Cadence Design Systems

Cadence Design Systems' second-quarter revenue grew 24% year over year to $1.58 billion, while non-GAAP operating margin expanded to 45.5%. Results were generally ahead of expectations, and management increased its full-year guidance.

Why it matters: Cadence reported results that were mildly ahead of our expectations and raised its full-year guidance. Current numbers and management's commentary are framing artificial intelligence as a tailwind for the firm and the electronic design automation industry, not an existential threat. Overall EDA revenue was up 19% year over year, roughly in line with our expectations, while system design and analysis and intellectual property revenue both beat our expectations, up 32% and 43%, respectively. IP growth was particularly impressive and is likely tied at least partially to Cadence's improving position in key foundry ecosystems, such as Intel and Samsung. SD&A growth also supports our overall thesis on incremental growth from systems design workflows.

The bottom line: We are increasing our fair value estimate for wide-moat Cadence to $350 from $330. Shares appear fairly valued. We think Cadence can grow at a low- to mid-teens revenue compound annual rate while expanding operating margins. All growth drivers seem to be firing. There are worries about the implications of recent chip design efforts by model companies using their own internal models to aid in chip design. Management argued that large language models will be used as enablers in Cadence's proprietary workflows, which we agree with. We thought the points on the need for deterministic, 100% accurate, and physically ground tools were important. We also highlight other issues with competing directly with the EDA firms, such as getting functional, advanced process design kits. We were encouraged by details on AI agent workflows and the ability for the EDA firms to charge for them. We see this as a robust revenue driver that is still in the early innings.

Our fair value estimate of $350 per share implies a 2026 adjusted price/earnings multiple of 43 times and an enterprise value/sales multiple of 14 times.

Our fair value estimate assumes Cadence will be able to increase revenue at a 12% compound annual growth rate over the next decade, above the historical mid- to high-single-digit rate that the overall semiconductor industry has typically seen. This is based on increases in design starts from AI-driven designs, rising chip complexity leading to greater use of EDA tools, increased demand for IP blocks, a slowdown in Moore’s law resulting in heightened demand for performance improvements through design changes, and an increase in systems companies utilizing Cadence’s offerings. We also expect Cadence will monetize the additional performance and efficiency value its AI tools bring, providing room for additional pricing power. Finally, we think the physical AI buildout will also be a key, long-term driver of demand.

Within our overall revenue growth projections, we assume that the core EDA business grows at a 10-year CAGR of 11%, IP grows at 14%, and systems design grows at 15%.

We expect operating margin to continually increase for Cadence, from 31% in 2025 to 40% by 2035. We see no reason to expect margin expansion to stop by 2035, either. This will be driven by strong growth in the IP business (where we expect margins to improve as the business scales), AI-driven efficiencies, pricing opportunities through selling higher-value-added products (value-add increases driven by increased demand along with product innovation, like AI enhancements), and more typical scale-based efficiencies (software companies do not need to double the salesforce or G&A to double sales, for example).

A stable business model and low cyclicality in revenue because of resilient demand during economic downturns give us confidence in assigning a 7.5% cost of equity.

We believe Cadence warrants a Wide Morningstar Economic Moat Rating due to switching costs and intangible assets associated with its vast product suite. Cadence provides leading-edge electronic design automation tools and intellectual property critical to the semiconductor chip design process. We believe that Cadence enjoys high customer switching costs and intangible assets in its sticky EDA segment, as well as intangible assets associated with its IP offerings. In our opinion, this customer stickiness and profitable IP portfolio will drive excess returns on capital for Cadence over the next 10-20 years. Fundamentally, we believe that as long as the semiconductor industry exists and is innovating, Cadence will be a necessary part of that innovation, earning excess returns on invested capital.

Electronic design automation software is what engineers use to design and build semiconductor chips. It is indispensable, functioning as the backbone for developing advanced microchips. Without EDA software, there would be no modern semiconductor industry. If a company is designing semiconductor chips, or its overall systems are heavily reliant on semiconductors (think modern automobiles), it is using EDA software.

