Baron Technology Fund | Q3 2025

During the third quarter, Baron Technology Fund® (the Fund) posted a positive return of 5.89% (Institutional Shares), but underperformed the MSCI ACWI Information Technology Index (the Benchmark), which gained 12.76%; the Invesco QQQ Trust (the QQQ), which increased 8.94%; and the broader S&P 500 Index, which rose 8.12%. For the year-to-date period, the Fund appreciated 18.25%, trailing the Benchmark, which returned 22.83%, but outperforming the QQQ and the S&P 500 Index, which gained 17.92% and 14.83%, respectively.*
 

Market Backdrop 

U.S. equities were broadly higher in the third quarter, building on gains from the prior period. The S&P 500 and NASDAQ Composite Indexes set new record highs throughout the quarter, most recently on September 22.

The dominant driver of market strength was the increased likelihood of Federal Reserve (the Fed) rate cuts, prompted by signs of weakness in the labor market and more dovish Fed commentary. Fed rate cut expectations rose in early August following a much weaker-than-expected July non-farm payrolls report and significant downward revisions to prior payroll numbers. Dovish Fedspeak intensified as the month wore on, with Chairman Powell hinting a possible interest rate cut while delivering remarks at the Fed’s annual Jackson Hole conference. The Fed ultimately resumed its rate-cutting cycle at the September meeting, lowering its policy rate by 25 basis points to a range of 4% to 4.25%, after being on hold since last December. Robust corporate earnings, a lull in trade uncertainties, a resilient consumer, increased M&A and IPO activity, and sustained AI optimism also contributed to market gains during the period.

Performance

The Fund posted positive gains but underperformed Benchmark during the quarter. Relative underperformance was largely due to stock selection. This was a combination of poor performance of certain stocks owned in the portfolio, some of which fall outside of the GICS defined Information Technology (IT) sector, and strong performance of mega caps we didn’t own. About a third of the relative shortall came from poor stock selection in IT, while most of the remaining underperformance came from unique investments in other sectors, particularly Communication Services and Consumer Discretionary.

In the IT sector, solid gains and relative performance from Taiwan Semiconductor Manufacturing Company Limited (the world’s advanced semiconductor manufacturing champion), Broadcom Inc. (AI networking and custom accelerator winner), Lam Research Corporation (leading semi-cap equipment company), and Shopify Inc. (e-commerce platform leader) were offset by sharp pullbacks in PAR Technology Corporation (restaurant technology provider) and Atlassian Corporation (software development and team collaboration tools leader), which are addressed below. The Fund’s relative underweight in Microsoft Corporation contributed to relative performance, while the Fund’s underweight in Apple Inc. detracted, accounting for 58% of our IT sector underperformance.

Regarding Apple, the stock rallied significantly after a U.S. court issued a favorable remedy in the Department of Justice’s antitrust case against Google, alleviating market fears of a worst-case scenario. The ruling was a positive outcome for Apple, preserving its more than $25 billion high-margin traffic-acquisition revenue stream from Google and other search providers, and leaving the door open for potential future AI-related payments. This news, along with positive early data on iPhone 17 demand, drove Apple’s stock to outperform.

In Communication Services, strong gains and relative performance from Reddit, Inc. (digital advertising leader) was overshadowed by poor performance from Spotify Technology S.A. (global streaming music and content leader) and The Trade Desk (digital advertising platform provider). Spotify, which remains a top performer on a year-to-date basis (up 56%) but retreated somewhat during the period, impacted relative performance because it is one of the largest overweight positions in the portfolio.

In Consumer Discretionary, strong returns and relative performance from Tesla, Inc. (electric vehicle (EV) and autonomous driving pioneer) was negated by disappointing performance from Amazon.com, Inc. (dominant e-commerce platform and cloud computing leader) and Duolingo, Inc. (leading language learning platform), discussed below. Amazon’s stock was roughly flat in the quarter but impacted relative performance because it is large overweight position in the portfolio.

