Written in Silicon
In this edition of the Smart Investor newsletter, we spotlight the design layer behind modern chips. But first, let’s review the latest Smart Portfolio developments.
1
1
Portfolio News and Updates
❖❖ America’s surging electricity demand, driven largely by AI data centers, is straining the power grid and pushing hyperscalers and data-center operators to look for creative solutions. Thus, Alphabet’s (GOOGL) Google has recently struck a landmark deal with Xcel Energy and Form Energy to deploy the world’s largest-capacity battery near Google’s forthcoming Minnesota data center. The system features a 300-megawatt / 30-gigawatt-hour iron-air battery capable of storing enough energy to power 3,000 average U.S. homes for a year, with up to 100 hours of discharge. Google will fully fund the project through a special Clean Energy Accelerator Charge (CEAC) tariff, ensuring no rate increases for existing Xcel customers. U.S. battery storage installations grew by roughly 30% last year, driven by falling costs and rising demand for resilience from data centers, factories, hospitals, schools, and commercial buildings.
Google, along with other hyperscalers, is rapidly expanding its power solutions. Facing massive grid bottlenecks and long interconnection delays, they are reducing dependence on the traditional shared grid through a mix of strategies: dedicated utility-scale resources (like GOOGL’s Minnesota deal), plus accelerated behind-the-meter and on-site options such as gas turbines, fuel cells, small modular reactors (SMRs), geothermal, and batteries. Many new large campuses now incorporate high levels of on-site control, enabling more independent operation during grid stress.
❖ In other news, Google DeepMind – Alphabet’s premier AI research lab – has launched Gemma 4, a new family of open AI models designed to run directly on devices. The models deliver advanced agent-like capabilities such as multi-step planning, autonomous actions, and offline processing, reducing reliance on cloud inference while improving privacy, lowering latency, and cutting costs. This addresses growing demand for fast, private AI on phones, laptops, and edge devices.
Gemma 4 is Google’s most capable open model family to date. It supports 140+ languages, excels at reasoning and agentic workflows, and is released under the highly developer-friendly Apache 2.0 license – a significant upgrade from previous versions. The release represents a major step in the open-source AI race and helps keep developers within Google’s ecosystem.
Google is also strengthening its partnership with NVIDIA (NVDA). The two giants collaborated to optimize Gemma 4 for NVIDIA GPUs, enabling high-performance reasoning, coding, and multimodal processing across RTX PCs, DGX Spark systems, Jetson edge devices, and data center GPUs – all while maintaining strong offline efficiency.
1
❖❖ NVIDIA (NVDA) is seeing a massive surge in demand for its H100 GPUs, driving rental prices up by nearly 40% since last October. Existing GPU capacity is effectively sold out across the board, with many companies that secured access earlier choosing to hold onto their allocations even as prices rise. As a result, securing available GPUs has become increasingly difficult, forcing some buyers to compete for much higher-priced spot instances on platforms like Amazon’s (AMZN) AWS.
The supply tightness extends beyond previous-generation chips to Nvidia’s new Blackwell GPUs, where lead times have now stretched into mid-2026 – signaling severe shortages amid soaring demand. This has created a counter-intuitive situation: instead of getting cheaper as newer, more efficient chips arrive, H100 prices have continued to rise strongly.
This dynamic shows that AI compute demand is still significantly outstripping supply. It counters near-term “AI bubble” fears by highlighting real, growing usage in inference, media generation, AI agents, and other workloads – not just large model training. The situation also strongly supports NVDA’s investment case, underscoring the company’s substantial pricing power and high revenue visibility for at least the next several quarters.
1
❖❖ IBM (IBM) is collaborating with Arm Holdings to develop new dual-architecture hardware for AI and data-intensive enterprise workloads. The partnership will combine IBM’s expertise in enterprise-grade systems design, reliability, and security with Arm’s power-efficient architecture and broad software ecosystem. This will give customers greater flexibility, portability, and scalability when deploying and running AI workloads in mission-critical environments. The project includes expanding virtualization technologies so Arm-based software can run efficiently on IBM’s enterprise platforms (like IBM Z), helping bridge existing mainframe environments with modern Arm-powered workloads. The collaboration reflects growing enterprise demand for infrastructure that balances performance, efficiency, and security as AI adoption accelerates.
