IBM Q3 Earnings Preview: Can Enterprise AI Accelerate Software Growth and Cash Flow?

 

IBM reports third quarter results and holds its earnings call at 5:00 p.m. Eastern on October 21, which is 05:00 Singapore time on October 22. That timing is confirmed on the company's investor relations page.

 

The company sits in an unusual position within the AI cycle. It carries none of the hundred-billion-dollar capital spending burden of the hyperscalers, yet it has to prove that enterprise AI budgets eventually land on its own income statement as software revenue and cash flow. The most recent quarter pointed the other way. Second quarter revenue of $17.16 billion grew just 1%, software growth slowed from 11% in the first quarter to 5%, and the company cut its full-year constant-currency revenue growth guidance from above 5% to a range of 4% to 5%.

 

The share price already reflects that doubt. IBM closed at $236.39 on October 19, up 1.83%, for a market capitalisation near $222.71 billion at a trailing price to earnings ratio of 21.02 and a forward multiple of 18.55, with a 2.86% dividend yield against a 52-week range of $199.19 to $332.46, leaving it roughly 29% below the high. Per Zacks, the stock fell 17.8% over the past year while trailing revenue still grew 7.9%.

 

What this report really has to answer is not whether the AI order book is large enough, but how long the gap between cumulative bookings and recognised revenue will take to close.

 

 

 

Timing and the bar. The call is at 5:00 p.m. Eastern on October 21, or 05:00 Singapore time on October 22, against a MarketBeat consensus of $16.82 billion in revenue and $2.85 in earnings per share, versus $16.33 billion and $2.65 a year earlier.

 

Last quarter decelerated. Second quarter revenue of $17.16 billion grew only 1% and came in about 1.8% below consensus, with software at $7.76 billion up 5%, consulting at $5.33 billion essentially flat, and infrastructure at $3.84 billion down 7.4%.

 

The AI figure is a cumulative metric. The most recent public number is more than $12.5 billion inception to date at the end of 2025, defined as software transactional revenue plus new SaaS annual contract value plus consulting signings tied to specific offerings, which is not revenue recognised in any single quarter.

 

Generative AI is already half of consulting signings. Second quarter consulting signings rose 6% with generative AI roughly half of them and more than 30% of consulting backlog, while consulting revenue grew 0%.

 

All the cash flow growth sits in the second half. First half free cash flow of $4.8 billion was flat year over year against full-year guidance of roughly $1 billion of growth on the $14.7 billion delivered in 2025.

 

The valuation premium persists. The forward multiple is 18.55, or 17.73 on the Zacks basis, against an industry average near 10.74.

 

 

The call is scheduled for 5:00 p.m. Eastern on October 21, which lands early on October 22 for Asian investors. IBM usually releases results after the close and holds the call immediately afterwards, so the numbers and management's commentary reach the market at virtually the same time.

 

The consensus bar is not demanding. MarketBeat's compilation calls for revenue of $16.82 billion and earnings of $2.85 per share, implying roughly 3% revenue growth and 7.5% earnings growth. Zacks looks for $16.84 billion and $2.87, holding the full-year estimate at $12.33 and 2027 at $13.20. Worth noting is that the third quarter is seasonally softer than the second for IBM, so a consensus below last quarter's $17.16 billion reflects normal cadence rather than further deterioration.

 

Three things actually matter. First, whether software growth can return from 5% toward the trajectory implied by full-year guidance of 6% to 8%. Second, whether consulting revenue begins to catch up with signings, which is the direct test of AI orders converting into revenue. Third, the quarterly free cash flow figure, because full-year guidance loads essentially all of the growth into the second half.

 

IBM has beaten the earnings consensus in seven of the last eight quarters but merely matched it in the most recent one. Zacks assigns an Earnings ESP of negative 0.63% and a Hold rating, indicating the sell side is not positioned for a meaningful beat.

 

 

The second quarter showed three entirely different curves inside one company.

 

Software revenue of $7.76 billion grew 5% but fell short of roughly $7.99 billion in consensus. Per Futurum's breakdown, about 80% of this segment is recurring, with recurring software annual recurring revenue reaching $24.6 billion up 8%, Red Hat growing 11%, OpenShift annual recurring revenue at $2.2 billion, and HashiCorp posting a record bookings quarter. The drag came from the remaining 20%, transactional revenue tied largely to mainframe-associated enterprise license agreements, which therefore moves with the hardware cycle. Short-term swings in software growth consequently need not reflect software demand itself.

 

Consulting revenue of $5.33 billion was flat year over year, up 1% at constant currency. Signings rose 6%, with generative AI roughly half of them and more than 30% of backlog. That gap between signings and revenue is the centre of this preview and gets its own section below.