AI Update

The market has recently absorbed a surge of activity around AI infrastructure buildouts. OpenAI is at the center of these developments, announcing partnerships with NVIDIA, Broadcom, and Advanced Micro Devices for 10 gigawatts (GW), 10 GW, and 6 GW of compute capacity respectively, along with more than 5 GW of commitments to Oracle for cloud infrastructure as part of the Stargate project (one or more of these vendors’ compute equipment will power Oracle-Stargate data centers). In addition, NVIDIA CEO Jensen Huang proclaimed in September that global AI-infrastructure spending could total $3 trillion to $5 trillion by 2030.

To be clear, in our industry and individual company models/estimates, we do not assume the underlying AI infrastructure market hits the figures implied by the announced partnerships or Jensen’s 2030 target, nor do we need the market to hit these levels to earn our targeted long-term returns. We assess announcements and projections like these and others across the broader AI ecosystem from first principles, conducting intensive research to measure the addressable markets and understand technical architectures and differences; talking to everybody (as Ron says), including business leaders, other investors and analysts, and industry experts; analyzing from top-down and bottom-up perspectives; and, as I always say, trying to separate the signal from the noise and focusing on what’s real rather than what’s said. With an appreciation that history does not repeat itself but often rhymes, we are rigorously examining the parallels and distinctions between today’s AI disruptions and inflections of today and past technology paradigm shifts (such as the “Dot-Com” era) and their broader economic implications, alongside lessons from historical infrastructure expansions. At the same time, we are quantifying the many AI addressable markets across sectors, while critically evaluating the usefulness and value of AI to real-world industries, businesses, workflows, and tasks. We remain attuned to ongoing debates around AI’s return on investment and potential bubbles and are closely monitoring pathways to monetization and value creation, including through cost efficiencies. We invite our shareholders to watch Michael Lippert’s recent appearance on Wealthtrack, “The State of the AI Investment Boom,” for further discussion of many of these topics.

From a top-down perspective, while even the low end of Jensen’s forecast and the announced compute programs seem mammoth at first blush, the global economy is expected to reach $140 trillion by 2030, meaning this $3 trillion market estimate would be just about 2% of the global economy. This strikes us as not unrealistic for the most critical infrastructure of the 21st century, and reasonably consistent with historical infrastructure builds that have generally peaked at 2% to 5% of GDP.

Moreover, as we have written before, technology spending today accounts for just 5% of global GDP. We live in a digital age. Every digital interaction from today forward, across business and consumer applications, will have AI at the core of the intelligence layer of the application. Much of incremental GDP growth will incorporate AI and technologies like agentic AI will displace large swaths of existing GDP. If AI innovations propel technology spending towards just 10% of global GDP by 2030, we’re talking about $14 trillion, quite a reasonable justification for large-scale AI infrastructure investments.

From a more bottom-up perspective, we attempt to break this down further and identify the true unlocks required from AI models and systems to justify the trillions of dollars that the leading AI labs now plan to deploy. The most immediate opportunities for AI are not in expanding the economic pie but in redistributing it across existing large markets. For example, the global e-commerce market is estimated to hit $6.4 trillion this year, representing about 20.5% of global retail sales. We find ourselves on the edge of the first holiday season where consumers will put AI commerce to the test, with such features as OpenAI’s Instant Checkout.⁴ Over the remainder of this decade, AI should influence and monetize a growing share of this massive consumer market. The global digital advertising market, projected to approach $850 billion this year, should also continue to see share shifts as AI reshapes consumer engagement. Similarly, the global software market, covering enterprise, productivity, and application software, is estimated to reach over $800 billion in 2025. AI features and services are already disrupting this market–AI user interfaces, AI predictive intelligence, AI agentic workflows assisting and replacing humans in code development and customer service, etc.–and are expected to drive future software spending and capture a large share of the market opportunity.

The second growth vector, labor replacement, is more transformative in theory but likely to be slower in practice. Human labor represents 45% to 50% of the global economy. Optimists believe AI agents will eventually replace significant portions of knowledge work, and physical AI like autonomous vehicles and robots will begin to displace a broad array of human labor. We are carefully monitoring product announcements and user adoption, but we do appreciate that this transition will evolve over the remainder of the decade and face social, political, and regulatory hurdles.

