Is Synopsys a Good Long-Term Investment? SNPS Growth, Profitability and Risks

Synopsys guided fiscal 2027 revenue to USD 11.10bn–11.20bn at its 2026 Investor Day in New York on 1 October, against a consensus of roughly USD 10.81bn — a midpoint of USD 11.15bn and about 15% growth. Adjusted EPS was guided to USD 19.04–19.12 against a USD 17.81 consensus on the same basis. Alongside it came two commercial agreements that change what kind of company this is: a multi-year Amazon Web Services deal valued at more than USD 1bn, and a multi-year OpenAI partnership that shares revenue on model usage. The shares closed roughly 13% higher.

The question of whether Synopsys is a good long-term investment therefore turns on something more specific than the guide. For years the investment case was a toll booth: chip designers cannot design chips without electronic design automation software, Synopsys and Cadence supply almost all of it, and revenue grows with design complexity at very high margins. The AWS and OpenAI agreements move the company from collecting a toll to participating in its customers' AI economics — licensing silicon IP into Amazon's own Nitro, Graviton and Trainium chips, and taking a share of revenue when customers use an AI model built on its design data. The numbers support a durable franchise getting a second growth engine. What has to go right is that the USD 11.15bn midpoint proves to be the first year of a re-rating rather than a pulled-forward one-off, which is a Growth and Profitability question before it is a valuation one.

What actually happened

Three things landed on the same day, which is why the reaction was the size it was.

First, the guide. Fiscal 2027 revenue of USD 11.10bn–11.20bn, roughly 15% growth at the midpoint and above the USD 10.81bn the market expected. Adjusted EPS of USD 19.04–19.12, comfortably above the USD 17.81 adjusted consensus. Free cash flow for the year is targeted at approximately USD 3.1bn.

Second, the long-term financial model. Management set out roughly 15% compound annual revenue growth from fiscal 2026 through fiscal 2030, with non-GAAP EPS growing in the mid-20% range over the same period — a framework that implies earnings compounding materially faster than revenue. The margin path underwrites it: a non-GAAP operating margin of about 44% in fiscal 2027, rising to about 50% by fiscal 2030. On capital returns, Synopsys expects to return up to 50% of free cash flow through share repurchases after reinvestment, with the stated intention of buying back approximately USD 1bn of stock over the coming months.

Third, the two agreements. The AWS deal is multi-year and valued at more than USD 1bn: Amazon licenses Synopsys silicon intellectual property and engineering software for its custom Nitro, Graviton and Trainium silicon, and becomes lead customer for an expanded application-optimised IP business. The OpenAI partnership is a multi-year arrangement to co-develop ‘GPT-Synopsys’, a model aimed at semiconductor design workflows, under which revenue is shared based on customer use.

The market read that package as a step change and marked the stock up about 13% on the day — a move normally reserved for an earnings surprise, not an investor day.

Why the market reacted the way it did

An investor day is usually a presentation of things already known. This one was not, for two reasons.

The first is the gap between the guide and consensus. Roughly USD 340m of revenue and USD 1.25 of adjusted EPS above expectation is a large reset for a company of this size, and because both figures were quoted on the same adjusted basis as the consensus they replaced, the beat is real rather than a basis artefact. On the margin path, the move from about 44% to about 50% over four years is the mechanism that turns 15% revenue growth into mid-20% earnings growth. That is the single most consequential number disclosed, and it is a forecast rather than a result.

The second is what the AWS and OpenAI deals do to the perceived ceiling. A toll booth has a natural cap: revenue grows roughly with the number and complexity of designs. Both agreements break that framing. The AWS deal is not a software licence renewal — it puts Synopsys IP inside Amazon's own accelerators, which means revenue scales with Amazon's custom silicon volumes rather than with its design-tool seat count. The OpenAI arrangement goes further by sharing revenue on model usage, which converts a fixed-licence relationship into something closer to a royalty on the productivity its customers extract.

There is also a positioning element. Chip-design software had been the subject of a running debate about whether AI tooling eventually disintermediates the tool vendors themselves. Having the two most visible AI buyers in the world sign multi-year deals with the incumbent is the most direct available rebuttal, and it arrived on the same day as an above-consensus guide.

The Openbook read

Momentum strengthens. A 13% single-day move on an investor day rather than an earnings release is a repricing of the forward model, not a reaction to a data point. That said, a move built on a five-year framework is a promise, and Momentum built on promises is the kind that reverses on the first quarter that misses the margin path. The score improves; its quality is lower than the magnitude implies.