 

Infrastructure revenue of $3.84 billion fell 7.4%, with IBM Z revenue down 42%. That number looks alarming until it is read against first quarter growth of 48%. The z17 program is running at nearly 130% of z16 at the same point in its cycle, management calls it the strongest mainframe cycle to date, installed MIPS exceed 140 million, and clients representing 85% of installed MIPS are maintaining or growing capacity. Infrastructure's issue is the comparison base rather than demand, though it feeds back into the software segment through transactional revenue, a linkage easily missed when reading the segments in isolation.

 

Financing revenue of $186 million grew 12.2%, small in absolute terms but a steady spread contributor in a higher-rate environment.

 

 

This is where reading IBM goes wrong most often, and it is the concept most worth settling before the report.

 

The most recent publicly disclosed AI figure is more than $12.5 billion inception to date at the end of 2025, given in the fourth quarter release. That release also supplies the definition: the metric equals software transactional revenue, plus new SaaS annual contract value, plus consulting signings tied to specific offerings. Those three components have entirely different accounting characters. Software transactional revenue is recognised revenue, new SaaS annual contract value is contracted value for the coming twelve months, and consulting signings are total contract value spread over several years. Adding them produces a measure of momentum, not revenue for any period.

 

More importantly, the figure is cumulative since inception. That $12.5 billion spans multiple years and does not sit on the same time axis as roughly $17 billion of quarterly revenue, so comparing the two directly yields badly distorted conclusions. IBM has stopped printing the number in the body of its releases as of the second quarter of 2026, moving the detail into exhibit tables, which makes whether and how it reappears this quarter an observation point in its own right.

 

Consulting provides the clearest sample of the gap. Generative AI accounts for roughly half of signings and more than 30% of backlog, and management noted in the first quarter that about 80% of the generative AI consulting book comes from net new clients. Those are all constructive signals. Over the same period, consulting revenue grew between 0% and 1%. The reason is straightforward, since consulting revenue is recognised as work is delivered, and large enterprise AI programs carry long evaluation periods and phased rollouts, leaving several quarters between a signature and recognised revenue.

 

A more reliable way to judge conversion is therefore to watch whether the spread between signings growth and revenue growth is narrowing or widening. Consulting revenue moving toward low to mid single-digit growth would say earlier signings are entering delivery. Signings continuing to grow while revenue stalls would mean programs are landing more slowly than expected, and in that case a rising cumulative book would actually obscure the problem.

 

 

On margins, second quarter gross margin of 57.7% fell 1.0 point year over year, while the non-GAAP pre-tax margin of 19.2% improved from 18.8%, non-GAAP earnings of $2.93 per share rose 5% and GAAP earnings of $2.27 fell 2%. The gap between the two measures comes mainly from acquisition-related intangible amortisation and restructuring charges, a persistent feature at IBM, which is why one basis has to be used consistently.

 

Cash flow is the part worth working through with a calculator. Second quarter free cash flow was $2.5 billion, down $0.3 billion year over year, bringing the first half to $4.8 billion, flat against the prior year. The company maintains guidance for full-year free cash flow roughly $1 billion above the $14.7 billion delivered in 2025. On that basis the year needs about $15.7 billion, leaving roughly $10.9 billion for the second half against about $9.9 billion in the second half of 2025. All of the year's growth therefore sits in the back half, with no cushion built in the first.

 

Third quarter free cash flow consequently says more than revenue. IBM's cash generation genuinely skews to the second half and especially the fourth quarter, a function of enterprise IT spending patterns. But zero growth in the first half means that if third quarter cash flow is again flat year over year, the full-year guide would require an extreme share of the increase to arrive in a single quarter, and the difficulty of hitting it rises sharply.

 

Two balance sheet figures deserve attention. Cash, restricted cash and marketable securities stood at $8.2 billion, down $6.3 billion from the end of 2025, while total debt reached $62.0 billion, up $0.7 billion year to date and including $13.0 billion of IBM Financing debt. With a 2.86% dividend yield to fund alongside acquisitions, the direction of the cash balance is worth tracking.

 

This also shows where IBM structurally differs inside the AI cycle. Its AI model runs through software and services rather than hyperscale data centre construction, so the pressure on cash flow comes from the speed of revenue conversion rather than a build-out schedule. Compute-side participants, by contrast, need elaborate financing structures to fund expansion, and the mechanics of those arrangements are covered in our analysis of AI infrastructure financing. The risk profiles differ completely, with insufficient growth on one side and leverage plus depreciation on the other.