The most compelling long-term vector is innovation itself. Human progress and GDP growth have historically been driven by research and development, which in turn depends on the number of people with the requisite intelligence and creativity in each field. This is the essence of the Artificial General Intelligence and Artificial Super Intelligence narratives: once machines reach or exceed human-level intelligence, we could, in principle, run millions, or even billions, of R&D agents and experiments in parallel inside AI data centers, accelerating discovery and unlocking entirely new drivers of global growth. As Elon Musk explained during the Grok 4 Livestream, “we’re at the beginning of an immense intelligence explosion…[w]e’re at the intelligence big bang,” but the “real test” for AI is “reality.” Will AI “invent a new technology…improve the design of a car or a rocket or create a new medication…Does the rocket get to orbit? Does the car drive? Does the medicine work?”

Here’s just one real-world example already in Baron portfolios. Earlier this year we participated in the IPO of Heartflow, Inc., a health care innovator that uses AI to analyze a coronary CT (computed tomography) angiogram scan, creating a personalized 3D model of heart arteries. The AI, combined with computational fluid dynamics and human analysts, simulates blood flow through this model to assess the impact of blockages and provide doctors with a “CT-derived fractional flow reserve” value, which helps diagnose coronary artery disease and personalize treatments. We are carefully monitoring the long-term AI innovation flywheel.

We remain optimistic about the broad trajectory of AI-driven progress but disciplined in distinguishing real, measurable advances from speculative ambition. Our focus is on identifying the best positioned players to capture enduring value amid this fast-moving, high-potential world of intelligence creation.

Below is a partial list of the secular megatrends we focus on:

• AI
• Semiconductors
• Cloud computing
• Software-as-a-service
• Digital media/entertainment
• Targeted digital advertising
• E-commerce
• Digital/AI healthcare
• Cybersecurity
• EVs/autonomous driving
• Electronic payments
• Robotics
• Data center power/cooling

We continue to run a high-conviction portfolio with an emphasis on the secular trends cited and listed. Among others, during the third quarter we initiated positions or increased portfolio weights in the following positions:

• Networking: Arista Networks, Inc., Lumentum Holdings Inc., and Coherent Corp.
• Data center power/cooling: AAON, Inc.
• E-commerce: Eternal Limited
• EVs/autonomous driving: Tesla, Inc.
• Software-as-a-service: Intapp, Inc., Samsara Inc., Synopsys, Inc., Via Transportation, Inc., Netskope, Inc., and Guidewire Software, Inc.

• Semiconductors: Taiwan Semiconductor Manufacturing Company Limited and NVIDIA Corporation

   

NVIDIA Corporation is a semiconductor and systems company specializing in compute and networking platforms for accelerated computing. NVIDIA has captured a dominant position in AI infrastructure with a comprehensive portfolio spanning semiconductor accelerators, networking solutions, modular and rack-scale systems, and software. Shares rose during the quarter as investor confidence in AI infrastructure expansion grew. For its July 2025 quarter, NVIDIA reported 56% total and data center revenue growth. Looking forward, NVIDIA disclosed near-term visibility of tens of GWs in AI buildouts, including the 10 GW agreement with OpenAI, with each GW representing an estimated $35 billion total addressable market (TAM). During its August earnings call, NVIDIA’s management declared: “We are at the beginning of an industrial revolution that will transform every industry. We see $3 trillion to $4 trillion in AI infrastructure spend by the end of the decade.” As AI infrastructure investment accelerates, NVIDIA’s leadership continues to strengthen through durable moats across compute silicon, networking, systems, software, and supply chain. We remain confident in AI’s potential to transform the global economy and in NVIDIA’s pivotal role as the leading enabler of that transformation, positioning it to capture significant long-term value in the AI era.

