IBM’s shares fell by 25 per cent in a single day, even though the company had not announced a slump in sales or any losses. Investors, however, saw something more significant: the AI boom is beginning to shift corporate spending priorities, and a broad technology portfolio does not guarantee that a supplier will capture this shifting budget.
IBM presented its preliminary results for the second quarter of 2026. Revenue rose by 1 per cent to $17.2 billion, whilst adjusted earnings per share increased by 5 per cent. At the same time, software sales grew by 5 per cent., whilst consulting remained flat and infrastructure recorded a 7 per cent decline. These results did not indicate a crisis for the company, but they were enough to wipe approximately $69 billion off its market capitalisation. It was IBM’s worst trading session in its history.
Such a strong reaction shows that the market was not judging the company on the basis of a single quarter alone. It questioned IBM’s ability to maintain predictable growth at a time when AI is rapidly reshaping customer spending. For businesses, this means that even a relatively small shift in revenue can alter a supplier’s valuation if it undermines the credibility of its core strategy.
AI isn’t cutting IT budgets. It’s changing their priorities
The most important information did not appear in the results table, but in the explanation provided by IBM’s CEO, Arvind Krishna. Towards the end of June, customers began shifting their spending towards servers, storage and memory modules to safeguard their hardware against shortages and further price rises. Fewer resources and less attention were devoted to software, services and projects related to IBM Z mainframes.
This does not mean that companies are cutting back on technology investment. Gartner forecasts that global IT spending will rise by 13.5 per cent in 2026, to $6.31 trillion. Spending on data centre systems is expected to grow the fastest, by 55.8 per cent. The market continues to expand, but funding is currently directed primarily towards the infrastructure required to run AI: computing power, servers, memory and cloud infrastructure.
For technology suppliers, this is a crucial distinction. Growth in the market as a whole does not automatically translate into growth in every segment. A company may be at the heart of the AI boom, yet still lose contracts if its offering is further down the customer’s shopping list.
The pressure to shift priorities is not solely an IBM argument. TrendForce forecasts that contract prices for traditional DRAM will rise by 13–18 per cent in the third quarter, and for NAND Flash by 10–15 per cent. Manufacturers continue to direct production capacity towards more profitable server and AI-related applications, limiting supply to other customers.
In such circumstances, purchasing components in advance becomes a form of risk management for a company. At the same time, every additional euro spent on memory, storage or servers may delay application modernisation, data integration or the implementation of consultancy services. AI therefore generates new expenditure, but also competes for resources with the rest of the technology stack.
A broad portfolio was not enough
In theory, IBM is well positioned. It sells infrastructure, software, services, hybrid cloud and AI solutions. This structure should enable it to capture spending regardless of which part of the market the capital flows into.
In practice, budgets have shifted to categories where IBM has failed to offset the decline in sales of mainframes and associated transactional software. Strong performance in selected areas was not enough to protect the company as a whole. The problem, therefore, was not a lack of products, but a failure to respond quickly enough to changes in customer behaviour.
Krishna admitted that IBM had failed to adapt in time, and that numerous major contracts had not been finalised as planned. According to the company, it was precisely these delays that accounted for most of the shortfall against expectations. Component shortages therefore explain the market shift, but do not absolve IBM of responsibility for delivery.
This is an important lesson for suppliers developing ‘end-to-end’ strategies. The breadth of a portfolio only creates a competitive advantage when a company is able to shift the conversation with the customer between hardware, software and services without losing the entire contract. If a client scales back an application project, an integrated supplier should secure business in infrastructure, financing, integration or environment management. Otherwise, a broad offering remains merely a collection of separate products.
Hardware can delay the business value of AI
From a business perspective, shifting the budget towards infrastructure solves one problem but creates another. Securing servers and memory mitigates the risk of price rises and component shortages. However, it does not guarantee that the company will be able to utilise the computing power it has purchased.
Value from AI is only realised once the infrastructure is integrated with data, applications, security, processes and staff expertise. If hardware consumes a disproportionate share of the budget, the company may end the year with greater computing power, but without any ready-to-use business applications. Rising memory prices therefore affect not only IT costs, but also the timetable for the entire transformation.
The case of IBM does not yet prove that software manufacturers as a group will lose out on AI investments. It does, however, show that market growth is becoming increasingly uneven, and that predicting the timing of customer spending is just as important as the direction of the technological trend.
IBM’s historic decline was therefore not a reaction to a single weaker quarter. It was a warning that, in the AI boom, value is captured not by the companies with the largest number of products, but by those that are quickest to follow the customer’s budget as it shifts between infrastructure, software and services.