❖ In other news, IBM received FedRAMP authorization for 11 of its AI and automation software solutions, representing a four-fold increase in its FedRAMP portfolio in just one year.
FedRAMP (Federal Risk and Authorization Management Program) is the mandatory U.S. government security standard for cloud products – essentially a critical stamp of approval required for federal agencies to use software at scale.
Several of the newly authorized solutions come from IBM’s flagship watsonx AI portfolio, along with its automation tools. This makes IBM’s enterprise AI immediately mission-ready for U.S. federal agencies. The approvals expand IBM’s access to a multi-billion-dollar federal AI market and further strengthen its position as a trusted supplier to the U.S. government — one of its most stable and high-margin customer bases.
All 11 solutions are deployed exclusively on Amazon’s (AMZN) AWS GovCloud as part of IBM’s strategic collaboration agreement with Amazon.
1
❖❖ According to media reports, Amazon (AMZN) is in advanced talks to acquire satellite communications company Globalstar in a potential $9 billion deal. The acquisition aims to accelerate and strengthen Amazon’s low-Earth-orbit satellite internet business, Amazon LEO (formerly Project Kuiper), as it pushes toward commercial service.
Amazon has already launched more than 180 satellites and is racing to meet a mid-2026 FCC deployment deadline for roughly 1,600 satellites. A deal would give Amazon immediate access to Globalstar’s operational satellites, spectrum rights, and infrastructure – helping it catch up to Elon Musk’s SpaceX Starlink, which maintains a substantial lead in both satellites deployed and active users.
1
❖❖ CrowdStrike (CRWD) announced a $500 million expansion of its buyback program, bringing total authorization to $1.5 billion. The company disclosed that it had already repurchased $150.6 million worth of shares following its record Q4 FY2026 earnings, reported on March 3. This move signals strong management confidence that the stock’s year-to-date decline is disconnected from the company’s robust fundamentals and AI-fueled growth momentum.
1
❖❖ Broadcom (AVGO) has signed a new agreement with Alphabet’s (GOOGL) Google to develop and produce future generations of Google’s proprietary AI chips – Tensor Processing Units (TPUs) – along with networking and other components for Google’s next-generation AI racks through 2031. This partnership positions AVGO as GOOGL’s key design partner for cutting-edge AI silicon, further cementing its leadership in custom-chip development and placing it at the heart of training and deploying large models.
In parallel, Broadcom, Google, and the Claude maker Anthropic announced an expansion of their existing collaboration. Under the add-on deal, Broadcom will facilitate Anthropic’s access to approximately 3.5 gigawatts of next-generation Google TPU-based AI compute capacity, beginning in 2027. The Amazon-backed AI startup stated that this infrastructure expansion will power its frontier models, helping meet accelerating demand that has driven its revenue run-rate to over $30 billion, up from about $9 billion at the end of 2025.
Zooming out, the Broadcom-Google-Anthropic deals illustrate a profound shift in the AI industry. Competition is increasingly moving beyond “who develops a better AI model” toward “who can secure compute capacity and power at industrial scale.” AI is no longer just a technology – it is rapidly becoming an infrastructure business, characterized by higher barriers to entry, massive capex requirements, and growing dependence on energy. Differentiation is shifting from models and algorithms toward control of physical resources: data centers, electricity, and communication networks.
Zooming back in, the agreement underscores Google’s transformation into a full-stack owner of the AI value chain – from hardware and networking to cloud and models. This strengthens its position versus Azure and AWS – though both have their own advantages – as well as versus NVIDIA (NVDA), although its GPU dominance remains unquestioned. Meanwhile, Broadcom’s evolution from a high-tech component supplier to a strategic AI infrastructure partner receives another significant acceleration.
1
1
Portfolio Stocks Under Review
❖ We are keeping Oracle (ORCL) under review despite its improved outlook, expanding AI portfolio, and analyst support.
