CoinWResearch Institute
IBM's second-quarter financial report can be summarized in three keywords: growth in servers and storage, software slowdown, and cash pressure. The company's revenue was $17.162 billion, a year-on-year increase of 1%; adjusted earnings per share were $2.93, both below market expectations. The market anticipated revenue of $17.86 billion and adjusted earnings per share of $3.01. The gap itself is not large, but the change in business structure is more noteworthy. Among them, the Power series servers and storage grew relatively quickly; excluding acquisition contributions, IBM's original software revenue showed virtually no growth, with significant declines in software for large mainframes. The company subsequently lowered its full-year revenue growth guidance for 2026, excluding currency effects, from above 5% to 4%—5%.
This set of changes reflects that some customers have adjusted their purchasing order. Management stated that in the last few weeks of the second quarter, some clients, concerned about tightening supplies and rising prices, prioritized purchasing servers, storage, and memory, leading to delays in previously planned software and large mainframe-related transactions. IBM's sales team did not adapt to this change in a timely manner, resulting in multiple large transactions not being completed as scheduled. Demand related to AI first manifested itself in servers and storage businesses with lower margins than software, while the growth rate of high-margin software slowed, putting overall gross margins under pressure.
The third quarter will be an important window for determining whether IBM's performance can recover. The timely completion of delayed software contracts, whether nearly $500 million worth of Power series server and storage backlog orders can be delivered, and whether AI consulting projects can enter implementation and billing stages will jointly reflect whether customer demand persists, and will test whether equipment growth can continue to be transformed into software, consulting, and cash. Meanwhile, quantum computing is becoming a new variable in IBM's valuation. The Trump administration plans to offer $1 billion in support for IBM's quantum chip manufacturing project, increasing market interest in its industrial prospects.
1. First Understand How IBM Makes Money
IBM primarily serves large organizations such as banks, airlines, manufacturing, telecommunications, and government agencies. Once critical systems like payment, ticketing, inventory, and risk management adopt IBM's hardware and software, the cost of switching suppliers is high, and the migration process may impact business continuity, which is why customers often purchase software, maintenance, and consulting services for the long term.
In terms of second-quarter revenue, software accounts for about 45%, making it the core of IBM's profit. IBM's software includes cloud software, data management, automation tools, and software that supports large mainframes. Most customers renew their subscriptions annually, and with each additional software revenue, the profit retained is usually higher than that from hardware and consulting. An increase in software growth rate indicates an improvement in revenue structure; a slowdown typically leads the market to lower its valuation of the company.
In terms of second-quarter revenue, consulting accounts for about 31%, which can reflect customer demand and impact future revenue. IBM helps businesses transform old systems and integrates AI into customer service, finance, supply chain, and risk management. Customers sign contracts first, and revenue is progressively recognized only after project initiation, implementation, and acceptance. Therefore, growth in signed contracts only indicates the willingness of customers to invest; revenue growth can demonstrate that the projects have entered the billing phase.
In terms of second-quarter revenue, infrastructure accounts for about 22%, which is an important source of quarterly performance volatility. This part includes large mainframes, enterprise servers, and storage devices. Large mainframes can be understood as high-performance computers used by banks and airlines to handle core transactions. Hardware revenue is affected by the new product cycle, component supply, and delivery times, with quarterly fluctuations typically being larger than for software and consulting.
The expenditures corresponding to these three types of businesses also vary: software requires continuous R&D and sales investment, consulting relies on professionals, and hardware still needs to procure components and prepare inventory. Acquisitions can consume cash and potentially increase debt and interest expenses. To understand IBM's financial report, one can remember four main lines: software looks at profit, consulting looks at future projects, hardware looks at quarterly delivery, and cash looks at how much performance ultimately retains. Next, these four lines can be used to break down the second-quarter report to determine whether IBM's problems stem from insufficient total revenue or changes in revenue structure.
2. The Most Worrisome Aspect of the Second Quarter Report Is the Software Slowdown
IBM's second-quarter revenue was $17.162 billion, a year-on-year increase of 1%, below the $17.86 billion market average expectation; adjusted earnings per share were $2.93, also below the pre-warning expectation of $3.01. The gap itself is limited, but the weakness in the software and large mainframe business, along with management acknowledging issues in sales execution, became the main reasons for the stock price decline.