Growth is the factor that changes most, and in the most defensible way. A 15% guided year against a consensus implying roughly 11% growth — USD 10.81bn on a fiscal 2026 base guided to USD 9.72bn — is a real step up, if a narrower one than the 13% move suggests, and the long-term model extends that rate to fiscal 2030. The AWS deal gives the near term a named, contracted anchor worth more than USD 1bn across its term. The honest caveat is that a multi-year licence recognised across several years can front-load a single fiscal year, and this is exactly the distinction between a re-rating and a pull-forward. Growth improves clearly; the verification comes from the second year, not the first.

Profitability is already exceptional and is guided higher. A non-GAAP operating margin around 44% in fiscal 2027 is in the top tier of listed software, and the path to about 50% by fiscal 2030 would be remarkable if delivered. Two labelling points matter here. These are non-GAAP figures and the gap to GAAP in this business is wide, largely because of stock-based compensation and acquisition-related amortisation — so the 50% target should not be compared with a GAAP margin from any other company. And the OpenAI revenue-share model is, by construction, lower margin than a software licence at the point of sale. A revenue mix that shifts toward usage-based sharing is accretive to revenue growth and dilutive to incremental margin. The Profitability score stays strong; the margin path is the part of the model carrying the most execution risk.

Solvency is not the binding constraint. A business targeting approximately USD 3.1bn of free cash flow in fiscal 2027 and committing to return up to half of it through buybacks is operating with substantial headroom. The item worth tracking is not leverage but the buyback itself: announcing an intention to repurchase roughly USD 1bn over the coming months, on the same day the stock rose 13%, means management is buying after the re-rating rather than before it.

Reward/Risk is where the long-term question resolves, and the picture is genuinely two-sided after a 13% day. The reward case is that a near-duopoly supplier with mid-40s operating margins has just added a usage-linked revenue stream with the two most aggressive AI buyers in the market, and guided to compounding earnings in the mid-20% range for four years. The risk case has three specific components. The guide is a forecast, not a result, and the fiscal 2027 print is a year away. The margin path from 44% to 50% has to survive a revenue mix shifting toward lower-margin revenue sharing. And a large share of the enthusiasm is attached to two customers whose own capital-spending plans are the most volatile variable in technology — concentration that cuts both ways. A reader comparing this against the rest of the sector will find the relevant multiples on our screener; our full factor detail sits on the Synopsys page.

The read-across

The clearest beneficiary besides Synopsys itself is Cadence Design Systems, which rose on the day. Synopsys and Cadence together supply the large majority of commercial EDA software, and the two share the same structural argument: AI makes chips more complex, complexity makes design software more necessary, and the barriers to entering this market are formidable. A credible 15% growth framework from one half of that pair raises the plausible growth rate for the other.

The caution on that read-across is that the two differentiating announcements are not shared. The AWS silicon-IP licence and the OpenAI revenue-share model are Synopsys-specific, and the IP business in particular is where the two companies' portfolios diverge most. A peer re-rating on the strength of Synopsys's deals would be a sympathy move, not a fundamentals one, until Cadence announces something comparable.

For the custom-silicon chain more broadly, the AWS deal is a disclosure about scale. Amazon is willing to commit more than USD 1bn over several years to license external IP for Nitro, Graviton and Trainium — which says the custom accelerator roadmap at the largest cloud provider is big enough to justify that outlay. That is a positive signal for the semiconductor IP vendors and foundry-adjacent names that sit in the same flow, and a reminder that hyperscaler in-house silicon is now a volume business rather than an experiment.

What to watch next

• Fourth-quarter and full-year fiscal 2026 results, due around early December 2026. This is the first checkpoint on whether the fiscal 2027 guide was built on a clean base. Watch the fiscal 2027 revenue range for any narrowing: a move in the floor would confirm the AWS contribution is contracted rather than assumed.
• Non-GAAP operating margin against the roughly 44% fiscal 2027 target. This is the load-bearing number in the whole model. Quarterly margin prints are where the path to 50% by fiscal 2030 either becomes credible or does not.
• Disclosure on the revenue recognition profile of the AWS agreement. How more than USD 1bn is spread across the term determines whether fiscal 2027's 15% is repeatable in fiscal 2028.
• Any revenue detail on ‘GPT-Synopsys’. A revenue-share arrangement is only a growth engine once it is quantified; until then it is an option with no strike price disclosed.
• Execution of the approximately USD 1bn buyback over the coming months, and whether repurchases are completed at the post-announcement price or paced.
• Free cash flow against the approximately USD 3.1bn fiscal 2027 target, which is the number that funds both the buyback and the reinvestment the 15% growth rate assumes.

What happened on 1 October was not a quarter beating expectations. It was a company restating what business it is in and attaching a four-year financial framework to it, with two of the most credible possible customers co-signing the premise. The franchise argument has rarely looked stronger. The open question is the ordinary one that follows a 13% day: how much of a four-year model has already been paid for in advance, and whether the margin path from 44% to 50% survives a revenue mix deliberately shifting toward sharing the upside with customers.