The latest earnings report confirmed that ORCL is already monetizing AI infrastructure demand, with its massive backlog starting to convert into real revenue. Just as important, the company clarified that much of this expansion is funded through customer-backed infrastructure models, significantly reducing its capital burden. While margins remain pressured by the pace of data center construction, this appears to be a timing issue rather than a structural concern. Much of the capacity under development is already contracted at attractive terms, pointing to improved returns as projects come online. The latest results also ease prior concerns around debt and negative free cash flow, as the customer-funded model shifts a meaningful portion of investment away from Oracle’s balance sheet.
Oracle’s cloud transformation is accelerating, driven primarily by AI demand. OCI remains the key engine, supported by demand that continues to exceed available supply, while the AI infrastructure business is already profitable, with further upside from higher-margin services and database offerings. Oracle is also deepening its AI stack, launching agentic capabilities across applications and database layers. At the same time, a series of federal contract wins reinforces Oracle’s $553 billion backlog narrative and OCI’s growing enterprise footprint. The expanded partnership with NVIDIA, alongside heavy investment in data center capacity, further positions Oracle as a full-stack AI infrastructure provider across compute, data, and applications.
Meanwhile, positive news is emerging on the financing and operational fronts. According to media reports, Related Digital is finalizing $16 billion of financing for an Oracle data center in Michigan, intended to power OpenAI workloads under the broader Stargate buildout. The project would give ORCL additional access to scarce AI infrastructure capacity, supporting future OCI revenue and OpenAI-related demand if utilization ramps as expected. While the scale of financing has drawn attention, it also reinforces Oracle’s positioning in one of the fastest-growing segments of the cloud market. The deal is structured so the debt sits in a separate entity – not directly on ORCL’s balance sheet – although Oracle retains lease obligations tied to the facility.
Street sentiment is turning more constructive. BofA reinstated coverage with a Buy rating and a $200 price target, citing upside tied to AI infrastructure demand. Bernstein reiterated its Buy rating with a $319 price target, noting ORCL’s story is stronger than expected – with lower financing needs, a clearer path to positive cash flow, and positioning as a key AI build-out beneficiary. JPMorgan highlighted ORCL’s resilient, sticky, and largely recurring revenue base, calling the risk-reward attractive, while Wedbush has recently reiterated its bullish call on defense AI, naming Oracle as one of the key beneficiaries of the massive spending thanks to its position as “picks and shovels” AI infra play.
Our conviction in Oracle as a core AI infrastructure player hasn’t changed, with clear evidence that monetization is underway and its role across the AI value chain is expanding. Still, near-term uncertainty remains elevated, as investors navigate a heavy news flow against a turbulent macro backdrop tied to the Iran war and broader economic implications. ORCL appears exposed to potential disruptions, alongside hyperscalers like Amazon and Google, amid threats to Middle East data centers.
Against this backdrop, even positive developments like ORCL’s new CFO hire, Hilary Maxson – a seasoned executive well suited for Oracle’s capital-intensive AI expansion, according to Barclays – are harder to digest. Investors are also weighing the company’s restructuring plan involving mass layoffs, which align with broader AI-driven cost-cutting trends and support cash preservation. Overall, with the equity picture still largely obscured by macro and geopolitical forces, we prefer to keep the stock under our magnifying glass for now.
1
1
Portfolio Earnings and Dividend Calendar
❖ The Q1 2026 earnings season will open this week, with JPMorgan Chase (JPM) and Citigroup (C) set to release their results on April 14, while Morgan Stanley (MS) is scheduled to report on April 15.
❖ The ex-dividend date for Oracle (ORCL) is April 9, while for General Dynamics (GD) it is April 10.
w
1
New Buy: Synopsys (SNPS)
Synopsys, Inc. operates at the center of the global semiconductor design process, providing the software and intellectual property used to develop and validate advanced chips. Its tools are embedded across the workflows of leading chip designers, foundries, and system companies, supporting everything from early-stage architecture to final verification and manufacturing readiness. The company’s portfolio spans electronic design automation, semiconductor IP, and software security, linking hardware development with increasingly complex software requirements. As chips become more intricate and development cycles more demanding, Synopsys plays a critical role in enabling the industry to move from concept to production – serving as a key infrastructure layer within the broader semiconductor ecosystem.
