Dan Ives Sticks With His $350 IBM Target While HPE Banks the Fallout

Dan Ives is sticking with a $350 price target on IBM (NYSE:IBM) one day before earnings that a prediction market says will probably disappoint again. Shares closed at $212.67 this week after a slide of more than 26% in five trading days, IBM’s worst stretch since 1968. Polymarket traders are pricing an 82.6% chance of another miss when the company reports Wednesday, July 22.

IBM’s rivals are cashing the checks its own customers used to write. A memory shortage has enterprise buyers redirecting cash from mainframes and consulting toward AI hardware, and Hewlett Packard Enterprise is one of the plainest beneficiaries of that shift. Ives is betting the redirection reverses before it hardens into something permanent.

Two Records Fall in the Same Week

IBM’s slide started July 14, when management preannounced Q2 2026 results well below consensus. Shares fell roughly 22% that session, marking IBM’s worst single trading day since the 1987 crash. The selling did not stop there. By the end of the week, the cumulative decline topped 26%, IBM’s worst five day stretch since 1968.

Krishna did not soften the message to investors.

We faltered. Numerous large deals failed to close.

That was Arvind Krishna, IBM’s chairman and chief executive, describing what happened as customers shifted spending toward AI hardware amid the memory shortage, with added pressure from the rollout of IBM’s next mainframe generation, known as z17. IBM’s market capitalization has fallen to roughly $200 billion.

The reaction was severe partly because IBM had beaten estimates for four consecutive quarters. Trailing free cash flow for the first quarter of 2026 had already softened to $2.22 billion, down 44.15% from a year earlier. A downgrade to sell from the research firm Wall Street Zen added to the selling. Software peers Salesforce and ServiceNow slid on the same news, while semiconductor names climbed.

Where the Missing Budget Actually Went

The mechanism is a structural memory shortage, not a supply hiccup that fixes itself. Samsung, SK Hynix, and Micron together control more than 95% of global DRAM production, the standard memory chips used in servers and PCs, and all three have shifted capacity toward high bandwidth memory built for AI accelerators. That memory earns three to five times the revenue per wafer of ordinary DRAM.

The price impact has been sharp. Conventional DDR4 ECC server memory rose roughly 60% to 80% between early 2025 and the first quarter of 2026, while DDR5 ECC server memory rose 100% to 116% over the same stretch. The International Data Corporation (IDC) projects DRAM supply growth of just 16% this year, well short of what buyers need.

• 16%: projected global DRAM supply growth for 2026, according to IDC, well below demand
• 100% to 116%: the rise in DDR5 ECC server memory prices since early 2025
• $18.46 billion: record enterprise SSD revenue in the first quarter of 2026 as contract prices jumped further
• 23%: high bandwidth memory’s share of DRAM wafer output, up from 8% before the AI buildout

Micron Technology’s chief executive, Sanjay Mehrotra, has warned that supply tightness is likely to persist beyond 2027, with new fabrication capacity in the United States, Japan, and South Korea not expected in meaningful volume before 2028. SK Hynix said its entire 2026 high bandwidth memory production was already committed under long-term contracts as of last October. Enterprise buyers facing those price curves have a plain incentive to pull hardware purchases forward, even when it means delaying mainframe refreshes and software renewals with vendors like IBM.

HPE Is Banking the Dollars IBM Lost

Hewlett Packard Enterprise (NYSE:HPE) is up 92.46% year to date, the sharpest move in this comparison. Its fiscal second quarter, ended April 30, brought revenue of $10.7 billion, up 40% year over year. Networking revenue alone reached $2.7 billion, up 148.2%, powered by the $14 billion acquisition of Juniper Networks.

Chief Executive Antonio Neri told investors that networking orders grew even faster than revenue, more than doubling. HPE raised its AI networking order outlook to $1.9 billion and now expects free cash flow of at least $3.5 billion this year, a target it had originally set for fiscal 2028. HPE reached its own three year plan two years early in the same stretch that IBM’s free cash flow shrank.

The Peer Group Splits Four Ways

Oracle and Accenture round out a peer group that has scattered in four directions this year. IBM trades at a forward price to earnings ratio of 17, the cheapest of the four on a pure earnings multiple, and is down 27.26% year to date against a roughly 9% gain for the S&P 500 over the same stretch.

