What is IBM?

What IBM actually does in 2026

Step inside any large bank, insurance company, or government agency in the United States and you will probably find IBM systems behind the scenes. The company sells four things in roughly this order of revenue importance: software subscriptions, consulting engagements, mainframe hardware, and financing for its own customers. None of those are growth-at-any-price businesses. All of them are sticky.

Software is the biggest slice and gets the most attention from sell-side analysts. Inside that segment, the pieces that matter are Red Hat (Linux and the OpenShift container platform), automation tooling, the data and AI stack including watsonx, and a transaction-processing layer that runs on mainframes. Software is where the recurring-revenue narrative lives, and where the multiple expansion thesis ultimately has to play out.

Consulting is the second-largest segment. IBM Consulting deploys advisors and engineers to implement technology programmes inside enterprises: ERP migrations, cloud build-outs, AI rollouts, security architecture work. The revenue is more cyclical than software because corporations cut services budgets first when conditions tighten. Signings versus revenue, often called book-to-bill, is the leading indicator analysts watch every quarter.

Infrastructure includes IBM Z mainframes, distributed servers, and storage. The mainframe still pays the bills here. Hyperscale clouds get the headlines, but the world's largest banks and credit card networks have not migrated their core transaction systems off Z, and most of them never will. Mainframe revenue oscillates around a two to three year hardware cycle tied to new product launches. The Financing segment, by contrast, is a small lender-of-last-resort that helps customers acquire IBM hardware and software, roughly 1 percent of revenue and largely a rounding error in the thesis.

How the business actually makes money

The thing to internalise about IBM is that the company has spent five years rebuilding itself around two acquisitions and one divestiture. The 2019 purchase of Red Hat for 34 billion dollars is the strategic foundation everything else now sits on top of. The 2021 spin-off of Kyndryl, the managed-infrastructure-services arm, removed a lower-margin business and tightened the focus on software and consulting. The 2024 separation from GRAIL (inherited via the Illumina entanglement story, then routed away) cleaned up the regulatory overhang on a different front.

What remains is a free-cash-flow machine. Annual free cash flow guidance has been running in the 13 billion dollar range, which funds the quarterly dividend, services the debt taken on for Red Hat, and leaves room for tuck-in software acquisitions. The dividend itself is the largest single capital return mechanism. IBM has raised it for around three decades, and the yield typically sits between 3.0 and 4.0 percent, putting the stock in dividend-aristocrat territory and giving it a buyer base that does not look like the buyer base for most other large-cap tech.

Gross margin tells the segment story cleanly. Software runs at gross margin in the high seventies to low eighties. Consulting sits in the high twenties to low thirties. Infrastructure varies with the hardware cycle. The mix shift toward Software is therefore the single most important driver of consolidated margin expansion over time, and analysts model it explicitly when projecting the multi-year operating model.

Watsonx is the bet on the next leg. The platform has three components: watsonx.ai for foundation-model training and deployment, watsonx.data for the underlying data store, and watsonx.governance for model risk management. The pitch is enterprise-grade AI for regulated industries where data residency, lineage, and on-premises deployment are non-negotiable. IBM discloses a cumulative AI book of business at every quarterly print, and the slope of that line is the cleanest available read on whether the watsonx story is working.

The earnings cadence and what gets revealed

IBM reports four times a year, always after the regular session closes. The pattern is reliable: late January for the fourth quarter, late April for the first, late July for the second, and late October for the third. The conference call begins at 5:00 PM Eastern Time and runs for around an hour. Sell-side notes hit the wire by the next pre-market.

Five things move on print night in roughly this order of importance. First, the Software segment revenue growth rate, particularly the constant-currency figure and the contribution from Red Hat. Second, the Consulting book-to-bill ratio, which tells you whether the services backlog is growing or shrinking relative to revenue being recognised. Third, free cash flow against the annual guide, because the entire capital-return thesis hinges on this single number. Fourth, the watsonx generative AI book of business, disclosed cumulatively and growing each quarter. Fifth, any quantitative refresh to the full-year guidance.

Historical earnings-day moves on IBM have been less violent than for the growth-tech peer set. A 4 to 9 percent same-session move is the typical range when guidance shifts materially. Misses on Software growth or Consulting book-to-bill tend to produce the larger declines, while upside surprises on FCF and watsonx bookings have driven the rallies of the last two years. Volume on print nights and the following morning runs at three to five times the trailing daily average.

