IBM just had its worst trading day in over a century. Shares dropped as much as 26% in a single session — the sharpest one-day fall since Black Monday in 1987. About $50 billion in market value disappeared before lunch. The stock now sits around $218, down 34% from its June high of $332.

The trigger was unusual. CEO Arvind Krishna released preliminary second-quarter numbers eight days before IBM’s scheduled earnings call. Revenue came in near $17.2 billion, below the $17.86 billion analysts expected. Earnings per share missed too, landing at $2.93 against a $3.01 estimate. Companies rarely pre-announce results this early. Doing so signals management felt the miss was too large to sit on.

Two things caused it. First, a sector-wide AI memory shortage is pulling corporate budgets toward hardware — servers, storage, memory — and away from software and mainframe deals, IBM’s bread and butter. Second, Krishna admitted a genuine execution problem: “numerous large deals failed to close on the timelines we expected.” Infrastructure revenue fell 7%, driven by weaker mainframe sales. Software still grew 5%. Consulting held flat. Buried in the same report: Red Hat revenue accelerated to 11% growth, a real bright spot nobody’s talking about.

Here’s what makes this hard to call. Wall Street didn’t panic in unison. Morgan Stanley and Oppenheimer raised price targets after the crash, calling the drop overdone. Evercore stuck with its Outperform rating. Bank of America trimmed its target but kept a Buy. HSBC, which had already downgraded IBM the day before on valuation grounds, now holds the Street’s lowest target. Across roughly two dozen analysts, ratings still skew bullish — about 15 buys against just 2 sells.

Nothing here points to financial distress. IBM’s Altman Z-Score of 3.28 sits comfortably in safe territory. The company generated nearly $15 billion in free cash flow last year. It’s a Dividend Aristocrat, paying dividends without interruption since 1916, with 30-plus straight years of increases. Because the payout wasn’t touched, the crash mechanically pushed the yield up to about 3.1%, above its five-year average.

IBM’s moat is intact where it matters most. Z-series mainframes remain deeply embedded in banks, insurers, and governments — switching costs are enormous, and this quarter’s stumble looks like timing, not customers walking away. Red Hat OpenShift gives IBM real leadership in enterprise Kubernetes. The softer spots are Consulting, smaller and less nimble than Accenture, and general enterprise software, where IBM competes credibly against Salesforce, SAP, and Microsoft without leading the category.

Krishna has run IBM since 2020 and gets real credit for a multi-year turnaround: spinning off the lower-growth Kyndryl business in 2021, doubling down on hybrid cloud through the Red Hat and HashiCorp deals, and pushing growth from roughly flat to double digits. He chose to face this miss head-on with an unusually blunt public letter rather than wait for the scheduled call. Still, this marks the second scare in five months — a February episode questioned whether AI coding tools might erode IBM’s mainframe business. Two sharp surprises in half a year raise fair questions about how predictable management’s near-term guidance really is.

The crash did real work on valuation, not just sentiment. Forward earnings multiples compressed from about 22x — already flagged as stretched — down to roughly 16x, right in line with the sector average. Nothing about the underlying earnings estimates has changed yet. That’s the open question the July 22 call will answer: does the miss stay contained to one quarter’s timing, or do full-year numbers need to come down too.

There are clear signposts either way. A stronger case for buying builds if the July call shows the deal slippage was purely timing — those delayed contracts closing in Q3 — and if management holds full-year guidance. The case weakens if infrastructure keeps softening beyond the expected mainframe cycle lull, or if Consulting starts losing share rather than just growing slowly. Watch Red Hat’s growth rate too: as long as it keeps accelerating, the core thesis that IBM is successfully pivoting toward hybrid cloud and software stays intact, regardless of what happens with the mainframe timing this quarter.

That uncertainty is why this sits as a Watchlist call with a constructive lean, not a confident Buy or a clear Avoid. A complete picture, including updated full-year guidance, arrives in about a week. Acting today means guessing at information that will simply be available soon.

Want the full breakdown — financials, valuation, competitive positioning, and scenarios? The complete IBM report is on our Reports page, alongside the rest of our coverage.