May mARKet Update Webinar Summary

Moderated by Portfolio Manager Dan White, CFA, ARK's May Market Update webinar opened with a macroeconomic overview by Cathie Wood, ARK's Chief Executive Officer and Chief Investment Officer. The broader discussion then turned to AI infrastructure economics and cloud computing, the analog semiconductor landscape, Tesla's autonomous vehicle progress, convergence opportunities in multiomics and AI, and digital assets. View the full webinar [HERE]  

Can you explain the investment case for CoreWeave (CRWV)? What kind of value does the company provide to its clients? What is your opinion of the company’s balance sheet strength?

Frank Downing, Director of Research, AI and Cloud, explained that CoreWeave is an AI-native cloud service provider specializing in building large-scale graphics processing unit (GPU)-based training and inference clusters for leading AI companies. The company counts nine of the top ten AI labs—including OpenAI, Anthropic, Meta, Microsoft, and Google—as customers and carries a backlog exceeding $100 billion. Semi Analysis, an industry research firm, rates CoreWeave as the only provider in its platinum tier for GPU cloud services, a reflection of the quality and reliability of its infrastructure. 

On balance sheet strength, Frank noted that CoreWeave structures its contracts so that customers pay roughly 20% upfront, with CoreWeave financing the remaining 80% through debt that typically is repaid within two and a half years. Concerns about accelerated GPU depreciation have proven overstated. Sustained AI demand is increasing prices for three- and four-year-old GPUs, allowing CoreWeave to re-rent aging infrastructure at higher rates as contracts roll off. The company's weighted average cost of debt declined by 600 basis points (6%) between 2023 and 2025, and its first investment-grade debt offering earlier this year lowered its cost of debt by an additional 80 (0.8%) basis points, reflecting improving financial quality as CoreWeave scales. 

What is your opinion about analog semiconductor companies? Are there any companies that you are looking into or have researched?

Jozef Soja, Research Analyst, AI & Cloud, noted that ARK's historical interest in analog semiconductors was based on their role in automotive electrification and software-defined vehicles—an exposure that has been reduced as the full autonomy opportunity has grown to overshadow that earlier growth vector. More recently, their data center exposure has boosted the attention to analog chip companies, specifically to the analog chips and microcontrollers that manage power systems, cooling, diagnostics, and other infrastructure functions that keep GPU and central processing unit (CPU) clusters running. 

Jozef highlighted NXP Semiconductors as a company that ARK has researched and holds in certain foreign portfolios. NXP generated ~$200 million in data center revenue in 2025 and expects that figure to grow to $500 million in 2026, on a total revenue base of roughly $12 billion last year and an estimated $14 billion this year. He drew a parallel to the re-rating CPUs have received as the market recognized that agentic AI workloads drive CPU demand alongside GPU demand. While ARK may revisit analog semiconductors through a data center lens, the team remains more focused on GPUs, CPUs, application-specific integrated circuits (ASICs), and networking and storage products, which ARK's research suggests could grow to a cumulative $1.4 trillion market by 2030. 

We're through the bulk of earnings, and AI continues to drive the narrative. Walk us through what you're seeing, the capital expenditure (CapEx) going in on one side, and the productivity coming out on the other.

Frank Downing described a market narrative that has been validated: the concern throughout last year that hyperscalers¹ Amazon, Google, Microsoft, and Meta are investing hundreds of billions into AI without visible returns has given way to clear evidence of both revenue growth and productivity gains. Projected 2026 capital expenditure across all four companies now exceeds $700 billion, and the results in their businesses demonstrate that the investment is paying off. Anthropic, for example, grew from ~$10 billion in annualized revenue at the end of 2024 to a reported $44 billion run rate, roughly a 4x increase in approximately four months. OpenAI's developer base on its Codex product expanded from the low hundreds of thousands to approximately four million this year, with annualized revenue now exceeding $25 billion. 

Cloud revenues are accelerating across the board, moreover, as Amazon Web Services (AWS) grew from 24% to 28% year-over-year and is now running at a $150 billion annual rate. Microsoft Azure ticked up from 38% to 39% growth, while Google Cloud was the standout, accelerating 63% year-over-year growth from the forties. Frank noted that Google is particularly well positioned to capture economics across the full infrastructure stack, earning both infrastructure-as-a-service revenue and token revenue through its Gemini model. On the productivity side, the capability of AI agents to complete extended tasks has expanded from five minutes in early 2025, to thirty minutes by year-end, to approximately three hours as of the time of recording,  with Anthropic's latest model in preview demonstrating this expanded capability and reflecting the rapid and compounding pace of improvement in model performance. 

If oil stays elevated because the war continues, shipping chips around the world may become a lot more expensive. What does that do to the semiconductor industry, and does it actually slow the AI rollout, or do hyperscalers just absorb it? Or could companies feel this is coming and there is a demand pull forward?

Frank Downing addressed the question directly, noting that while rising oil prices instinctively raises concerns about shipping costs, GPUs are extremely high-value relative to their weight—a single card costing $30,000 to $40,000—making freight a rounding error relative to the value of the systems being deployed. On the broader question of energy costs, the annual power and operating expenses for a one-gigawatt data center run approximately $1 billion against $15 billion in potential cloud revenue, meaning that even a significant increase in electricity prices would not undermine the unit economics of AI infrastructure. Frank also drew a clear distinction from the COVID-era supply shock, which disrupted unit production for three years; the current environment does not involve that kind of demand destruction. 

Within ARK's convergence framework, where multiple innovation platforms intersect, which position do you think the market is directionally right on but dramatically underestimating the magnitude of impact? And which amongst ARK's ETF best captures that asymmetry today?

