IBM's Worst Week Ever: Stock Plunges 26% as AI Hype Fades and Transformation Proves Long and Hard

IBM's stock fell 2.92% on Friday (17th), closing at $212.66 per share, resulting in a cumulative weekly decline of approximately 26%—the worst weekly performance in the company's history.

IBM's stock had risen 0.93% on Monday, closing at $290.23, but plummeted 25.21% on Tuesday after an early release of its second-quarter results, closing at $217.07—the largest single-day drop in company history. The stock continued to fluctuate afterward, falling another 2.70% on Wednesday, rebounding 3.72% on Thursday, and dropping 2.92% again on Friday.

This week's sell-off reflects not only the second-quarter results falling short of expectations but also growing investor skepticism about IBM's ability to maintain its dominance in the mainframe market and successfully transform into a key beneficiary of the AI era.

Early Earnings Release Shocks Market, Wiping Out Nearly $70 Billion in Market Value

On Tuesday, IBM unusually released its second-quarter financial results ahead of schedule. Both revenue and profit missed Wall Street expectations, causing the stock to crash 25% that day—the largest single-day drop in company history—and wiping out nearly $70 billion in market value.

IBM is typically very cautious about its financial reporting schedule, with the last early release occurring in October 2008 during the global financial crisis, when the company aimed to reassure investors it could still meet financial targets. This time, however, the early release revealed multiple issues in business growth and execution.

Before the earnings announcement, IBM's stock had hit an all-time high on June 2. Investors had high expectations, betting on the company's leadership in mainframe computing, its quantum computing initiatives, and its potential to benefit from AI infrastructure. The disappointing results amplified market disappointment, fueling a wave of sell-offs.

Infrastructure Revenue Drops 7% as Major Deals Fail to Close on Schedule

IBM faced slowing growth in its software and consulting businesses in Q2, pressured by enterprise customers shifting tech budgets toward AI infrastructure such as chips and servers. The biggest drag came from the infrastructure segment.

IBM had anticipated a decline in infrastructure revenue, but the actual 7% drop was faster and more severe than expected. Enterprises are not only reducing purchases of mainframe hardware but also cutting spending on high-margin software used for banking and credit card processing.

IBM CEO Arvind Krishna attributed the performance gap to poor execution. In a letter to shareholders, he admitted the company failed to adjust quickly enough and that several major deals did not close as scheduled—key reasons for the Q2 shortfall.

IBM still dominates the mainframe market. A 2022 study commissioned by IBM and conducted by Celent found that over half of global transactions by value are processed by IBM Z-series mainframes. Yet, this market advantage did not prevent a sharp slowdown in infrastructure business during Q2.

Software Growth Falls Short of Expectations, Analysts Turn Cautious

IBM's software division grew revenue by 5% in Q2, far below Oppenheimer's initial forecast of 12%. After the earnings release, Oppenheimer downgraded IBM's stock rating and began questioning whether the company can achieve double-digit software revenue growth by 2027.

Analysts believe IBM may need to close the large deals that failed to materialize in Q2 or make significant acquisitions of faster-growing software assets to bridge its current growth gap.

Stefan Slowinski, an analyst at BNP Paribas, noted that IBM's strategy is to use cash flow to acquire high-growth software companies and improve its overall growth profile. However, IBM's current consulting, software, and mainframe businesses all show only low single-digit organic growth rates, leaving very limited room to accelerate growth organically.

AI Could Be Eroding the Mainframe's Moat

A deeper market concern is whether AI is weakening the competitive moat IBM has built over decades in its mainframe business.

AI startup Anthropic released a COBOL modernization guide for Claude Code in February, claiming it can significantly simplify the process of updating legacy programs. COBOL remains widely used in IBM mainframes and critical enterprise systems like banking.

In the past, migrating systems away from IBM mainframes required high costs and complex engineering, forming a protective moat for IBM's high-margin mainframe business. If AI can reduce the difficulty of rewriting code and migrating systems, this advantage could gradually erode.

Valuation Has Fallen But May Not Be Cheap Enough—Market Demands Proof of Execution

By Wednesday's close, IBM's forward 12-month P/E ratio had dropped to 16.54x, the lowest since June 2024. In contrast, on June 2, IBM's forward P/E briefly exceeded 25x, higher than the S&P 500's 21.52x at the time.

Slowinski believes the stock decline has made IBM more attractive, but the company's organic growth rate and future outlook remain among the lowest in his coverage universe, justifying a valuation discount. He currently maintains an 'Underperform' rating on IBM.

Dan O'Regan, Managing Director of Equity Trading at Mizuho Securities, pointed out that IBM had become a popular AI infrastructure beneficiary stock, with its price near all-time highs—making any execution misstep severely punished. However, the 25% single-day drop also suggests the market may already be pricing in a more persistent business slowdown than management admits.

O'Regan stated that IBM's future stock performance will no longer depend solely on AI hype but on whether management can prove this earnings miss was a one-time execution error, not the start of structural demand decline. Until IBM delivers tangible improvements, the market will not easily trust it again.