Commentary: Large Cap Growth ESG Strategy

• The Strategy meaningfully underperformed the Russell 1000 Growth Index through one of the most powerful rallies in its history, with underweights to mega cap AI beneficiaries as well as lower-quality AI-related names the largest headwinds. Stock selection in health care also weighed on performance.
• We believe resurgent growth in the economy ex-tech from fiscal stimulus and easing monetary policy will lead to broadening participation within our benchmark. In particular, we are targeting opportunities in health care and short-cycle industrials.
• ClearBridge evaluates environmental and social shareholder proposals on a case-by-case basis, considering whether the ask from the proposal has merit and whether its wording diminishes or enhances shareholder value.

Large cap growth stocks maintained market leadership in 2025 as accelerating spending and enthusiasm related to generative artificial intelligence (Gen AI) and continued strong earnings growth from mega cap companies enabled the benchmark Russell 1000 Growth Index to finish up 18.6%. The technology-heavy NASDAQ Composite rose 20.4% for the year while the S&P 500 Index advanced 17.9%.

After outperforming in a negative first quarter, the ClearBridge Large Cap Growth ESG Strategy trailed the benchmark’s powerful 53.9% rally from the post Liberation Day lows in early April through late October. It continued to lose ground through the rest of the fourth quarter, underperforming the benchmark’s 1.2% quarterly advance by about 170 basis points (gross of fees). In aggregate, the Strategy underperformed the Russell 1000 Growth Index (RLG) by approximately 900 bps for the year.

Over the history of the index, the RLG’s recent rally has only been topped by rebounds from the COVID-19 and Global Financial Crisis in 2020 and 2009, respectively, as well as the run-up to the dot-com bubble in 1999 (Exhibit 1). What surprised us most about this year’s surge was the pace of capex increases by cloud hyperscalers, the sudden emergence of privately held OpenAI and Anthropic as sources of billions of dollars in AI spending and the move by Alphabet to start selling custom AI chips to its competitors.

Exhibit 1: Best Six Month Returns for Russell 1000 Growth Index

The majority of the Strategy’s 2025 underperformance — roughly 500 bps — can be attributed to our differentiated AI and mega cap positioning compared to the benchmark (with poor stock selection in health care the second-largest detractor). We were right on Nvidia, a stock we first purchased in 2018 and that has been one of the top portfolio holdings for the last several years. Additionally, we knew an AI revolution was developing, but we underestimated the magnitude of the AI spending growth we witnessed in 2025. We also misjudged the extent to which the market would discount the demise of application software companies in an AI world. Fast moving markets are a challenge for a low-turnover portfolio, although we were quite active from Liberation Day on in seeking to upgrade our AI participation through the purchases of Broadcom, Marvell Technology, Datadog and Oracle, as well as exits from names where we had less confidence in long-term earnings power including Accenture and application software makers Workday and Adobe.

In hindsight, we did not scale up our positions in Broadcom and Alphabet enough. Broadcom shares rose more than 40% for the year but the maker of custom silicon chips for AI workloads was one of the Strategy’s largest relative detractors due to our roughly 300 bps underweight. We had been consistently adding to Alphabet since repurchasing the stock in April 2024, as it had been trading at a sub-market multiple, but regret not increasing our exposure more aggressively after the DOJ antitrust case concluded with less onerous penalties than feared. The company’s success with its Google Gemini chatbot and TPU chips for AI workloads boosted its shares more than 60% for 2025, causing our ~350 bps underweight to create a significant performance drag.

Exhibit 2: Divergence Emerging Among Mega Caps

Much of the broadening within IT has occurred among more speculative names in contrast to our buckets approach to portfolio construction, which limits our exposure to this group of holdings. For example, not owning AI operating system software maker Palantir Technologies was a headwind but, at current levels, we think very aggressive future growth assumptions are already factored into the valuation. In fact, we have seen wide dispersion between momentum stocks and quality stocks in 2025, particularly in the second half of the year (Exhibit 3). Our relative overweight to quality and relative underweight to momentum factors negatively affected results.