1
Logical Synthesis
Synopsys was founded in 1986 to commercialize logic synthesis technology1 at a point when chip design was becoming too complex to manage manually. That early focus on automation shaped the company’s trajectory, as it expanded from synthesis into a broader suite of design, verification, and semiconductor IP capabilities. Over time, Synopsys became embedded across the semiconductor design process, evolving into one of the core software providers supporting the global chip industry.
A more defining shift has taken place over the past five years. As chip architectures became more complex and system-level constraints more binding, Synopsys moved to strengthen its role in the later stages of design and signoff, where performance, reliability, and manufacturability converge. In 2023, the company acquired Silicon Frontline Technology, adding advanced electrical verification, electrothermal analysis,2 and ESD-focused tools3 that deepened its presence in signoff and system-level validation.
At the same time, SNPS expanded its use of AI within its own platform. Tools such as DSO.ai and the broader Synopsys.ai framework have been integrated into customer workflows to automate parts of the design process, reduce iteration cycles, and improve optimization across increasingly large design spaces. These capabilities have become more relevant as chipmakers push into advanced nodes, heterogeneous integration, and AI-specific architectures.
The company also invested in emerging areas tied to long-term industry shifts. In 2022, Synopsys and Juniper Networks formed a joint venture called OpenLight, focused on silicon photonics. Synopsys retains a stake in the business, which develops integrated photonics solutions4 aimed at improving data transfer speeds and energy efficiency in data centers and high-performance computing environments.
Portfolio decisions have been equally deliberate. In 2024, Synopsys agreed to sell its Software Integrity business, marking an exit from application security testing and sharpening its focus on semiconductor design and system engineering. Around the same period, it also divested its Optical Solutions Group to Keysight, exiting a hardware-oriented business that sat outside its core software and IP model. These moves streamlined the company around higher-value, design-centric capabilities and reduced exposure to adjacent segments with limited strategic overlap.
The largest step came in 2025 with the completion of the Ansys acquisition. The combination brings together electronic design automation, semiconductor IP, and multiphysics simulation,5 extending Synopsys’ reach into system-level engineering where thermal, mechanical, and electrical constraints increasingly shape design outcomes.
More recently, Elliott Investment Management disclosed a multibillion-dollar stake, pointing to opportunities to improve growth and profitability. The timing is notable – it comes as Synopsys has expanded its scope and tightened its focus, with a broader platform and clearer strategic direction already in place, supporting Elliott’s view that SNPS’s unique position in the chip industry can be better monetized.
—
1 – Logic synthesis technology – software that converts high-level chip designs into gate-level circuits ready for manufacturing.
2 – Electrothermal analysis – simulation of how electrical activity generates heat and how that heat affects chip performance and reliability.
3 – ESD-focused tools – tools that design and verify protection against electrostatic discharge, which can damage semiconductor devices.
4 – Integrated photonics solutions – technology that uses light instead of electricity to transmit data within and between chips.
5 – Multiphysics simulation – modeling that combines electrical, thermal, mechanical, and other physical effects within a single system.
—
1
Complexity Rules
Synopsys sits at the design and engineering core of the semiconductor industry, providing the software and models that define how chips and increasingly entire electronic systems are built. Its platform combines electronic design automation, semiconductor IP, and simulation into a unified stack that extends from transistor-level design to full system behavior. With the addition of Ansys, that scope now includes multiphysics simulation, allowing electrical, thermal, and mechanical interactions to be modeled together earlier in the development cycle.
The business is anchored in two primary segments. Design Automation – the largest contributor and margin driver – provides the tools used to design and verify chips at advanced nodes. Design IP contributes a smaller share of revenue but carries disproportionate strategic weight, supplying high-speed interfaces and connectivity solutions that sit at the center of AI and data center architectures. As bandwidth and energy efficiency become constraints, demand for advanced interconnect IP is increasing, particularly in chiplet-based and multi-die systems. Ansys adds a third dimension, bringing simulation into the same workflow and enabling Synopsys to participate earlier in product design across automotive, aerospace, and industrial systems.
