Commentary: All Cap Growth Strategy

• Leadership shifted among growth stocks as investors moved away from high-momentum semiconductor names toward companies demonstrating clearer AI monetization, while surging Treasury yields and higher oil prices weighed on cyclical areas of the market.
• The Strategy outperformed on strong stock selection, as the unwind in AI momentum rewarded our more disciplined, valuation-sensitive approach and software holdings rebounded.
• We used AI-related volatility to broaden our participation in AI infrastructure through new positions in Advanced Micro Devices, Micron Technology and Vertiv.

U.S. equities saw shifting leadership in the third quarter, with a selloff in higher-beta AI stocks offsetting corporate earnings strength in a period that saw Treasury bond yields rise to multidecade highs. The benchmark Russell 3000 Growth Index rose 0.5%, trailing the 2.3% return of the Russell 3000 Value Index, while mid and small cap growth stocks moved lower.

Within growth, leadership moved away from the semiconductor winners of the first half of 2026 toward hyperscalers and software companies demonstrating clearer paths to AI monetization. The June Russell index reconstitution was also consequential, materially increasing semiconductors’ weight in large cap growth indexes while reducing the weight of software, altering both the benchmark’s character and the relative performance backdrop.

AI was the main source of volatility during the quarter. A sharp momentum selloff in July unwound crowded positions in certain semiconductor and AI-exposed industrial stocks as investors questioned the durability of and payoff from capital spending. These concerns resurfaced in September after a coordinated call among the leading AI labs to slow the pace of frontier model development. Anthropic’s delayed initial public offering and a significant rise in usage of lower-priced open-source models intensified the debate around AI monetization.

Macro forces also shaped sector returns within the benchmark. Continued conflict with Iran drove oil prices higher, making energy the benchmark’s best-performing sector. The Federal Reserve’s first rate increase since 2023 and a sharp rise in Treasury yields weighed on rate-sensitive and cyclical stocks, leaving utilities and industrials, where AI power and infrastructure names also pulled back, as the weakest sectors.

The ClearBridge All Cap Growth Strategy outperformed its Russell 3000 Growth benchmark during the third quarter, benefiting from strong stock selection as the unwind in AI momentum rewarded our more disciplined, valuation-sensitive approach. Stock selection contributed in seven of the nine sectors in which we were invested, led by IT, industrials and consumer discretionary, and much of the benefit came from what we did not own.

IT was the largest contributor to relative performance. Information security provider CrowdStrike led the way as several of our software holdings rebounded after coming under pressure in the first half of the year. The rebound helped security companies such as Palo Alto Networks and infrastructure software makers like Snowflake, which now appears positioned to benefit from the proliferation of AI agents. Micron Technology, a new position, also contributed, as our purchases came after much of the memory maker’s July decline. Semiconductor equipment makers Applied Materials, KLA and Lam Research, which had surged in the second quarter, fell sharply as crowded AI semiconductor trades unwound, and not owning them contributed meaningfully to relative returns.

Our underweights to Nvidia and Microsoft — a function both of our fundamental views of the companies and position size limits in a concentrated, evolving benchmark — were the largest detractors. Nvidia’s data center revenue continued to grow rapidly, while Microsoft’s shares were supported by continued Azure cloud momentum, encouraging Copilot usage data and signs of recovery in enterprise software during an optimistic period for AI monetization. Broadcom declined with AI semiconductors and after issuing a fiscal fourth-quarter revenue forecast slightly below expectations, and ASML, not in our benchmark, declined with the broader semiconductor capital equipment group as elevated expectations and uncertainty over export controls constrained upside, even as ASML’s second-quarter results were strong.

In industrials, stock selection more than offset the cost of our overweight to one of the benchmark’s weakest sectors. Not owning Caterpillar helped as growing public resistance to data center construction raised questions about the power generation demand behind its earlier rally. However, Axon Enterprise, a maker of TASER devices, body cameras and public safety software, detracted as second-quarter profitability fell and inventory rose.