The EDA industry is dominated by three key players that control over 90% of the market: Synopsys, Cadence, and Siemens EDA (formerly Mentor Graphics). Among these three, Synopsys and Cadence are the most dominant market leaders, each controlling close to 40% of the market. This dominance is reinforced by near 100% retention rates and an 80%-85% recurring revenue model, reflecting the critical role that EDA tools play in chip development and the loyalty the top EDA firms have from their clients. These are essentially the strongest retention rates a software company can have, implying customer lifetimes well over 20 years. The primary reason a client stops using the services of Cadence is that it goes out of business. We believe the top EDA firms’ market dominance (essentially a stable oligopoly), evidence of extremely high switching costs (retention rates near 100%), and position as indispensable tools for processes where firms cannot afford failure (a bad chip design could put a younger firm out of business, or at the very least have a severely negative impact on a more mature firm) all support a wide moat based on switching costs and intangible assets.

Switching costs represent the primary moat source for Cadence. Chip design is an intricate process involving thousands of hours dedicated to fine-tuning workflows, integrating tools, and aligning processes. Cadence’s tools are tightly integrated with its clients' workflows, and transitioning to an alternative could lead to significant implementation risks, such as a drop in productivity and complex data migration challenges. In a rapidly moving technological industry, we believe that the costs of switching vendors far outweigh the small technological benefit the client would get from switching vendors. In addition, engineers invest considerable time in becoming proficient with specific tools, and changing vendors would mean more training, additional ramp-up time, disruption to ongoing projects, and delayed time to market.

Further, firms cannot afford mistakes in the design process, as it puts future revenue and the future of the business at risk. Minor design flaws can result in significant disruptions and exacerbated losses down the line, making it necessary for these tools to be reliable and trustworthy. Clients would find it impractical to change providers without jeopardizing project deadlines and the quality of the final product.

In support of intangible assets, we point to the overall reputation of top EDA firms, like Cadence. When a company is considering which EDA tools to use, there are only three reliable choices. We find it highly unlikely that a company would entrust its design processes to an unknown, unproven upstart. It is simply not worth the risk. This hesitancy toward new vendors, along with gradual industry consolidation over time, has led to minimal investment in new alternatives. This is in stark contrast to other parts of the semiconductor industry, such as chip design (the clients of EDA firms), where new capital and companies jumped at the opportunity presented by new waves of demand, such as designing AI-optimized chips.

The barriers to entry in the EDA business are substantial, involving massive upfront capital, deep technical knowledge, and cumulative industry expertise that cannot be easily replicated. Synopsys' and Cadence’s continuous acquisitions to build their software portfolio have led to market consolidation and a robust pace of innovation, which has put them leagues ahead of smaller competitors. We observe that if either lacks a feature or process, they have both the resources and technical know-how to catch up and maintain their competitive advantage. A major advantage that Cadence has over smaller competitors is its ability to offer end-to-end design flow—a full suite of tools needed to design a chip from start to finish—thus reducing the appeal of small competitors that have fragmented solutions.

Another industry feature that gives us confidence in our wide moat rating is the fact that EDA is not cyclical. During uncertain times and economic downcycles in the semiconductor market, firms can cut budgets, but one area where firms are loath to cut spending is their core R&D budget. Innovation is essential to maintaining a competitive edge in the semiconductor industry. If a company starts cutting its EDA budget significantly, it means it is essentially firing its engineers, which also likely means it is facing serious financial difficulty, likely even on the verge of going out of business.

Over the long term, we expect demand for EDA to only increase. As chips get smaller and more complex, and therefore more expensive to design, customers are willing to pay a premium for the latest tools that Cadence has to offer. Increasing complexity and increasing difficulties in chasing improved performance also increase the value of the solutions EDA firms offer. To meet this demand, Cadence has some of the highest R&D spending in the software space to ensure it remains at the cutting edge.

Another factor increasing our confidence in the long-term demand outlook is the increasing diversity of Cadence’s client base, extending beyond semiconductor-focused companies to include what are called system-level customers. Fifteen years ago, nearly all of Cadence’s clients were semiconductor companies; today, close to 45% are companies that produce the end products rather than the chips alone. For example, an automotive firm that has to design a car that uses many different semiconductor components will design the car with the overall system, including the semiconductors, in mind. They can use EDA software for this system-level design process. With products only becoming more complex and more electronic over time, we expect this demand trend will remain in place for decades to come. With the EDA industry consolidated around a few players, Cadence’s brand and unique technological advantage position it to capture pricing uplifts as well as volume increase from a broadening customer base, along with industry-leading margins, all of which ultimately flow into better returns on capital.