Synopsys operates in a triopoly alongside Cadence Design Systems and Siemens, where decades of accumulated models, workflow integration, and foundry alignment create high switching costs. Its tools are embedded across the design process, often serving as the reference layer for verification and sign-off, which ties customers into long development cycles and recurring engagement.
Demand is shaped by rising chip and system complexity. Advanced nodes, multi-die architectures, and heterogeneous integration require deeper verification and tighter coordination across domains. At the same time, development is shifting toward simulation-first workflows, where digital models increasingly replace or precede physical prototypes. This expands the role of multiphysics simulation and increases the amount of computation required before a design reaches fabrication.
AI reinforces this shift from both directions. It increases the complexity of the systems being built, while also improving how those systems are designed. SNPS is embedding AI across its tools to accelerate workflows and automate parts of the design process, while hardware-assisted verification systems and advanced tools are capturing a growing share of high-performance computing workloads.
The company’s position within the AI ecosystem has become more explicit. Its expanded partnership with NVIDIA integrates accelerated computing directly into design and simulation workflows, enabling larger-scale verification and digital modeling. Collaboration with Arm extends that reach into next-generation CPU architectures, placing Synopsys within the standard toolchain for AI and data center compute platforms. Together, these relationships place the company as a central layer connecting compute infrastructure with design and validation.
Growth is increasingly tied to hyperscalers and AI-driven customers, where custom silicon programs are expanding rapidly. That shift is increasing demand for advanced IP, verification, and simulation, while broadening the customer base beyond traditional semiconductor firms. Ansys strengthens this trajectory by enabling cross-selling and deeper integration into these workflows, particularly in areas where simulation intensity is rising.
Near-term conditions remain uneven. AI-related design activity is strong, while consumer, automotive, and industrial markets are softer, creating a two-speed demand environment. Parts of the IP business are in transition as the portfolio shifts toward higher-value applications, and monetization is evolving toward more flexible, usage-based models that will take time to scale.
This context also helps explain the early-2026 “SaaSpocalypse” selloff, where SNPS was grouped with enterprise software despite operating under a fundamentally different model. Its revenue is tied to design activity, complexity, and compute intensity, not seat-based licensing, and its tools are embedded in mission-critical engineering workflows where substitution risk is low.
Elliott’s recent stake builds on that foundation, highlighting Synopsys as an indispensable but under-monetized platform. The gap is in how effectively that position is translated into pricing, margins, and share of customer spend. Industry growth has outpaced EDA monetization for years, and the combination of AI-driven demand, system-level expansion, and deeper customer integration creates a clear path to close that gap.
The next phase centers on execution. The platform is broader, ecosystem integration is deeper, and the Ansys combination creates a business that scales directly with engineering complexity and compute demand. What remains is translating that position into stronger monetization, higher margins, and more consistent growth – as the demand environment that justifies all of it continues to expand.
1
Current Conversion
Synopsys entered fiscal 2026 with a financial profile that reflects both scale expansion and transitional friction. First-quarter revenue reached $2.41 billion, up sharply year-over-year following the inclusion of Ansys, while non-GAAP EPS of $3.77 exceeded expectations, extending a pattern of consistent earnings outperformance with only a single miss since 2021. Profitability remained elevated, with a non-GAAP operating margin of roughly 42%, underscoring the strength of the core business even as integration costs and mix shifts move through the model.
The addition of Ansys has reshaped the company’s financial structure. The simulation business contributed approximately $886 million in Q1 and is expected to generate around $2.9 billion for the full year, bringing total revenue guidance to $9.56-9.66 billion. This expands SNPS beyond its traditional semiconductor base while introducing a new layer of growth tied to system-level engineering. At the same time, the underlying mix has become more differentiated. Design Automation, which now accounts for more than 80% of revenue, continues to operate at margins in the mid-to-high 40% range, closely aligned with Cadence. Design IP, representing roughly 15-20% of revenue, remains the outlier, with margins closer to the mid-teens following a 6% year-over-year revenue decline and continued investment in next-generation IP.