Broadcom Inc. is a leading semiconductor and enterprise software company, generating approximately 60% of revenue from semiconductors and 40% from software. The company is strategically positioned at the intersection of high-performance AI compute and networking infrastructure, while also demonstrating disciplined execution in software. Broadcom has extended its leadership in networking silicon from the cloud era into the AI era and is regarded as the most reliable silicon partner for AI foundational model builders designing custom chips to train frontier models. Shares rose during the quarter on accelerating momentum in Broadcom’s AI product lines and strong business visibility into next year. In its July quarter, Broadcom reported almost $16 billion in total revenue, up 22%; $5.2 billion in AI revenue, up 63%; and $6.8 billion in software revenue, up 17%. Broadcom continued to demonstrate excellent profitability, with EBITDA margins over 67% and free cash flow margins at 44%. While Broadcom continues to execute with its key custom AI accelerator customer, Google, it is on track for volume production with two additional customers (likely Meta and ByteDance), has secured a fourth customer with orders worth $10 billion next year, and announced a 10 GW deal with a notable fifth customer, OpenAI. Beyond AI, Broadcom is advancing VMware integration, while its non-AI semiconductor businesses appear to be bottoming and may gradually recover in the coming quarters. We retain our long-term conviction in Broadcom’s position within the AI ecosystem.

Tesla, Inc. designs, manufactures, and sells EVs, related software and components, solar products, and energy storage solutions. Tesla shares contributed to our quarterly performance due to three key catalysts. First, Tesla’s core automotive business saw an uplift in sentiment and growth. The company announced record third quarter results for both vehicle deliveries and energy storage deployment, surpassing strong expectations. While the expiration of U.S. consumer EV tax credits likely positively influenced domestic sales, investors also identified enthusiastic consumer response to a new Model Y variant in China, elevating expectations for this variant’s traction. Second, investor confidence in the company’s ambitious long-term vision and in Elon Musk’s leadership was reinforced by a newly proposed CEO compensation package and nearly $1 billion in personal share purchases by Musk. To fully benefit from the compensation plan, Elon and Tesla will need to achieve close to an 8-fold increase in market cap and over a 25-fold expansion in adjusted EBITDA. Alongside these financial goals, the company must also deliver on its revolutionary AI initiatives. Over the coming decade the company will need to deploy one million robotaxis in commercial operations, selling one million humanoid robots, reaching 10 million active Full-Self-Driving (FSD) subscribers, while continuing to sell millions of cars to consumers each year. Finally, Tesla’s AI initiatives continue to advance rapidly. Tesla’s Austin robotaxi network saw its operational footprint grow from 20 to over 170 square miles in just two months after its June 2025 launch, with further expansion on the horizon, including in new cities. Additionally, the forthcoming FSD Version 14 is anticipated to be a critical step toward broader autonomous capabilities. Elon Musk previously stated that V14 “is the second biggest update to Tesla AI/Autopilot ever after V12. It feels alive,” noting that “With FSD V14, we’re on track to enable unsupervised driving in select areas by the end of the year, assuming regulatory approval.” In addition, Tesla’s humanoid robot is expected to transition from design to volume production next year as Tesla finalizes its latest Optimus design.

PAR Technology Corporation is a leading software, hardware, and service provider to the food service industry. Shares of PAR fell during the quarter as the company lowered its annual recurring revenue (ARR) growth outlook for the year to 15% (from prior 20% expectations) due to weak macro conditions for its restaurant customers and the purposeful delay of some new rollouts to focus on winning late-stage opportunities with some of the largest restaurants in the industry. Additionally, management indicated that it believes the right core ARR growth rate for the business going forward is now 15%, not the prior 20% target. Despite the weaker start to the year and lower “core” growth outlook, for the next 12 months the company remains confident it can deliver 20% ARR growth supported by a strong pipeline of contracted customers, including its rollout to Burger King. Additionally, we have confidence that the company is well positioned to win at least one of its three late stage “mega deals,” which would meaningfully improve the growth outlook for 2026 and 2027, given they are sized as equal to or multiples of the Burger King deal, which is PAR’s largest deal to date. In general, as more enterprise-scale restaurants look to upgrade their technology stacks, PAR remains well positioned to win an outsized share as the leading cloud-based multi-product platform in the industry. Also, despite the lower growth outlook the company can manage expenses in lockstep, which means our estimates for rapidly scaling profitability in the coming years remain virtually unchanged.