 

 

The guidance path across three quarters is itself a piece of information. Alongside full-year 2025 results in January 2026, the company guided 2026 to constant-currency revenue growth above 5% with currency expected to add about half a point. April's first quarter kept the above-5% framing. July's second quarter cut it to 4% to 5% and moved the currency assumption to neutral for the year.

 

At the segment level, per Futurum's summary of second quarter guidance, software is guided to 6% to 8% with the low end assuming current enterprise spending dynamics persist, consulting to low to mid single-digit growth accelerating through the year, and infrastructure to low single-digit growth, with pre-tax margin expansion of roughly 100 basis points after an upward revision. The target of roughly $1 billion of free cash flow growth has never moved.

 

That combination explains the valuation standoff. A forward multiple of 18.55 is not cheap for single-digit revenue growth, and the 17.73 figure on the Zacks basis carries a clear premium to the industry average near 10.74. Supporting the premium are margin improvement from a rising software mix, double-digit Red Hat growth, and a meaningful dividend plus buyback. Undermining it are decelerating revenue alongside a cut guide, with AI orders yet to show up fully in reported results.

 

The average target among 25 analysts is $242.65, only about 2.65% above the October 19 close, with a Buy consensus overall. A target that close to the current price says the sell side sees no directional conclusion at this level either, which leaves the report itself as the main near-term input to pricing.

 

 

What this earnings season is really repricing is the lag between AI investment and AI return, and different companies occupy different points along that lag.

 

Hyperscale platforms sit on the investment side. When Meta reported second quarter results on July 30, it raised the low end of its 2026 capital expenditure range to leave it at $130 billion to $145 billion, while net income fell 14% to $15.85 billion. The central question for that cohort is the payback period, discussed further in our coverage of Meta earnings and AI spending.

 

Semiconductor and equipment makers sit on the receiving side, where revenue and margins expand together during a build-out while valuations stay acutely sensitive to cycle length.

 

IBM occupies a third position, and the hardest one to price. It is neither the supplier of compute nor the principal bearer of AI spending, but the implementer and software vendor for enterprises fitting AI into existing processes. The advantage is light capital intensity with steady cash flow and dividend support. The disadvantage is that revenue recognition depends on how fast customer programs advance, which is exactly the category of spending enterprises defer first when the macro picture turns uncertain.

 

IBM's results therefore carry unusual informational value for the whole chain. Hyperscaler capital spending tells the market how much supply is being built, equipment orders tell it when capacity arrives, and IBM's consulting revenue plus software transactional revenue tell it whether enterprise customers are actually using any of it. That third signal is currently the weakest of the three, which is precisely why October 21 matters.

 

 

 

For James Mitchell, the value of this report lies less in whether IBM clears $2.85 than in the fact that it is one of the few samples in the AI narrative that directly tests enterprise adoption. Hyperscaler capital spending measures conviction, equipment orders measure capacity, and only enterprise software and consulting revenue measures actual usage. Right now conviction is strongest, capacity second and usage weakest, and that ordering is the source of the valuation disagreement.

 

The easiest misreading involves the metric itself. The cumulative figure of more than $12.5 billion adds recognised revenue, twelve months of contracted value and multi-year signings into one inception-to-date total, so it measures momentum rather than revenue. Setting it beside quarterly revenue, or inferring reported growth from its growth, leads to the wrong conclusion. The sturdier approach watches only two spreads: consulting signings growth minus consulting revenue growth, and recurring software growth minus total software growth. The first measures conversion efficiency, the second strips out noise from the mainframe cycle.

 

Cash flow is the second focus that tends to be overlooked. First half free cash flow of $4.8 billion was flat while full-year guidance calls for roughly $1 billion of growth, which implies about $10.9 billion in the second half against roughly $9.9 billion a year ago. IBM's cash generation does concentrate late in the year, but a flat first half has already consumed the cushion. If third quarter cash flow is flat again, what comes under pressure is not the quarter but the credibility of the full-year guide, and for a stock held substantially for its dividend and buyback, the credibility of the cash flow guide matters more than the revenue growth rate.

 

Across assets, IBM functions as a downstream verifier of the AI cycle rather than an upstream beneficiary. That role tends to be persistently underpriced early on, because capital naturally flows to higher-beta links in the chain. Once the market starts demanding evidence of return, however, its revenue data becomes one of the most frequently cited exhibits. The real question for investors is therefore not whether IBM clears the bar this quarter, but whether the slope of the enterprise adoption curve is steepening or flattening. One report cannot settle that, though it will provide the clearest reading currently available.

 

 

 

The company releases results and holds its call at 5:00 p.m. Eastern on October 21, which is 05:00 Singapore time on October 22, as confirmed on IBM's investor relations page. Results typically arrive after the US market close with the call immediately afterwards, so figures and management commentary reach the market almost simultaneously, and after-hours moves tend to cluster around those two moments.