Software revenue was $7.761 billion, a year-on-year increase of 5%. Part of this growth came from acquisitions; excluding acquisition contributions, IBM's original software revenue showed virtually no growth. Excluding the impact of exchange rates, cloud software revenue grew by 11%, and data business grew by 18%, which also includes revenues from newly acquired companies. Software that supports critical businesses such as banking payments and airline ticketing declined by 9%, becoming the main influencing factor in the software business.
Consulting revenue was $5.327 billion, with signed contracts growing faster than revenue. Excluding the impact of exchange rates, consulting revenue grew by 1%, while the amount of new contracts increased by 6%. AI-related projects accounted for about half of new contracts signed in the consulting business and over 30% of incomplete projects. This indicates strong customer willingness to invest, but IBM still needs to complete project implementation and acceptance for the corresponding contracts to gradually convert into revenue.
Infrastructure revenue was $3.835 billion, declining by 7% after excluding the impact of exchange rates, with servers and storage performing well, while large mainframes remained weak. After excluding currency effects, large mainframe revenue fell by 42%, while distributed infrastructure revenue primarily from Power series enterprise servers and storage grew by 37%. The backlog of orders for Power series servers and storage is close to $500 million, and IBM has not committed to delivering all these orders in the third quarter. Management believes that demand related to AI infrastructure, new product upgrades, and customers' procurement rhythms have collectively driven growth, but the company did not disclose how much of this is attributable to AI revenue.
Management lowered its full-year revenue growth guidance, excluding the impact of exchange rates, from above 5% to 4%—5%, and revised the software growth target for the full year, excluding currency impact, down to 6%—8%. Whether revenue growth can recover in the second half of the year primarily depends on the timely signing of delayed software contracts, gradual delivery of the backlog of Power series servers and storage orders, and accelerating consulting projects entering implementation and billing phases.
Key metrics to watch in this financial report include whether IBM's original software can resume growth, if cloud software can maintain double-digit growth, whether gross margins can stabilize, and if profits can convert into free cash flow. These four indicators did not improve concurrently in the second quarter, and AI-related demand did not simultaneously bring higher profits and more cash. Among them, servers and storage can confirm revenue relatively quickly, while software and consulting require signing, implementation, and acceptance. This time lag explains the performance pressure in the second quarter and raises the next question: how IBM can convert equipment demand into higher-margin, longer-lasting software and service revenue.
3. AI Demand Has Emerged; IBM Needs to Convert Equipment Orders into Long-Term Revenue
In the second quarter, demand for equipment and consulting related to AI deployment has appeared, but this demand has not yet fully converted into software revenue and cash. The speed of purchasing and revenue recognition for equipment, software, and consulting varies; some clients will first replenish their servers, storage, and memory, then purchase data management and operational software, and introduce consulting services during the project implementation phase. The specific order varies by customer, but equipment is typically delivered faster, large software contracts need to be signed, and consulting still requires implementation and acceptance.
In the last few weeks of the second quarter, management observed that some large customers, concerned about tightening supply and rising prices, prioritized purchasing servers, storage, and memory and delayed some software and large mainframe transactions. This change explains the combination of rapid growth in Power series servers and storage and weaker performance in large mainframes and related software. It does not represent that all customers’ AI budgets are flowing in the same way, nor can all equipment growth be attributed to AI revenue. What can be confirmed is that the revenue increment recognized in this quarter has largely come from server and storage businesses with lower margins than software.
Changes in customer purchasing explain only part of the pressure. Management acknowledged that multiple large transactions were not completed as planned, and the sales team did not timely adapt to the new purchasing order. The large mainframes and accompanying software became the most significant influencing factors for the quarter. If delayed contracts are still not signed or recognized in the third quarter, concerns about weakening demand and execution capabilities in the market will further escalate.