Ovid Amadi, PhD, Multiomics Portfolio Manager and Director of Research, highlighted Tempus as one of the strongest convergence stories in the portfolio and among the most mispriced. Tempus operates across at least four intersecting innovation platforms: multi-omics sequencing in oncology and rare disease; AI applied to molecular and phenotypic data to support pharmaceutical drug development; AI-enhanced interpretation of molecular diagnostics; and AI-powered cardiovascular algorithms capable of identifying patients at risk of atrial fibrillation or heart failure from routine EKG data. Despite that multi-layered platform, Tempus trades at ~4.5 times 2027 revenue compared to roughly 8 times for peers with similar growth profiles a discount of nearly 50%. 

Tesla has mentioned that AI 3 vehicles will not have the capability for robotaxi, when in the past they said they could. What gives you confidence that AI 4 will solve robotaxi and not AI 5?

Tasha Keeney, CFA, Director of Research, Autonomous Technology & Robotics / Director of Investment Analysis, noted that Tesla is already conducting driver-out operations on the AI4 chip, providing direct real-world evidence that the platform can support unsupervised autonomy. Tesla's history with the AI3 hardware involved similar back-and-forth about its capability threshold, making the current announcement about AI3 consistent with historical precedent rather than a new concern. The company also has confirmed that FSD version 15, the software update planned for robotaxi deployment, is compatible with AI4, the chip installed in Tesla's existing passenger vehicle fleet. 

Beyond the technical evidence, Tasha pointed to Tesla's manufacturing scale as a structural advantage that no competitor has yet to replicate. Waymo, the most established robotaxi operator, has a reported fleet of ~3,000 vehicles, a quantity that Tesla can produce in a single day.  

There was an announcement about Circle's ARC network token, and that ARK Invest was an investor. What is Circle's ARC and why is it important? 

Lorenzo Valente, Director of Research, Digital Assets, explained that Circle—the second-largest stablecoin issuer, with ~25% market share by supply and over 60% of USDC transaction volume—has spent years building a three-layer ecosystem around stablecoins. The application layer includes agent toolkits, a digital wallet, and the Circle Payments Network—a cross-border business-to-business (B2B) payment infrastructure. The asset issuance layer encompasses USDC, EURC (a Euro stablecoin), USYC (a tokenized money market instrument), and the forthcoming Circle Bitcoin (BTC). What had been missing until now was the third layer: purpose-built blockchain rails to serve as the base infrastructure for the entire ecosystem. 

Conclusion:

The May webinar illustrates how ARK's convergence framework is maturing from an analytical thesis into a set of tangible, measurable outcomes. AI infrastructure economics are proving more durable and scalable than the market anticipated; a diagnostics company is applying machine learning across four distinct innovation platforms; a robotaxi program is bolstered by manufacturing scale that no competitor has matched; and a stablecoin issuer has constructed institutional-grade blockchain infrastructure. The common thread? Declining costs, expanding data advantages, and compounding platform capabilities are widening the gap between innovators and incumbents. At ARK, we remain focused on identifying and holding positions whose general direction the market might read directly but whose ultimate potential magnitude remains dramatically underestimated. 

Make sure to check out our In The Know video series for deeper insights. 

Important Information 

Investors should carefully consider the investment objectives and risks as well as charges and expenses of an ARK Fund before investing. This and other information are contained in the ARK Funds’ prospectuses and summary prospectuses, which may be obtained by clicking here. The prospectus and summary prospectus should be read carefully before investing.  

An investment in an ARK Fund is subject to risks and you can lose money on your investment in an ARK Fund. There can be no assurance that the ARK Funds will achieve their investment objectives. The ARK Funds’ portfolios are more volatile than broad market averages. The ARK Funds also have specific risks, which are described in their respective prospectuses.  

The ARK DIET Buffer ETFs have characteristics unlike traditional investment products and are not suitable for all investors. 

The information provided in this material is for informational purposes only and should not be used as the basis for any investment decision and is subject to change without notice. It does not constitute, either explicitly or implicitly, any provision of services or products by ARK, and investors should determine for themselves whether a particular investment management service is suitable for their investment needs. All statements made regarding companies or securities are strictly beliefs and points of view held by ARK and are not endorsements by ARK of any company or security or recommendations by ARK to buy, sell or hold any security. Historical results are not indications of future results.  

Certain of the statements contained in this material may be statements of future expectations and other forward-looking statements that are based on ARK's current views and assumptions and involve known and unknown risks and uncertainties that could cause actual results, performance or events to differ materially from those expressed or implied in such statements. ARK assumes no obligation to update any forward-looking information. ARK and its clients as well as its related persons may (but do not necessarily) have financial interests in securities or issuers that are discussed. Certain information was obtained from sources that ARK believes to be reliable; however, ARK does not guarantee the accuracy or completeness of any information obtained from any third party. 

Digital assets, often referred to as cryptocurrencies, such as bitcoin and ether are relatively new investments, which have unique and substantial risks and which may be more volatile than other types of investments. ARK strongly encourages any investor considering an investment in any digital asset to consult with a financial professional before investing. All statements made regarding digital assets are strictly beliefs and points of view held by ARK and are not recommendations by ARK to buy, sell or hold any digital asset. Historical results are not indications of future results. 

To view the top ten holdings for ARKK click here. 

To view the top ten holdings for ARKQ click here. 

To view the top ten holdings for ARKW click here. 

To view the top ten holdings for ARKG click here. 

To view the top ten holdings for ARKF click here. 

To view the top ten holdings for ARKX click here. 

To view the top ten holdings for ARKVX click here. 

Holdings are subject to change 

Foreside Fund Services, LLC, distributor.