IBM's USD $67 Billion Wipeout and the implications to software sector

• IBM suffered the worst single-day decline in its 115-year history on 14 July 2026, plunging 25.2% to close at $217.07 — surpassing the 23.7% loss it recorded on Black Monday in 1987.

• IBM issued an unusual pre-announcement disclosing preliminary Q2 2026 results: revenue of $17.2 billion (consensus $17.85 billion, +1% year-on-year) and non-GAAP EPS of $2.93 (consensus $3.01).

• IT budgets are growing, but infrastructure price increases — particularly in memory — are growing faster than those budgets.

• Near-term effect of AI investment may be to delay portions on enterprise software spending, even if long-term demand remains intact.

• Overall, we have yet to see significant reductions in company spending on software renewals or new licenses, but it is something to watch during the upcoming results season. For now, we continue to hold our positive view on the software sector.

IBM suffered the worst single-day decline in its 115-year history on 14 July 2026, plunging 25.2% to close at $217.07 — surpassing the 23.7% loss it recorded on Black Monday in 1987. The sell-off erased roughly $67 billion of market capitalisation, was accompanied by trading volume 551% above the three-month average and immediately dragged down software and IT services names across the board. The catalyst was not standard earnings miss.

It suggested that AI infrastructure spending may be beginning to cannibalise budgets previously allocated to enterprise software.

An unusual pre-earnings announcement

Ahead of its scheduled 22 July earnings date, IBM issued an unusual pre-announcement disclosing preliminary Q2 2026 results: revenue of $17.2 billion (consensus $17.85 billion, +1% year-on-year) and non-GAAP EPS of $2.93 (consensus $3.01). The shortfall was concentrated in two segments. Infrastructure revenue fell 7%, driven by weak mainframe sales against the difficult comp of the z17 launch, while software — now 45% of IBM's revenue and the centrepiece of CEO Arvind Krishna's turnaround thesis — grew just 5%, well below the double-digit trajectory IBM had been running. Consulting was flat. Full-year guidance was deferred.

The contrast with Q1 was stark. Just three months earlier, IBM had delivered software revenue growth of 11% to $7.05 billion, beating on both the top and bottom line and affirming the narrative that hybrid cloud and AI were structurally lifting the business. That narrative weakened materially in a single quarter. Before this, consensus revenue growth for the software segment is around 10% y/y, now company management had guided half of that.

Figure 1: software segment forecast decline

The Root Cause: Hardware Crowding Out Software

In a letter to investors, Krishna explained that the shortfall is due to over the last few weeks of June, enterprise clients shifted their quarterly capex spend toward servers, storage, and memory purchases to secure supply-constrained infrastructure ahead of expected price increases. IBM had anticipated some supply-chain-related drag, but not the magnitude of the capex reprioritisation. Numerous large deals failed to close on expected timelines, driving most of the shortfall.

The mechanism is straightforward: IT budgets are growing, but infrastructure price increases — particularly in memory — are growing faster than those budgets. When memory and server prices rise fast enough, enterprises pull forward hardware purchases at the expense of software renewals and consulting engagements. DRAM prices rose between 100% and 116% in Q1 2026 as AI data centre demand motivated memory makers to divert wafer capacity toward more profitable high-bandwidth memory (HBM). SK Hynix has warned that the tightest supply conditions arrive in 2027, meaning similar budget reallocations Krishna described could persist if memory supply remains tight.

A secondary pressure compounded the miss: cybersecurity budget diversion. Krishna cited "rapidly evolving, industry-wide cybersecurity concerns" as another draw on client budgets, noting that Anthropic's AI cybersecurity model, Mythos, is causing enterprises to pause new deals while they reassess how much they need to spend on cyber. The important nuance is what is being paused. Enterprises are not cutting cyber budgets — they are freezing new software and consulting deals with vendors like IBM while they figure out how much more to allocate toward security. Total enterprise’s cyber spend is rising, not falling.