The divergence between these segments is central to the current setup. The core business is already operating near peer-level profitability, while the IP segment is in a transition phase, with revenue timing weighted toward the back half and margins compressed by ongoing development spend. This creates near-term volatility, but also a clear path for margin expansion as volumes recover and monetization evolves, particularly through royalty-style models tied to AI and data center demand.
Despite these moving parts, cash generation remains a consistent strength. Operating cash flow exceeded $850 million in the quarter, with free cash flow of roughly $822 million, and full-year expectations around $1.9 billion. That cash flow is already being reflected in the balance sheet. Following the Ansys acquisition, Synopsys carried roughly $10 billion of debt, but repaid a $4.3 billion term loan within months, signaling both financial discipline and deleveraging capacity. Additional flexibility came from NVIDIA’s $2 billion equity investment, reinforcing the company’s position within the AI ecosystem while strengthening its capital base.
Visibility remains unusually high for a software business. Backlog stands at approximately $11.3 billion, supported by a large deferred revenue base, with the majority expected to convert within three years. The revenue model – a combination of time-based licenses, upfront agreements, and maintenance – continues to provide a stable and predictable foundation even as the company expands into new domains.
The near-term challenge is less about demand than about clarity. Fiscal 2026 is widely viewed as a transition year, shaped by the integration of Ansys, mixed segment performance, and evolving monetization models. Design Automation remains strong, supported by AI-driven design activity, while Design IP and non-AI end markets introduce variability. At the same time, reported earnings are affected by acquisition-related amortization and tax changes, creating a gap between GAAP optics and underlying profitability.
That setup brings the focus back to monetization. SNPS operates at the center of a semiconductor ecosystem that continues to expand, yet EDA revenue still represents only a small fraction of total industry value. Elliott’s involvement reflects a push to close that gap – not by changing the business, but by improving pricing, margins, and share of customer spend as demand continues to scale.
From a financial standpoint, the structure is already in place. The core business delivers high margins and strong cash flow, the balance sheet is stabilizing, and the Ansys combination introduces additional growth and cross-selling opportunities, including a $400 million revenue synergy target. The next phase depends on execution across these levers – particularly in lifting IP profitability, integrating simulation more deeply into the platform, and translating rising demand into sustained earnings expansion.
1
Forward Bias
Synopsys operates within a narrow group of U.S.-listed companies tied to the design and engineering layer of the semiconductor and broader electronics ecosystem – where scale, technical depth, and workflow integration define competitive positioning. Cadence Design Systems provides the closest comparison, sharing the same customer base, product scope, and margin structure, and serving as the primary benchmark for execution and monetization. Keysight Technologies – a core Smart Portfolio holding – sits adjacent in the design-to-validation chain, combining software, simulation, and test capabilities across high-performance and RF systems; its more cyclical, hardware-exposed model offers a useful contrast to Synopsys’ higher-margin, software-driven economics while reinforcing the shared exposure to AI infrastructure and advanced chip development. Autodesk adds a broader engineering software perspective, with a similarly entrenched position in mission-critical design workflows and a monetization model built on pricing power and long-duration customer relationships. Together, these peers place Synopsys between direct competition in EDA, adjacent exposure to semiconductor systems, and the pricing dynamics of high-end design software.
Stock performance across this group has varied widely over the past year, reflecting differences in end-market exposure and execution. Keysight has been the clear outperformer, up more than 125%, driven by direct leverage to AI infrastructure buildout and strong demand across high-speed and test-related applications. At the other end, Autodesk is down roughly 2%, with performance tied more to enterprise software dynamics and pricing execution than to the AI-driven semiconductor cycle. Cadence has gained over 19%, despite being pulled into the same early-2026 “SaaSpocalypse” selloff as SNPS. The recovery has been supported by consistently higher margins and a steady pattern of beat-and-raise quarters, reinforcing confidence in execution.
Synopsys, by contrast, is up about 4%, a more muted outcome that reflects a combination of factors – the same sector-wide selloff, a transition year following the Ansys integration, and ongoing questions around IP performance and monetization. The gap versus CDNS is less about demand and more about timing, with SNPS still in the process of converting its expanded platform and AI-driven exposure into earnings momentum. With NVIDIA’s validation and Elliott’s involvement, that conversion process may accelerate, creating room for performance to catch up with positioning.
