Arvind Krishna's Letter to IBM Investors

IBM reported Q2 revenue of $17.2B, up 1%, but warned the quarter’s results are disappointing, with Infrastructure revenue down 7% and GAAP gross margin at 57.7% (down 100bps). Non-GAAP EPS rose 5% to $2.93, supported by Red Hat software growth accelerating to 11% and Red Hat/HashiCorp/Confluent traction, but management cited a worse-than-expected z17/z performance shortfall plus a capex reprioritization toward servers/storage/memory and cybersecurity-related deal delays. IBM generated $7.8B net cash from operations YTD and $4.8B free cash flow, while highlighting ongoing AI initiatives (Lightwell GA July 8; $5B commitment) and plans to invest $10B+ in quantum over five years; the company will discuss full-year expectations on July 22.

Analysis

This reads less like a demand collapse than an execution reset, which is usually worse for the multiple. IBM’s core risk is that investors have been paying for software durability and cash conversion; when deal timing slips and mainframe attach weakens, the market tends to reprice both 2026 growth and the quality of earnings, not just one quarter.

The second-order winners are the infrastructure names where customers can reallocate budget quickly: server/storage/memory vendors and, to a lesser degree, cybersecurity vendors if security spending stays elevated. That makes the relative setup better for HPE, DELL, SMCI, and memory suppliers like MU than for IBM’s software stack, because the spend is moving toward fungible hardware and away from large-solution purchases that depend on close rates.

The catalyst is the July 22 call. If management has to temper full-year software or FCF expectations, IBM likely sees another leg down over 1-3 months; if they frame this as timing and reaffirm backlog, the stock can recover because the cash engine is still intact. The quantum/AI messaging is option value, not earnings value: the market should discount it unless there are partner wins and milestone-driven capital efficiency by 2027-2029. Contrarian view: this may be overdone if the June capex shift was a one-off and deferred enterprise deals reclose in Q3; that would favor a squeeze, but only if guidance remains unchanged.

Key Decisions for Investors

• Short IBM into the July 22 earnings call, or buy 2-4 month put spreads, to express downside from a likely guide reset; risk is limited if management keeps FY software growth and FCF intact.
• Pair long MU or HPE vs short IBM for the next 1-3 months: thesis is that budget reallocation into storage/memory and infrastructure is real, while IBM’s execution shortfall is idiosyncratic and may persist.
• If IBM trades down hard but Q3/FY guidance is reaffirmed, cover shorts into the flush; that would signal the move is a timing issue rather than a structural demand break.
• Do not chase the quantum/AI announcements as a near-term long; treat them as long-dated option value until there is evidence of revenue conversion or a lower capex path.