The Trade Desk, a leading digital-advertising demand-side platform (DSP), detracted from performance this quarter after reporting results and guidance that fell short of expectations. This frustrated investors amid strength in the broader advertising market and following strong first quarter results that exceeded both Baron and consensus forecasts. We believe several factors contributed to second quarter revenue growth decelerating more than anticipated: (i) a slower-than-anticipated rollout of the company’s new platform, Kokai (which has since re-accelerated); (ii) broad macro uncertainty around tariffs that impacted larger brand advertisers, who appear to have pulled-forward advertising spending into the first quarter in front of the administration’s April tariff roll out; and (iii) the company’s restructuring of sales and account coverage, particularly focused on its largest clients. Importantly, our conversations with industry participants suggest that Amazon’s competing DSP has not yet taken meaningful share from Trade Desk and that most advertiser interest in Amazon appears limited to its exclusive Prime Video inventory. That said, we are closely monitoring Amazon’s push to scale its offering by undercutting Trade Desk on fees and signing new publishing partners like Roku, Netflix, and Spotify. Despite these issues, our research supports our view that Trade Desk remains the product leader in the DSP space and is well positioned to benefit from strengthening brand advertising trends in the second half of 2025. Trade Desk no longer commands a premium valuation, trading more in line with peers and reflecting tempered growth expectations. The company remains financially stout, with over $1.7 billion in cash on its balance sheet, strong free cash flow generation, and the capacity to aggressively repurchase a significant portion of its market cap. We continue to hold our position in Trade Desk, while keeping a close eye on the evolving competitive landscape.

As stated above, global streaming music and content leader Spotify Technology S.A. remained a top performer on a year-to-date basis (up 56%) but retreated somewhat during the period. It is reported as a third quarter detractor because it is a large position in the portfolio.

Despite language learning platform Duolingo Inc. reporting a strong second quarter with 40% daily active user growth and 41% revenue growth, the stock traded down intra-quarter following an OpenAI product demonstration around language learning and third-party data trends that indicated slowing user growth on the app. While the OpenAI demo was impressive, we believe that OpenAI’s offering requires substantial effort on the part of the user, lacks Duolingo’s expert-approved exercises, and most importantly, lacks the engaging gaming mechanics that Duolingo is known for and keeps users motivated to keep coming back. Duolingo was also negatively impacted by third-party data trends showing a deceleration in user growth, following social media controversy over the company’s decision to use more AI to slow hiring. While we believe that daily active user growth has slowed from second quarter’s 40-plus percent year over year level, we cbelieve that the company should continue to grow users north of 20% (and revenue growth ahead of user growth) English speaking regions continue to maintain strong engagement and user growth, and advanced English courses unlock a larger addressable market in the hundreds of millions. The company’s newest AI offering, Max, allows users to practice speaking with a virtual tutor, allowing for greater engagement, improved learning, and higher monetization for the company. We continue to closely monitor trends in user growth and monetization.

Atlassian Corporation is a leading team collaboration and productivity software vendor. The company initially focused on serving software engineers (over 20 million worldwide), but its newer products, features, and use cases address a much larger set of users, including business teams engaged in product development (over 100 million worldwide) and the much broader group of knowledge workers (over 1 billion globally). Despite solid results and guidance, shares declined on continued fears that AI software code development tools would pressure the number of software developers and Atlassian’s own growth from this segment. Our research indicates that AI will spark more software code to be written and applications to be adopted, resulting in more—not less—demand for Atlassian’s portfolio of products. We believe these products will be monetized through a combination of both seats and consumption elements. We are carefully monitoring all risks related to AI but have yet to see any impact on the company’s business.

We invest in companies of any market capitalization that we believe will deliver durable growth from the development, advancement, and/or use of technology. We invest principally in U.S. securities but may invest up to 35% in non-U.S. securities.

At the end of the third quarter, the largest market cap holding in the Fund was $4.5 trillion and the smallest was $880 million. The median market cap of the Fund was $29.7 billion and the weighted average market cap was $1.4 trillion.