In consumer discretionary, our underweight to Tesla helped as the stock fell sharply after second-quarter profit missed expectations on lower vehicle prices and higher AI spending. In communication services, Meta Platforms advanced following a strong reception to the launch of its Muse personal AI agent, whose early downloads surpassed those of the ChatGPT and Gemini debuts, reinforcing our view that agentic commerce is becoming a broader opportunity. Also in communication services, AppLovin, which operates a leading AI-powered platform for mobile application marketing and monetization, detracted as an expected step-up in model performance shifted from the second quarter into the third, leaving results short of the company’s own guidance.

"The selloff in AI-related stocks created attractive entry points to build positions in companies we had not owned." 

In health care, Natera, a leader in molecular residual disease (MRD) testing for cancer recurrence, reported an increase in quarterly MRD test additions and raising its full-year guidance, helping shares rise. Thermo Fisher Scientific, a fully active position, beat earnings, with resilient biopharma consumables and pharmaceutical product development end markets and healthy third-quarter guidance supporting the prospect of an accelerating business into 2027.

Our portfolio activity during the quarter was guided by two objectives: thoughtfully expanding exposure to AI and semiconductor infrastructure while maintaining appropriate risk following the June Russell reconstitution.

The selloff in AI-related stocks created attractive entry points to build positions in companies we had not owned. We initiated positions in Advanced Micro Devices, which offers differentiated exposure to high-performance computing, and Micron Technology, which holds a particularly strong position in DRAM and high-bandwidth memory. AI data centers require significantly more memory than traditional computing systems, and we see Micron as a differentiated way to gain exposure to the AI buildout while diversifying our semiconductor holdings. We also added Vertiv, which stands to benefit from increasingly power- and cooling-intensive data center architectures. Encouraged by a recovery in software, we added to our position in ServiceNow; our aggregate software exposure remains weighted toward security and infrastructure investments.

We continued to refine our mega cap exposures after the Russell index reconstitution, more than doubling our position in Alphabet and further reducing Amazon and Microsoft. To fund the additions, we exited Roche, a position where our conviction had diminished, and sold e.l.f. Beauty and Tractor Supply as decelerating growth and questions around long-term return potential weakened our confidence. We also sold Insulet on concerns about competition and execution.

The recent AI selloff has created a more constructive backdrop for our process. We remain fundamentally positive on AI and believe adoption is still early, but the market had moved quickly into many infrastructure names to capitalize on high expectations. A usage shift from proprietary to open-source models does not diminish the need for compute; in many cases, their broader use reinforces demand for the underlying infrastructure.

Times of change and uncertainty create opportunities for us as bottom-up stock pickers, as the market often has a habit of painting industries and subindustries with a broad brush, as it did when software was sold indiscriminately over much of the last 12 months. Classic measures of market breadth remain weak even as individual business fundamentals are sound, a combination we believe argues for selectivity. We would not be surprised to see bifurcation in the market as earnings reveal the ability for AI to either accelerate near-term business momentum or erode it, and our balance across sectors and the spectrum of growth leaves the Strategy well positioned to identify the winners and losers as AI’s implications broaden.

The ClearBridge All Cap Growth Strategy outperformed its Russell 3000 Growth Index benchmark during the third quarter. On an absolute basis, the Strategy had positive contributions from five of the nine sectors in which it was invested. The leading contributors were the IT, health care and communication services sectors, while the industrials sector detracted the most.

On a relative basis, overall stock selection contributed to performance, while overall sector allocation detracted. Stock selection in the IT, industrials, consumer discretionary, health care, communication services and materials sectors benefited performance. Conversely, an overweight to the industrials sector weighed on relative results.

On an individual stock basis, the biggest contributors to relative returns were CrowdStrike, Natera and Meta Platforms, as well as not owning Applied Materials and KLA. The largest detractors from relative returns were AppLovin, ASML and Broadcom, as well as underweights to Nvidia and Microsoft.

In addition to the transactions discussed above, we exited positions in Chipotle Mexican Grill in the consumer discretionary sector, Doximity in the health care sector, Builders FirstSource in the industrials sector and Vistra in the utilities sector.