 

 

MarketBeat's consensus calls for revenue of $16.82 billion and earnings of $2.85 per share, against $16.33 billion and $2.65 a year earlier, implying roughly 3% revenue growth and 7.5% earnings growth. Zacks looks for $16.84 billion and $2.87 while holding the full-year estimate at $12.33. IBM has beaten the earnings consensus in seven of the last eight quarters but matched it in the most recent one.

 

 

The most recent public figure is more than $12.5 billion inception to date as of the end of 2025. The definition matters: it equals software transactional revenue, plus new SaaS annual contract value, plus consulting signings tied to specific offerings, accumulated since the program began. It measures business momentum and does not equal revenue recognised in any single quarter. IBM has stopped printing the number in the body of its releases as of the second quarter of 2026.

 

 

Because consulting revenue is recognised as work is delivered. Second quarter signings rose 6% with generative AI roughly half of them and more than 30% of backlog, yet consulting revenue was flat year over year. Large enterprise AI programs carry long evaluation periods and phased rollouts, leaving several quarters between signature and revenue. The way to judge conversion is to watch whether the spread between signings growth and revenue growth narrows or widens.

 

 

Second quarter software revenue of $7.76 billion grew 5%, down from 11% in the first quarter. About 80% of software revenue is recurring, and that portion reached $24.6 billion in annual recurring revenue up 8%, with Red Hat growing 11%. The slowdown came from the remaining 20% of transactional revenue, which is tied to mainframe-associated enterprise license agreements and therefore moves with the hardware cycle. Recognising this avoids mistaking a hardware cycle effect for weakening software demand.

 

 

Because all of the year's growth sits in the second half. First half free cash flow of $4.8 billion was flat year over year while the company maintains guidance for roughly $1 billion of growth on the $14.7 billion delivered in 2025. That implies about $10.9 billion in the second half against roughly $9.9 billion a year earlier. IBM's cash generation does concentrate late in the year, but a flat first half leaves no cushion, so the third quarter figure feeds directly into the credibility of the full-year guide.

 

 

It depends on the reference point. IBM trades at a forward multiple of 18.55, or 17.73 on the Zacks basis, against an industry average near 10.74, a clear premium. Supporting it are a rising software mix, guidance for roughly 100 basis points of margin expansion and a 2.86% dividend yield. Working against it are single-digit revenue growth and a full-year guide already cut from above 5% to a 4% to 5% range. The average target among 25 analysts is $242.65, only about 2.65% above the October 19 close.

 

 

The information above is provided for general market information and analysis only and does not constitute investment advice, financial advice, legal advice, tax advice or a recommendation to trade. Prices of equities, crypto assets and other related financial assets can fluctuate sharply, and past performance, financial metrics, analyst estimates and company guidance do not guarantee future results. The financial data, business metrics, guidance and share prices cited here reflect publicly available information at the time of publication, while actual results will be defined by IBM's official release and investor relations materials. The book of business and signings metrics referenced here use the company's own definitions, which differ from revenue recognition under accounting standards and should not be treated as equivalent. Stock perpetual futures and similar derivatives carry leverage, volatility expands sharply around events such as earnings, and losses can exceed the initial outlay. Readers should conduct their own research and make decisions based on their own financial circumstances, investment objectives and risk tolerance, consulting a qualified professional where appropriate. The MEXC Crypto Pulse team accepts no liability for any direct or indirect loss arising from the use of this information.

 

 

James Mitchell specializes in technical analysis, market trends, and trading strategies for both Bitcoin and altcoins. Based in London, he has over 10 years of experience in financial markets. Before joining MEXC Learn, James worked as a senior analyst at a leading European investment firm, where he developed expertise in risk management and quantitative trading. His transition to cryptocurrency markets began in 2017, and he has since become recognized for his data-driven approach. He holds a Master's degree in Financial Economics from the London School of Economics. His analytical approach combines traditional technical analysis with on-chain metrics to provide readers with actionable insights.

 

Areas of Expertise: Technical Analysis, Market Trends and Cycles, Trading Strategies, Bitcoin and Altcoin Analysis, Risk Management.

 

 

 

Want the fastest access to MEXC's latest updates? Join our official Telegram group now!

MEXC Platform: Download App | Zero Fee Trading | Buy USDT Now

Join MEXC Community: X (Twitter) | Telegram | Discord

Get Insights: MEXC Learn | Terms of Service | MEXC Blog | MEXC News

Account Verification: Understand KYC | How to Complete KYC

External Content Platforms: Substack | Medium | Paragraph | LinkedIn | X(News)