Cadence’s other primary business is its IP segment, which is becoming an increasingly important part of the company. As of fiscal 2025, IP growth had outpaced Cadence's core EDA growth, representing around 14% of the company’s revenue.

Cadence’s IP solutions can be compared with Lego building blocks. When designing a chip, it is helpful for engineers to use prebuilt blocks for parts of the chip that are common and undifferentiated, allowing designers to focus on customizing the unique, differentiated aspects of their creation. For example, part of the chip may have to deal with interacting with a USB connection. Building out this part of the chip each time a new chip is designed is not an efficient use of engineering resources. Companies like Cadence will prebuild this part of the chip and then sell that part of the design as IP. This allows clients to purchase these modular, proven, undifferentiated parts of a chip’s design for a fraction of what it would cost them to develop them in-house with their own engineering team. Cadence can sell this IP at an attractive price because it is able to depreciate its cost of development over a large base of clients. Finally, speed to market is also important, and having an established IP business at the center of the design ecosystem allows Cadence to compete for the fastest time to market with verified IP blocks.

We believe this business benefits from an intangible asset moat source, driven by the breadth of its IP library and Cadence’s unique position at the center of the semiconductor design workflow.

Cadence has steadily expanded its IP library (through acquisitions and internal development), building a robust repository of standardized and custom solutions that differentiate it from smaller competitors. The extensive library, combined with continuous innovation in sectors like automotive and AI-focused applications, keeps Cadence at the forefront of client demands and the pace of technological developments. Nearly half of Cadence’s workforce is dedicated to building and maintaining this library, a testament to the company’s commitment to quality and breadth. This breadth makes Cadence a natural first place to look for IP solutions as a designer.

The second key ingredient to Cadence's moat in IP is its unique position at the center of semiconductor design. In theory, anyone could create many of these IP blocks. However, it is fundamentally more efficient at the industry level to have one or two firms do it, and then sell those standardized design blocks to the rest of the industry, as opposed to having each design firm rebuild what are functionally the same design blocks for each and every project. This allows a firm like Cadence, which is naturally already at the center of the design process for the industry with its software, to have a natural distribution advantage. This further allows Cadence to depreciate the cost of development for each IP block across a wider base of customers. We hesitate to award the firm a cost advantage for this because, in theory, it does not cost Cadence any less to develop these IP blocks, but the ability to sell them across a wider base of customers does allow Cadence to sell them at a price which would be unpractical for a firm with too few customers, and at a price which is cheaper than what it would cost any of its customers to develop these IP blocks in house.

Cadence grew its market share in the IP business to 6% in 2024, growing faster than the overall IP industry. Arm remains the dominant player overall, with 44% share, while Synopsys is number two with 24%. The consistent quality of Cadence’s IP fosters long-term partnerships and deepens client loyalty. Furthermore, client-specific customized solutions developed with Cadence make switching to other providers challenging, as it risks compromising design integrity and reliability.

Lastly, the type of IP portfolio matters. Arm has the most dominant market share, but it is focused only on processor IP. Synopsys and Cadence concentrate on nonprocessor IP, which is growing fast; we forecast it will soon account for more than half of the IP market. Given its fast-growing IP margin profile, we believe Cadence’s IP business will be a significant contributor to strong and improving ROICs and future growth.

With limited competition and products that are essential to the semiconductor industry, which generates more than $500 billion in annual revenue, Cadence’s moat is wide and durable.

Secular tailwinds such as physical AI and AI tools will drive new chip design starts, raise R&D intensity, and open new markets for semiconductor-focused EDA tools, leading to excess revenue growth for 10 or more years.

Margins should only improve, supported by improving IP margins, greater AI-driven wallet share, and internal AL/machine learning optimizations that enhance operational efficiency.

The market is already giving Cadence full credit for the upcoming growth cycle. Any hiccups along the way will lead to declines in valuation.

Geopolitical tensions with China and China’s focus on supporting domestic companies could affect Cadence’s ability to do business in the region and lead to slower revenue growth.

Cadence has less share with the overall IP market, and currently the analog space is not growing as fast as the digital semiconductor industry, given the AI demand boom, so Cadence could underperform peers.