IBM – SaaSpocalyspe?

An example of one to have avoided.

• My main investment theme of 2026 is to pick the thrivers from the SaaS wreckage caused by AI panic, and the most recent warning from IBM is a good example from the original analysis of one to avoid.
• This came into sharp relief on 14^th July when IBM issued a warning regarding its Q2 26 performance, which triggered a 24.6% collapse in the share price.
• To be fair to IBM, the scale of this collapse is mostly correcting the circa 50% rally that it enjoyed as a result of it making large investments in quantum computing and the hype that that generated.
• I am bullish on quantum computing in the long term as it is by far the leading technology to take over from silicon when it reaches the end of the road, but there is still a very long way to go before this becomes an economic reality.
• Hence, one should not be pricing anything into IBM on a fundamental basis for these investments as they should be viewed as long-dated options rather than contributors to IBM’s investment case.
• However, the warning itself is troubling and is a sign that the have-nots of AI are going to encounter great difficulty in stopping their customers from spending their budgets elsewhere.
• This is exactly what IBM is seeing as Q2 revenues will be $17.2bn up 1% YoY and below consensus of $17.9bn while adj-EPS will be $2.93, up 5% YoY but below consensus of $3.01.
• The problem is simple in that a large part of IBM’s infrastructure business is selling data centre-sized racks that do transaction processing, which is a far cry from training and running AI inference.
• Hence, clients who are building in-house AI capacity prioritised spending on servers, memory and storage, which resulted in IBM’s infrastructure business declining by 7% YoY.
• The problem for IBM now is that this is the first concrete evidence of IT spending shifting away from legacy businesses and migrating towards AI, and the market now has reasons to fear that there is worse to come.
• I have not liked IBM due to its legacy exposure for some time, but also when one looks at it against the SaaS peer group, IBM has a higher valuation than many others but offers lower growth.
• Furthermore, that growth now looks to be at greater risk than some of the others, and so I think that the valuation could go lower from here.
• Again, to be completely fair to IBM, it has been through a transition like this before in the 1990s, and unlike many technology companies that no longer exist, it changed and it thrived.
• Hence, I think there is every chance that there is another lease of life awaiting Big Blue, but the transition is going to be hard as it was before.
• Furthermore, if this future is quantum, it is not going to bail IBM out before the mid-2030s, which is well beyond my investment horizon, and I am very long-term relative to the market that is myopic beyond the next quarter.
• This means that any recovery within an investment horizon will need to be a pivot towards AI, and here the legacy business is going to cause problems for some time to come.
• Every podcast that one listens to these days basically says that AI usage is exploding, but at the same time, every enterprise is trying to find ways of making it more cost-effective.
• This means running AI on one’s own infrastructure and also having orchestrators or harnesses that allocate tasks to the models best suited to each task, both in terms of performance and cost.
• This is why I think that companies like ServiceNow, Salesforce are going to thrive, as they already have products that do precisely this, meaning that their steady growth should continue.
• Adobe also remains well positioned as it offers control and copyright indemnification in a world where more content than ever is being created using AI.
• I don’t think consumers care about this at all, and so I would not be surprised to see Adobe’s consumer business go to zero, but enterprises are not going to take this risk.
• Hence, I continue to think that ServiceNow, CRM, Adobe, SAP. Snowflake will continue to do well, and of all of these, ServiceNow and Adobe offer by far the best risk-reward profile.
• I continue to have a position in both and am looking for a substantial rerating once the market realises that they are not about to be slaughtered on the gallows of AI.
• I would consider IBM at some stage, but I think it has a rocky patch ahead while it sorts out the transition, and it is expensive and higher risk compared to the positions I hold.