Exhibit 3: Momentum Stocks Have Dominated Since Liberation Day

Meanwhile, the Strategy’s more diversified exposure to AI through ownership of Eaton, Accenture and Equinix did not add significant value in 2025. Eaton, a manufacturer of electrical components critical to the operation of data centers, has had supply constraints that limit its ability to vastly exceed its targeted growth rate which may also limit its ability to garner price increases. Accenture, an IT consultant helping enterprise customers deploy AI technologies, was hurt by its association with application software makers and its outsourcing business was deemed at risk of being disintermediated by sophisticated large language models (LLMs). We exited our position in Accenture in the third quarter as part of our AI repositioning. Data center operator Equinix should benefit from the growing importance of sharing data across clouds, but the returns on a large capital spending project will delay revenue growth acceleration until 2027. We view the company as a later stage AI beneficiary, especially if all the current spending begins to produce a positive return on investment.

The performance of our health care holdings was another disappointment. UnitedHealth Group, a consistent contributor in the portfolio’s stable bucket for more than a decade, suffered from a combination of negative sentiment and severe mismanagement that caused the stock to lose about half its value in 2025 by the time we exited the position in August. While we had been trimming UnitedHealth consistently since the fourth quarter of 2024, we also believed the company deserved some leeway to turn things around given its long-term track record. In hindsight, UnitedHealth disclosure had always been below average due to the nature of its regulated businesses, as well as the size and scope of the company’s operations, but that lack of disclosure also made it hard to assess the execution issues that the company faced in 2025. Ultimately, despite a decade of solid returns, we lost confidence in UnitedHealth’s ability to navigate a turnaround under new leadership and exited the position.

We were too early in exiting pharmaceutical maker Eli Lilly, which we sold in July at what we viewed as fair value and due to uncertainty over reimbursements for GLP-1 treatments for diabetes and obesity and the continued prevalence of lower-priced competition. Since then, the U.S. government has struck a deal with Lilly to offer GLP-1s to Medicare and Medicaid patients, opening up millions of additional prescriptions, while readouts on the company’s oral GLP-1 treatment indicated a broader market than we had expected. Lilly shares surged in the fourth quarter, causing our lack of exposure to be a drag on relative results.

As with technology, we repositioned our health care exposure with the proceeds of these sales throughout the year by purchasing high-quality biotechnology company Vertex Pharmaceuticals, a leading contributor in the fourth quarter, while adding to our positions in Thermo Fisher Scientific and Intuitive Surgical.

After a flurry of activity post Liberation Day, our portfolio actions in the quarter were more limited. We closed a position in Starbucks to concentrate our exposure in restaurants into Chipotle Mexican Grill. While we still believe in the turnaround story at Starbucks under CEO Brian Niccol, improvements are taking longer than expected to accelerate the business. Both Starbucks and Chipotle are being impacted by macro headwinds in the restaurant sector due to a challenged low-income consumer; however, we believe Chipotle has faster long-term store growth potential and better unit economics, and we feel more comfortable in Chipotle’s price versus value provided to consumers.

We initiated positions in ServiceNow and Arista Networks in the IT sector. ServiceNow provides an end-to-end software-as-a-service (SaaS) platform to help enterprise customers automate and standardize business processes in areas like IT, customer services, sales and data security. ServiceNow also has a monetizable generative AI product being adopted by its customers which provides a measure of support in an environment where most application software companies are being pressured by fears of AI disintermediation.

Arista makes high-speed switches and networking equipment critical to cloud platforms and data centers. In addition to being well represented among hyperscalers and benefiting from the AI infrastructure capex cycle, Arista has a low but growing share in its enterprise business and is poised to continue gaining as companies update their campus networking infrastructure.

We will continue to be active in adding to positions or initiating new ones during pullbacks, as we did with Palo Alto Networks following its announcement of the Cyber Ark acquisition and Netflix when the stock sold off following disclosure of its intent to purchase Warner Bros. Discovery.

While AI capex has exceeded expectations thus far, we are still very early in the technology’s development.

Exhibit 4: More AI Capex Ahead

We see AI as the most transformative technology of our investment careers but, as with past innovations, there will be changes in perceived winners and losers as new LLMs are released, semiconductor innovation evolves and as monetization models become clearer. The growth market right now is momentum oriented and sentiment driven, with the latest data point from an AI bellwether determining near-term direction — similar to what we experienced in 2021. This is not a durable environment, and our challenge is positioning the portfolio so that sentiment shifts do not impede multiyear performance.

This backdrop causes us to be thoughtful in analyzing the potential long-term AI winners and opportunistic in taking advantage of the market volatility that comes with a secular growth trend as it evolves, while remaining cognizant of the downside risks inherent in such emerging trends. Oracle is an example of this volatility. The shares jumped over 30% after our purchase in July but are down 40% from their peak as investors question how soon AI capex will lead to revenue and profitability. While Oracle is a highly levered company, it generates free cash flow and has a large backlog of signed contracts. While there is uncertainty and execution risk with a spending ramp of the size and scale that the company has announced, the market is currently attributing very little value to Oracle being successful in its cloud business.