That distinction explains an otherwise puzzling dynamic: pure-play cybersecurity vendors are simultaneously raising prices. CrowdStrike's renewal pricing rose approximately 27% year-on-year across both SMB and enterprise tiers and the broader cybersecurity platforms category has seen similar upward pressure. This is a renewal-side story rather than a new-deal story — endpoint security is operationally sticky, switching costs are high, and customers who have layered identity and cloud workload modules onto the Falcon platform cannot easily let their licences lapse while broader strategy is being reassessed.

Agentic AI is compounding the pressure by creating a wave of new machine identities that need to be secured alongside human users, expanding the attack surface and giving vendors further justification for price increases. The result could be a growing share of IT wallet absorbed by cyber at the direct expense of discretionary categories like application software and consulting.

The market priced these two sides of the same coin in real time. On the same day IBM crashed 25%, CrowdStrike surged more than 10% to a 52-week high, the Global X Cybersecurity ETF (BUG) jumped over 6%, and Fortinet, Okta, and Palo Alto Networks all rallied. Investors read Krishna's warning as confirmation that cybersecurity is "must-have" spend that enterprises are protecting — and that incumbent cyber vendors retain pricing power on renewals — even as budgets for application software and mainframe purchases are being cut, which reinforced our positive call on the cybersecurity sector. 

Table 1:IT services most exposed to IT budget reallocation.

Source: Gartner (February 2026). iFAST estimates as of 16 July 2026. Brighter color represents larger amount.

 Software and Services Under Pressure

The sell-off did not stay contained to IBM. ServiceNow fell approximately 7%, Salesforce lost 5%, Accenture dropped 8%, and Cognizant declined 7%. None of these companies share IBM's mainframe exposure or its z17 supply-chain problem. What they share is a client base that is now demonstrably reprioritising hardware over software contracts, and investors treated that commonality as reason enough to sell first and ask questions later.

The flipside of the same dynamic lifted the direct beneficiaries: Micron rose approximately 5% and SanDisk nearly 6% on the same day, reflecting the fact that the memory names are on the receiving end of the very budget shift that is crushing software.

For software investors who have been paying elevated multiples on the premise that AI adoption would accelerate deal cycles and expand budgets, IBM's pre-announcement introduced an uncomfortable possibility: the near-term effect of AI investment may be to delay portions on enterprise software spending, even if long-term demand remains intact.

Delayed, not cancelled

A memory shortage is inflating hardware costs and crowding out software and services spend from the same IT wallet. This is a budget-allocation story, not a demand-destruction story. If the delayed deals close in H2, the revenue would be delayed rather than lost.

We believe that among IT budgets, IT consultations and IT services are the easiest line item to defer without operational risk. Digital transformation project can be delayed by a quarter with limited immediate operational disruption while endpoint security lapse and core software licence renewals are non-negotiables, you cannot stop renewing your software licenses as this will impact your business operations.

Overall, we have yet to see significant reductions in company spending on software renewals or new licenses, but it is something to watch during the upcoming results season. For now, we continue to hold our positive view on the software sector.

The Bigger Picture

IBM's pre-announcement marks the moment the memory-wall thesis crossed over from a semiconductor story to an enterprise software story. For two years, the AI buildout has been a rising tide for infrastructure names — chipmakers, cloud hyperscalers, memory producers. IBM's Q2 provides one of the clearest indications so far that the same spending wave is not just lifting some boats but actively swamping others.

The divide in the technology sector is no longer between AI winners and AI-unrelated names; it is between companies building AI infrastructure and companies whose budgets are increasingly being redirected toward AI infrastructure.

Table 2: Recommended Products

Sector	Recommended Products
Software	iShares Expanded Tech-Software Sector ETF (BATS: IGV)
Cybersecurity	Global X Cybersecurity ETF (NASDAQ:BUG)

Declaration:

For specific disclosure, at the time of publication of this report, IFPL (via its connected and associated entities) and the analyst who produced this report hold a NIL position in the abovementioned securities.

This research report was prepared with the assistance of artificial intelligence (AI) tools. iFAST Financial Pte Ltd does not rely exclusively on AI for content generation; the content of this report – including all investment theses, ratings, price targets and conclusions – has been independently reviewed and verified by the research analyst(s) to ensure accuracy and professional integrity.