The growth outlook reinforces this dynamic, building on the past year’s trajectory. Revenue expanded at the fastest rate in the peer group and is expected to remain the growth leader over the next 12 months, while previously lagging EBITDA growth is projected to accelerate, closing the gap and potentially surpassing Cadence. Meanwhile, Synopsys’ valuation does not yet reflect this improving growth and profitability profile. Its trailing and forward non-GAAP P/E, EV/EBITDA, EV/Sales, price-to-cash-flow, and forward PEG ratios remain below all peers, with the exception of Autodesk, whose stock performance has lagged over the past year. Wall Street largely views this gap as an opportunity, with the consensus price target for SNPS implying a potential upside of over 37% from current levels.
Along with the potential for stock price appreciation, Synopsys also returns capital to shareholders via a discretionary buyback program. In February 2026, the board replenished the existing authorization with up to $2.0 billion in new buyback capacity. This came after the company had largely repaid the additional debt taken on for the Ansys acquisition, freeing up capacity for capital returns. In early March 2026, the company executed the first tranche by entering into a $250 million ASR agreement with Bank of Nova Scotia. Analysts largely expect Elliott’s involvement could lead to clearer targets or more aggressive use of the authorization to support shareholder returns.
Synopsys now combines leading growth in the group with a valuation that still reflects hesitation – a gap that may narrow as execution catches up with its expanded platform and positioning in the AI-driven design cycle.
1
Investing Takeaway
Synopsys is moving into a phase where its expanded platform and market position are starting to align with how the business is reflected in its results. Its role at the center of chip design and system-level engineering places it directly in the path of rising complexity driven by AI, custom silicon, and increasingly integrated products. The combination with Ansys broadens that reach beyond semiconductors, while partnerships with NVIDIA and Arm deepen its position within the core design ecosystem. Near-term execution remains the key variable, particularly in translating platform breadth into consistent earnings growth. If that conversion continues to improve, Synopsys has the characteristics of a durable compounder – with a business that is already essential, and a valuation that still leaves room for that importance to be more fully recognized.
1
1
New Sell: Vertex Pharmaceuticals (VRTX)
We are exiting Vertex Pharmaceuticals – not because the business has weakened, but because the stock’s current setup no longer fits our portfolio in an environment where visibility has declined and uncertainty has increased.
VRTX remains a high-quality biotechnology company. Its cystic fibrosis franchise continues to deliver strong, durable cash flows, and recent FDA label expansions further reinforce its leadership in the space. At the same time, its pipeline across kidney disease, pain, and gene-editing therapies provides meaningful long-term optionality. The fundamental story remains intact.
The issue is not VRTX’s business, but visibility and market dynamics. Over the past several months, the stock has underperformed the S&P 500 and increasingly traded in line with broader healthcare indices rather than on company-specific developments. Even positive updates have not translated into sustained momentum, reflecting a market that is currently driven more by macro and sentiment than by fundamentals.
In this environment, where geopolitical developments and macro pressures are shaping market behavior, assessing stock-specific outlooks in complex sectors like healthcare has become more difficult. VRTX requires ongoing evaluation of clinical data, regulatory developments, and policy dynamics – a level of sector-specific analysis that is not justified given its role as a non-core holding in the portfolio.
This is a portfolio discipline decision as much as a stock-specific one. With limited near-term visibility and no clear catalyst to shift sentiment, we prefer to redeploy capital into higher-conviction opportunities where we have stronger analytical edge.
1
1
Smart Investor’s Winners Club
The Winners Club represents stocks from the Smart Investor Portfolio that have risen at least 30% since their purchase dates.
A positive, if volatile, week ended with the Club gaining two members, now counting 22 stocks: AVGO, GE, TSM, EME, HWM, ANET, APH, VRT, PH, IBKR, MTZ, GOOGL, ORCL, KEYS, RTX, BK, ATI, ASX, CSCO, JBL, MS, and CRWD.
The first runner-up is now JPM with a 21.58% gain since purchase. Will it return to the ranks, or will another stock outrun it to the finish line?
1
1
|