The other recurring date is the mainframe cycle. IBM has signalled the z17 platform refresh, with the z16 cycle now in its mature phase. Whenever a new Z is launched, the Infrastructure segment shows a large positive year-on-year revenue contribution for roughly four quarters and then decelerates into the next cycle. Modelling IBM without smoothing across the mainframe cycle is a common rookie mistake.

What actually moves the share price

IBM does not trade on a single macro factor. The share price reflects a blend of segment-mix progress, capital-return discipline, and where the hardware cycle happens to sit. Five drivers do most of the work.

When IBM is liquid and when it is not

IBM trades during the New York Stock Exchange regular session, 9:30 AM to 4:00 PM Eastern Time, Monday through Friday. Pre-market and post-market sessions exist on the underlying equity but are not part of the CFD trading window at LHFX. The most active block is the first hour after the open and the final 90 minutes before the close, when index-rebalancing flow and end-of-day positioning move large-cap names.

The single highest-volume window of any quarter is the 4:00 PM Eastern release of the earnings press release and the 5:00 PM conference call. On a print night the underlying ADV runs at three to five times average. Spread widens around the open and tightens through the morning; on a typical session, the IBM CFD spread at LHFX is one of the tighter US single-stock CFD spreads available, owing to the size of the underlying equity and the depth of book on NYSE.

Month-end and quarter-end sessions see additional flow tied to index reweighting and dividend record dates. IBM is a constituent of the Dow Jones Industrial Average, the S&P 500, the S&P 100, and many sector and dividend indices. The Dow membership in particular brings index-fund flow through the closing auction on rebalance days.

CFD on IBM versus owning the share

There is no futures market on individual large-cap US single names like IBM, so the realistic comparison is between a CFD position and direct cash-equity ownership. The mechanics differ on several dimensions.

The CFD is the tactical vehicle. Cash equity is the long-term holding vehicle. Most traders who use both treat them as complementary, not substitutable.

Trading IBM as a CFD versus owning the share

If you buy 50 shares of IBM through a brokerage account, you become a registered shareholder. You receive the quarterly cash dividend, you can vote at the annual meeting, you can lend the shares to short sellers, and you can hold the position indefinitely without any overnight financing charge. The downside is that you cannot easily go short, the leverage available is modest (usually 1:2 inside a US margin account), and you pay a per-trade commission plus a regulatory fee on every fill.

A CFD on IBM is a different instrument that references the same share price. You and the broker agree that the cash difference between the price when you open the position and the price when you close it will be settled to your account. You do not appear on the IBM shareholder register. You do not receive dividend cash. You do, however, receive a dividend adjustment to your CFD position on the ex-dividend date: long positions are credited the equivalent amount, short positions are debited. The mechanism replicates the economic effect of the dividend without the legal ownership.

The other practical differences are in leverage, financing, and shorting. CFD leverage at LHFX on a US single-stock name like IBM is capped at 1:20, which is materially higher than the cash-equity margin a retail broker will extend. Overnight financing is charged on any open CFD position held past the daily rollover (around 5:00 PM Eastern Time), with the rate based on the relevant benchmark plus a spread. Shorting is symmetric to going long: same leverage cap, same commission, same execution venue. There is no locate requirement and no separate short-borrow fee structure.

One thing the CFD route is not designed for is long-term buy-and-hold accumulation of dividend income. The overnight financing cost compounds against you over months and quarters. If the thesis is to own IBM for the yield and the capital-return profile across multiple years, cash equity is the better vehicle. If the thesis is a tactical view around earnings, the watsonx narrative, the mainframe cycle, or a sector rotation, the CFD is the more efficient structure.

Trading IBM at LHFX

IBM at LHFX is a CFD on the NYSE-listed share, routed STP/ECN and quoted in MetaTrader 5. The contract specs below are the working set every position depends on, from margin sizing to overnight cost.

Full contract details live on the IBM instrument page, with cost breakdowns at spreads and fees and margin tables at leverage.

What can go wrong on this trade

Three risks deserve explicit attention because they are specific to IBM rather than generic to single-stock CFDs. Treat them as standing inputs to position sizing rather than as tail events to ignore.