One of the lessons we have learned this year is the importance of being proactive and moving quickly when a secular theme takes hold. In addition to targeting new ideas at advantageous valuations, we are constantly revisiting the case for owning every holding, moving away from companies with lower visibility and subpar execution and reprioritizing our highest-conviction names. We are currently vetting several ideas in health care and seeing signs of improving activity in areas like short-cycle industrials. These exposures could prove beneficial should growth and earnings revisions in technology-related sectors become more normalized compared to the rest of the market. If not a leadership change from the mega cap names that have dominated RLG performance for the last three years, resurgent growth in the economy ex-tech in 2026 from fiscal stimulus and easing monetary policy would at least support broadening participation within our growth universe.

The ClearBridge Large Cap Growth ESG Strategy underperformed its Russell 1000 Growth Index benchmark in the fourth quarter. On an absolute basis, the Strategy delivered positive contributions across three of the nine sectors in which it was invested (out of 11 sectors total). The primary contributors to performance were the health care and consumer discretionary sectors while the communication services and IT sectors were the main detractors.

Relative to the benchmark, overall stock selection detracted from performance. In particular, stock selection in the communication services, IT and industrials sectors weighed on performance. On the positive side, stock selection in the consumer discretionary and consumer staples sectors and an underweight to IT contributed to performance.

On an individual stock basis, the primary detractors from relative performance for the quarter included Netflix, Eaton, underweights to Alphabet and Apple as well as not holding Eli Lilly. The leading contributors to relative returns were Intuitive Surgical, Thermo Fisher Scientific, ASML, Vertex Pharmaceuticals and an underweight to Microsoft.

Of the tools public equity investors can use to advocate for sustainable business practices, proxy voting is one of the more visible and powerful. It was vigorously debated in 2025. Throughout the year the SEC tightened parameters for shareholder proposals, strengthening the grounds on which they can be excluded from annual meetings.¹ It announced it would no longer “respond to no-action requests for, and express no views on, companies’ intended reliance on any basis for exclusion of shareholder proposals under Rule 14a-8,” with minimal exceptions.² The likely result will be to enable companies to exclude proposals without having to seek SEC approval, leading to fewer shareholder proposals making it to a vote.

Against this backdrop, the broad trends of the 2025 proxy season were a decline in environmental and social proposals and heightened scrutiny on governance issues. Major topics of environmental proposals filed included emissions disclosures and climate risk and plastic pollution. Social proposals, which were reduced in number, showed continued concern with workforce-related risks like pay equity, workplace safety, and diversity and inclusion. Like environmental proposals, social proposals received less support in 2025 than in previous years, although many of these proposals filed were perhaps “overly prescriptive, duplicative of existing disclosures, or insufficiently tailored to company-specific issues,”³ a reminder that such proposals need to be judged on a case-by-case basis.

Declines in environmental and social proposals and an increase in governance proposals (which received steady support, all told) were also reflected in ClearBridge’s voting activity in 2025 (Exhibit 5).

Exhibit 5: Shareholder Proposals Voted on by ClearBridge

The continued — and apparent increase in — relevance for governance topics reflects our view that good governance is a catalyst for value creation: board and chair independence reduces insular oversight; separating CEO and board chair roles reduces the potential for conflicts of interest; diversity on the board leads to more varied views and strengthens governance; board tenure should balance experience with innovation; linking compensation with sustainability factors could improve environmental stewardship and ensure the social license to operate. We have seen incremental improvements across many of these goals in recent years, and they remain worthy of supportive company dialogue.

Voting on a Case-by-Case Basis

Per ClearBridge’s Proxy Voting Policy, we evaluate certain environmental and social proposals on a case-by-case basis. While we would generally be supportive of ESG proposals, we also consider whether the ask from the shareholder proposal has merit and whether the wording in the proposal diminishes or enhances shareholder value.

We also take note if a proposal does not seem to recognize substantial improvements by the issuer on the requests being addressed. This is an important element of ClearBridge’s approach to proxy voting and our partnership approach to active ownership: we engage with CEOs, CFOs and other company leaders regularly about all factors that could materially affect value creation. This provides a valuable information component for assessing the merits of shareholder proposals.