IBM's business model dictates that equipment orders are merely the starting point. Management provided the long-term experience that each dollar of hardware revenue can lead to more than three dollars in long-term software contracts. This ratio reflects the company's operational experience and does not represent revenue recognized in the second quarter. It highlights the true value of large mainframes and servers, which comes from subsequent data management, transaction processing, maintenance, and support software. Hardware sales are just the starting point, with software dictating whether profits can continue.
The stability of software currently mainly comes from renewals, with about 80% of software revenue derived from subscriptions, usage-based fees, and long-term support; the annual recurring revenue for software is $24.6 billion, up 8% year-on-year. The remaining approximately 20% primarily comes from non-recurring transactions such as permanent licenses and enterprise-level large contracts, which saw a rapid decline in the second quarter. Stable renewals support baseline revenue, while delays in large contracts slowed growth, indicating that IBM will still be influenced by the timing of a few large contracts.
IBM's role in the AI market is mainly to assist large institutions in integrating AI into existing systems and data. Banking, airlines, manufacturing, and government clients need to manage data security, access permissions, and operational records, requiring a longer time from trial to official launch. Management indicated that nearly half of the clients chose configurations that could accelerate AI operations. This shows that customers are beginning to deploy AI in core systems, which cannot be directly equated to a proportionate increase in AI revenue.
To go from equipment to long-term revenue, two steps through software and consulting are required. Cloud software helps clients run AI applications between their own data centers and various cloud platforms, and data tools transmit the latest business information to AI systems; the consulting team then integrates these tools into customer service, finance, supply chain, and software development. If equipment and consulting projects further stimulate software renewals, maintenance, and support, IBM's AI-related revenue will become more stable.
IBM has not separately disclosed AI product revenue; cloud software, data tools, servers, and consulting also serve other businesses. Whether AI can become a continuous source of growth for IBM will depend on the complete conversion process: equipment orders generating revenue, followed by software and consulting growth, gross margins stabilizing, and eventually leading to free cash flow. The second quarter only completed the first half, and the third quarter will need to validate whether software, consulting, and cash can continue to follow up.
4. The Third Quarter Will Verify Whether Software, Profits, and Cash Can Improve Simultaneously
In the third quarter, it is necessary to verify whether the pressure from the second quarter mainly came from timing issues and observe whether revenue, profit margins, and cash flow can improve simultaneously. The adjusted earnings per share for the second quarter increased by 5% year-on-year, while the adjusted gross margin fell from 60.1% to 59.4%. The software gross margin decreased from 83.9% to 82.6%, infrastructure gross margin decreased from 61.5% to 58.4%, and consulting gross margin increased from 27.5% to 28.9%. The reduction in high-margin software large contracts, coupled with the faster growth of servers and storage with lower margins than software, is the main clue to the overall decline in gross margins.
The adjusted pre-tax profit margin rose from 18.9% to 19.2%, in contrast to the gross margin's movement. Management stated that reducing external procurement, increasing sales and marketing efficiency, improving supply chain and project delivery, and internal AI automation jointly drove a slight rise in the adjusted pre-tax profit margin. IBM did not separately disclose the contributions of these measures. They can protect short-term profitability, but it remains to be seen whether revenue growth can replace cost reductions as the primary driver of profit improvement.
Free cash flow reveals another side, as it is the cash that the company can use for dividends, debt repayment, and acquisitions after completing daily operations and necessary investments. The free cash flow for the second quarter was $2.54 billion, a year-on-year decrease of $305 million; for the first half of the year, it was $4.76 billion, remaining basically flat year-on-year. The company still expects free cash flow for the full year to increase by about $1 billion year-on-year; to achieve this goal, the second half requires accelerating the signing of delayed contracts, delivering backlog equipment orders, implementing consulting projects, and improving cash collections. At the same time, the third quarter will validate whether software, profits, and cash can improve together:
The first validation is delayed software contracts. Management stated that approximately one-third of the delayed transactions had been signed in the early third quarter, calculated by the number of transactions. The amount of each transaction varies, and completing signing does not equate to the same proportion of revenue being recognized or cash being collected. A narrowing decline in large mainframe supporting software and renewed growth in IBM's original software would better indicate that demand still exists, and adjustments in the sales team are beginning to yield results.