We had investments in 46 unique companies. Our top 10 positions accounted for 59.1% of net assets.

To end the quarter, the Fund had $103.2 million in net assets.

Fund flows were slightly negative in the third quarter but remain meaningfully positive this year.

* Individual weights may not sum to the displayed total due to rounding.

We initiated a position in Lumentum Holdings Inc. this quarter. The company is a leading photonics and optical components company specializing in lasers (light emitters) used for optical communication within data centers. As data centers become larger and more communication-intensive, the industry is transitioning toward newer classes of laser technology, an area where Lumentum holds a leading market share. The supply of these advanced lasers remains structurally constrained, positioning Lumentum to exercise pricing power and gain share in the photonics industry. Moreover, Lumentum’s leadership in next-generation optical technologies, such as co-packaged optics (optical components integrated directly with silicon chips) and optical circuit switching (enabling direct optical routing without electrical conversion), positions the company well to capture emerging markets that could redefine data center architectures. These technologies enhance performance, reduce power losses, and expand the TAM for optical solutions. We believe Lumentum is exceptionally well positioned to benefit from both the secular growth of traditional optical communication and the emergence of new markets that open massive incremental opportunities in the buildout of the AI infrastructure ecosystem.

We added shares of Duolingo, Inc. in the quarter, as discussed previously.

This quarter, we re-established a position in Arista Networks, Inc., a leading provider of high-performance networking solutions for data centers, cloud providers, and enterprises. Arista’s advanced switching and routing platforms, powered by its proprietary software, offer enhanced scalability, automation, and flexibility. The company generates revenue through hardware sales bundled with software and post-contract support services, serving major cloud players like Microsoft, Meta, and Oracle, along with a growing range of enterprise customers. We are witnessing an unprecedented buildout of AI infrastructure, where networking is becoming an increasingly critical component. While NVIDIA offers a comprehensive technology stack for AI data centers, Arista stands out as the leading networking company with best-in-class Ethernet solutions. Its products not only interconnect servers within data centers and link multiple data centers together but will also extend to emerging architectures that connect AI accelerators within the rack. We believe Arista is well positioned to capture a meaningful share of the data center networking stack as AI cluster builders prioritize performance optimization and vendor diversification.

We exited our position in CyberArk Software Ltd. after its acquisition by Palo Alto Networks Inc. was announced.

We also exited our position in Reddit, Inc. due to valuation.

We sold LPL Financial Holdings Inc., MNTN, Inc., Park Systems Corporation, and eDreams ODIGEO SA to allocate capital to higher conviction names.

We trimmed monitoring and security software platform Datadog, Inc. and reallocated capital to new and existing software names, such as Intapp, Inc., Samsara Inc., Via Transportation, Inc., and Netskope, Inc.

Looking Ahead

Looking ahead, our focus is unwavering: owning category-defining technology businesses that sit at the heart of durable, secular growth trends. Led by visionary, execution-driven management teams, these companies convert breakthrough innovation into expanding free cash flow. We believe this combination uniquely positions the Fund to compound our investors’ capital over the long term.

Fund Action

On or about December 12, 2025, Baron Technology Fund® (the “Fund”) will convert from a mutual fund into an exchange-traded fund (ETF). We are converting the Fund to an ETF to provide shareholders with greater flexibility, lower costs, and increased transparency.

Upon conversion, a new fee structure will go into effect where a unitary management fee of 0.75% will be adopted. There will be no change to the investment strategy or the portfolio management team.

All shareholders will receive a direct notification.

• If clients hold shares in a brokerage account that can hold ETFs:
  Shares will automatically convert—no action is needed.
• If clients hold share in accounts that cannot hold ETFs:
  Shares need to be transferred to an existing or new brokerage account that allows ETF investments before the conversion date. If no transfer is made, shares will be liquidated at the time of conversion, and proceeds will be paid in cash (net of any applicable fees and expenses).

For questions, please call the Baron Capital customer service line at 1-844-673-0984 Monday through Friday from 9:00 a.m. to 6:00 p.m. EST.