Here we offer highlights of some recent ClearBridge votes and our thinking behind them.

Companies Are Making Sustainability Improvements

Amazon.com is a good example of a company that has made substantial improvements in areas where it nevertheless continues to see proposals: in 2025, for example, we examined a shareholder proposal asking the company to report on efforts to reduce plastic packaging. The company has received similar proposals for the past five years but has been making significant progress, addressing the resolutions of the proposals with improvements each year.

We chose not to support this proposal this year on the grounds that the company has already been reporting its plastic packaging reduction efforts and has quantified and published the improvements to the public each year. Such improvements include transitioning away from plastic in its outbound packaging and working with its vendors to let them ship in their own brand packaging via their Ships in Product Packaging (SIPP) program — reducing the use of an Amazon box on top of the product packaging. In addition, as of October 2024, Amazon has removed all plastic air pillows from delivery packaging used in its global fulfillment centers, which to date is the biggest decrease in plastic packaging in North America.

Moreover, through innovation and investment in technologies, processes and materials since 2015, Amazon has been able to reduce the weight of the packaging per shipment by 43% on average and avoided more than three million metric tons of packaging material. There are other achievements in packaging (both plastic and other materials) that the company has reported publicly.

Amazon is advancing partnerships and research to improve recycling infrastructure, engaging with organizations such as the Ellen MacArthur Foundation and The Recycling Partnership and demonstrating its efforts to align with industry peers, even if Amazon is not formally a signatory to the New Plastics Economy Global Commitment. We would still like to see Amazon publish an overall baseline of plastic used across its entire supply chain, to add to its robust reporting levels for outbound packaging practices.

Voting Requires Deep Knowledge of the Company

Our portfolio managers chose not to support a shareholder proposal asking Microsoft to report on the risks of its European Security Program (ESP) being used for censorship of free speech. We thought this proposal appeared to conflate a cybersecurity initiative with speech regulation and could mislead investors on the nature of the ESP. The company launched the ESP in response to the sharp rise in ransomware and cyberattacks involving espionage, data theft and disruption of democratic institutions.

Microsoft’s ESP provides structured, limited-scope support to governments by sharing insights into these threats and aligns with Microsoft’s Information Integrity Principles, which emphasize trusted information and freedom of expression rather than content moderation, surveillance or speech regulation. The company also participates in the Global Network Initiative (GNI), which independently evaluates its adherence to principles protecting privacy and free expression.

Executive Compensation Should Be Reasonable

We actively engaged UnitedHealth Group’s Board of Directors over the course of 2025 about the appropriateness of the compensation for their executive team.

The company serially missed earnings expectations, resulting in underperformance relative to the S&P 500 Index by 20% in both 2023 and 2024. Further, UnitedHealth had a major cybersecurity incident that jeopardized payments throughout the U.S. health care system, and public sentiment toward the company was at historic lows. Despite poor results, United asked investors to support pay increases for the CEO and CFO, while withholding any bonus payment to the family of murdered executive Brian Thompson. We opposed the proposed pay scheme, as did 40% of voting investors, and we accordingly expressed our views to the board.

Following the proxy vote, UnitedHealth announced it would replace both the CEO and the CFO. UnitedHealth’s board failed to hold either outgoing executive accountable for poor performance, and it allowed both of them to keep very significant unvested compensation. We again expressed our dissatisfaction to the board about its compensation decision.

Seeking to Enhance Shareholder Value

In voting proxies, we are guided by general fiduciary principles. Our goal is to act prudently, solely in the best interest of the beneficial owners of the accounts we manage. We attempt to provide for the consideration of all factors that could affect the value of the investment and will vote proxies in the manner that we believe are consistent with efforts to maximize shareholder values.

Among these factors would also be issuance of preferred shares. For example, the ClearBridge Emerging Markets Strategy portfolio managers considered a proposal at Localiza, a Brazilian car rental company, which held an out-of-cycle extraordinary general meeting to approve the creation of preferred stock.

Although the issuance of preferred stock adds complexity to common shareholders, the background here was telling: Brazil was to initiate a new dividend tax in January 2026 and companies were advancing dividends and bonus share issues to use up distributable reserves before the year end.

We judged that shareholder voting rights were being maintained and the company was attempting to issue bonus shares before the year-end tax increase. Ultimately, we agreed with management that the share issue was in the interest of shareholders and voted in favor of the proposal.