Oracle (NYSE:ORCL) trades at $126.41 after a 30.88% drawdown in the past month, carrying the widest analyst implied gap of any name here, and its own ratings lean bullish, eight strong buys and 29 buys against five holds and a single sell. Accenture (NYSE:ACN) trades at $143.57, down 45.07% year to date, with ratings tilting neutral. Recent analyst revisions on Accenture have leaned toward cuts. IBM’s setup still looks cleaner in one respect: less capex burn, an existing free cash flow machine, and a target that would return the stock toward its $332.46 52 week high rather than push it into uncharted territory.

Ives and the Crowd Are Pricing Different Bets

Wedbush analyst Dan Ives has raised his IBM target three times in about a year, from $300 to $325, then to $350 on June 2, weeks before the crash. IBM sits on his “IVES AI 30” list of AI winners, and this is not his first test of that conviction.

In February, IBM shares fell 13% in a single session after Anthropic said its Claude model could help modernize old COBOL code, threatening to bypass IBM’s installed base. Wedbush called the episode the “AI Ghost Trade” and argued AI would speed up legacy modernization work, not sidestep it. Those fears faded within weeks, and bulls making a similar case now are reaching back further, to a turnaround playbook IBM ran in 1993.

Ives points to fundamentals still moving in IBM’s favor. IBM Z mainframe revenue grew 51% year over year in the first quarter, and Krishna has said a fully loaded Z system now runs “about 450 billion inferences a day,” a figure Ives cites as evidence the Street still under models the mainframe’s AI role.

Polymarket’s own contracts do not tell one simple story either. The broad question of whether IBM beats overall Q2 estimates was pricing an 82.6% chance of a miss this week, up from just a 25.5% implied beat probability on July 14, meaning the crowd grew more pessimistic as the print approached. A separate contract asking whether IBM’s software revenue clears $7.9 billion was pricing roughly 80% odds in favor, suggesting traders expect the software line to hold up even if the headline number does not.

Morgan Stanley analyst Erik Woodring raised his target 10% to $293 ahead of the print, well below Ives but still pointing to meaningful upside. Alpha Vantage’s tracker shows 3 Strong Buy, 12 Buy, 7 Hold, 0 Sell, and 1 Strong Sell across 23 analysts covering the stock, a group that remains constructive even after the crash.

What Would Prove the Bulls Right?

Wall Street gets its answer this week. A results call built around temporary, z17 related timing, reaffirmed cash flow guidance, and sequential improvement in the AI book keeps Ives’ bet alive. Stalling consulting growth and a second straight quarter of shrinking free cash flow hand the argument to the bears.

Three things on Wednesday’s call would specifically support the bull case:

• Krishna’s timing explanation for the miss holds up, and enterprise wallet share turns out intact
• Management reaffirms roughly $1 billion of free cash flow improvement and 5%-plus constant currency revenue growth for the full year
• The generative AI book of business, at $12.5 billion inception to date as of the fourth quarter of 2025, shows sequential improvement

That path gets the stock back toward the $273.75 consensus target and gives Ives’ $350 print room to work. A dividend aristocrat with weakening cash generation is exactly how a value trap starts, even one paying a 3.07% yield after 31 straight years of increases.

Frequently Asked Questions

When Does IBM Report Q2 2026 Earnings?

IBM’s full Q2 2026 report and conference call are scheduled for Wednesday, July 22, after markets close. The preliminary numbers released July 14 already showed revenue of $17.2 billion, about $660 million short of consensus, with the full report expected to add segment detail and updated guidance.

What Is Polymarket and How Do Its Odds Work?

Polymarket is a prediction market where traders buy shares tied to yes or no questions, including whether a company beats earnings. Each share pays $1 if the outcome happens and nothing if it does not, so the share price doubles as an implied probability backed by real money rather than a poll.

Is IBM’s Dividend at Risk After the Free Cash Flow Decline?

IBM’s dividend still looks covered for now, even with free cash flow softening. Management still projects roughly $1 billion of free cash flow improvement for the full year, and the increase streak stands at 31 straight years. A second consecutive quarter of cash flow deterioration is the trigger analysts say would turn that into a real risk.

Why Isn’t HPE Getting Hurt by the Same Memory Shortage?

Hewlett Packard Enterprise sells the AI servers, networking gear, and storage that the shortage is making more valuable, instead of buying memory as an input the way IBM’s enterprise customers do. The same scarcity pushing IBM’s clients to delay software and mainframe purchases is showing up as record networking orders at HPE.

Disclaimer: This article is for informational purposes only and does not constitute investment advice. IBM, Oracle, Accenture, and HPE shares carry market risk, and analyst price targets are opinions, not guarantees. Consult a licensed financial advisor before making investment decisions. Figures are accurate as of July 21, 2026.