The second validation is the nearly $500 million backlog of Power series servers and storage orders. IBM has not promised that all these orders will be delivered in the third quarter. When observing the delivery progress, it is necessary to simultaneously look at relevant business revenue, inventory changes, and operating cash collections. Improvements in revenue and cash collections can demonstrate that some orders have been delivered and converted into cash; if inventory remains high, it must be considered in conjunction with new orders and stocking needs; changes in inventory alone cannot lead to conclusions.
The third validation is whether AI consulting contracts can enter the implementation and billing phases. In the second quarter, related projects accounted for about half of new contracts signed in the consulting business and over 30% of unfinished projects. If consulting revenue growth, excluding the impact of exchange rates, exceeds the 1% growth of the second quarter, it can be seen as a signal of improvement in project conversion; simultaneously, the implementation progress of backlog projects and cash collections should also be observed.
Therefore, the simultaneous emergence of IBM's original software returning to growth, accelerated consulting revenue, stabilized gross margins, and improved free cash flow would prove that IBM's revenue structure has genuinely improved.
5. IBM's Value Relies on Core Business Recovery and the Long-Term Potential of Quantum Computing
Overall, the main line in IBM's second quarter is clear: servers, storage, and AI-related projects grew first, while high-margin software, consulting revenue, and free cash flow did not improve simultaneously. The backlog orders for Power series servers and storage are close to $500 million, and cloud software and data tools continue to grow, but IBM has not disclosed the direct contribution of AI. The next step is to see if this demand can extend to software renewals, consulting, and cash collections. Meanwhile, beyond the core business, quantum computing is becoming a long-term variable in IBM's valuation judgment. In May of this year, the Trump administration signed a letter of intent for support with nine companies, planning to provide a total of $2.013 billion in quantum computing support, of which IBM plans to receive $1 billion to build quantum chip manufacturing facilities.
Currently, the core of the debate on Wall Street is whether its orders are merely delayed or whether there has been a more permanent change in enterprise technology budgets. Bank of America and JPMorgan maintain buy or overweight ratings, focusing on software renewals, order replenishment, and cash; BMO and Bernstein maintain neutral ratings, increasingly concerned about software slowdowns and full-year goal execution; HSBC maintains a sell rating. Target prices also reflect this divergence, with Goldman Sachs maintaining a buy rating and lowering the target price from $335 to $270; Morgan Stanley maintains a hold rating and lowers the target price from $293 to $190. Despite differing ratings, target prices have generally declined, indicating that institutions are lowering expectations for short-term growth and realization speed.
IBM’s stock price decline reflects both the pressures on the company’s fundamentals and second-quarter financial results falling below expectations, as well as the macroeconomic environment and overall weakening of the software sector. In the first half of this year, the market worried that the rapid innovation of AI large model companies undermines the product value and revenue growth of traditional software enterprises, resulting in continued sluggish overall performance of software stocks. Against this backdrop, for IBM to reverse market expectations in the third quarter, it needs to not only push delayed contracts to convert into software revenue but also expedite the delivery and collections of backlog equipment orders, facilitating consulting projects to enter implementation stages, thereby driving improvements in gross margins and free cash flow.
If only Power series server and storage revenue bounce back, the market will still worry that growth comes from lower-margin, more volatile equipment businesses; a simultaneous improvement in IBM's original software, consulting revenue, and cash collections will make performance recovery more credible. Meanwhile, CoinW TradFi has launched the IBMUSDT perpetual contract, which is priced and settled in USDT, supports around-the-clock trading, as well as long and short positions and leverage, allowing users to trade directions or hedge risks around financial reports, product releases, policy changes, and sudden market events, providing traders with more options.
The value of IBM needs to be viewed through the lens of revenue realization and innovation potential. Whether the company can continue to convert equipment and project demand brought by AI into software, consulting, and cash will be verified by the third quarter's contract signing, equipment delivery, consulting project implementation, and collection situation; whether quantum computing can leverage policy support to advance manufacturing capability construction and gradually form actual applications and revenues will determine whether IBM can open new growth spaces. The recovery of core business will support valuation, while the ongoing monetization of AI demand and the industrialization of quantum computing will decide whether the market is willing to reassess IBM's growth potential.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
