IBM Just Had Its Worst Day Since Black Monday 1987. I'm Using It to Buy More. Here's Why.

This looks less like a permanent demand air-pocket and more like an enterprise budget rotation from services into infrastructure, which is structurally worse for IBM’s mix than a simple timing miss. The first-order loser is IBM’s higher-margin software and consulting attach rate; the second-order winner is the AI/server ecosystem that absorbs the redirected wallet share, including NVDA-adjacent compute suppliers and server OEMs like DELL/HPE. If customers bought hardware early to preempt price hikes, the next procurement wave may be smaller than management hopes, because some of the budget is now already spent.

The key risk is that ‘temporary’ becomes a multi-quarter reset if backlog conversion and bookings do not reaccelerate. In the next 1-3 months, the market will care less about the one-quarter revenue miss than about whether deferred services demand reappears in the order book; if not, IBM’s multiple deserves to de-rate from low- to mid-teens rather than rebound. Over 6-18 months, AI-related bookings can still support the equity story, but only if they convert into realized revenue without cannibalizing legacy mix and margins.

Consensus is probably underpricing how sticky budget reallocation can be in large enterprises: once capex is pulled forward, consulting and software projects often get reprioritized, not merely delayed. The contrarian read is that the selloff may be overdone tactically, but the recovery path is more fragile than the bulls assume. The falsifier is simple: if the next quarter shows weak bookings/RPO or another guide-down in software and consulting, this is not a one-off and the value case breaks quickly.