial intelligence would drive customer demand for its infrastructure software, enabling enterprises to connect and manage different AI models more efficiently.
However, this earnings warning has reignited market doubts about IBM's transformation path. In February this year, after AI startup Anthropic introduced a tool that helps modernize legacy programming languages on IBM mainframes, IBM's stock price experienced significant volatility. Meanwhile, the second-quarter results falling short of expectations have further reinforced market concerns that its legacy business is being squeezed by AI infrastructure investments.