Commentary: Dividend Strategy

• Markets rebounded sharply driven by narrow gains within the semiconductor sector.
• We took advantage of surging stock prices to remix our semiconductor holdings, gaining exposure to GPUs and cutting-edge production while broadly maintaining our overall sector weighting.
• With record market concentration, diversification is paramount. We have measured investments in many companies at the heart of the AI ecosystem, but we also are focused on finding overlooked gems in other corners of the market.

After a challenging first quarter that saw the S&P 500 Index decline 4.3%, markets rebounded sharply in the second quarter. The S&P 500 climbed 15.2%, driven by a surge in information technology (IT) stocks, particularly semiconductor stocks, which soared 44.0%. The ClearBridge Dividend Strategy rose nicely but lagged the exuberant gains of S&P 500 mainly due to our lower technology exposure.

While no one likes to trail in sharp market rallies, diversified dividend growth portfolios tend to lag in these narrow, momentum driven markets. Our research shows, however, that these periods tend to be followed by periods of broader market participation in which diversified dividend-growth portfolios shine.

Semiconductor memory stocks were the standout performers. Shares of Micron, the largest producer of storage chips, skyrocketed 240%, adding over $900 billion of market cap — more than the entire market capitalization of Exxon Mobil, Johnson & Johnson or JPMorgan Chase. Our underweight to Micron and a select group of other non-dividend-paying data center plays, like Advanced Micro Devices, Intel and SanDisk, were the largest drivers of our relative underperformance. We previously owned Intel but sold it in 2024 when the company eliminated its dividend.

While the market’s leading contributors were concentrated in IT, our top contributors were diffuse. Shares of CVS Health rose as earnings recovered significantly in its health insurance segment. Public Storage shares ascended as rental trends bottomed and the company announced a compelling acquisition of National Storage. MetLife’s stock gained due, in part, to improved variable investment income.

Our underweight to IT reflects a long-standing emphasis on broad-based diversification. We have rarely had more than 20% in one sector, whereas IT currently represents 38% of the S&P 500 — the highest a single sector has ever represented in the composite. Healthy markets are typified by broad stock participation, whereas recent gains (Exhibit 1) have been very narrow.

Exhibit 1: U.S. Market Breadth is Narrow

Despite the crucial role computer chips have played in technological advancement over the last 40 years, they have not typically been more than 4% of the S&P 500. In the last few years, and particularly the last few months, however, semiconductor equities have come to dominate the U.S. stock market (Exhibit 2).

Exhibit 2: Semiconductor Weighting in the S&P 500 Index

With insatiable demand for chips far outstripping supply, pricing and profitability have soared. Pricing for DRAM (storage chips) is up 900% in less than a year (Exhibit 3). Operating profits for the semiconductor sector are up 2,000 basis points. As the charts below vividly depict, we are in unprecedented territory.

Exhibit 3: DRAM Spot and Contact Prices

Exhibit 4: Semiconductor Sector Operating Margins

Record current profitability explains the sector’s recent earnings windfall and stock market gains, but it is irrelevant to its forward-looking prospects. Long-term business and stock performance will be driven by future profitability, and nobody knows how that will unfold. Competitive industries operating at cyclical peaks tend to see profitability revert toward long-term averages as rising prices destroy demand and production capacity expands to bring the market back into balance.

In June, Apple raised prices for many of its products by 20% to protect its profitability as soaring storage prices inflate its input costs. With the 800-pound gorilla leading the way, other electronics companies are sure to follow suit. Higher prices for consumers will crimp sales of tablets, computers, etc., ultimately reducing demand for storage chips. On the supply side, Micron plans to spend $50 billion next year to expand capacity — twice what it spent this year and six times what it spent two years ago. As that new capacity comes online, supply constraints should ease.

"Competitive industries operating at cyclical peaks tend to see profitability revert toward long-term averages." 

Over the last three years, hyperscaler¹ annual capex has quadrupled to north of $600 billion. Do such staggering sums reflect the surge of AI’s initial buildout or the new trendline? Could technological advancements change the way AI works and reduce the demand and/or the cost of chips? Will the AI business model evolve profitably enough to generate returns sufficient to justify the trillions of dollars being spent, or will disappointing returns lead to reductions in capacity additions? These are big questions to which the market does not yet have answers.

We currently have about 9% of the portfolio in semiconductor-related stocks. We believe this gives us enough exposure to benefit from the wonderful growth that chip stocks enjoy without overly exposing the portfolio to a highly volatile sector currently trading at cyclical highs. During the quarter we initiated a meaningful position in Nvidia, after the company announced a 25-fold increase in its dividend, taking its yield from almost zero to 0.5%. We also increased our exposure to Taiwan Semiconductor, which, as the largest producer of computer chips in the world, should prosper regardless of whose chips are used for AI computing. We funded these purchases, in part, through pruning Broadcom and Texas Instruments, both of whose stocks soared during the quarter and were strong contributors. By remixing our semiconductor holdings, we gained exposure to GPUs and cutting-edge production, while limiting our overall exposure and reducing positions in equities that have been substantially revalued higher.

With investors myopically focused on AI, we are finding terrific opportunities in other areas of the market. Indeed, our outlook for the portfolio is the most constructive it has been in years. We recently built significant positions in ADP, Blackstone, Marsh & McLennan and Otis and added materially to positions in Apollo, Haleon and Unilever. These are fabulous companies with robust strategic positions and excellent financial characteristics, and we bought them at compelling absolute and relative valuations.

In recent years ADP, Blackstone, Marsh & McLennan and Otis have traded at large premiums to the market because of their quality (they average a return on invested capital of 30% versus the market’s ~10%), but today the market is making no distinction between these high-quality businesses and the average business (Exhibit 5). Not since the Global Financial Crisis have we bought so many terrific companies at such compelling prices.

Exhibit 5: Exuberant Market Not Rewarding High-Quality Businesses

The case for the ClearBridge Dividend Strategy is as strong as ever. Our portfolio delivers an attractive upfront yield, robust dividend growth and conservative risk management. With record market concentration despite uncertainty introduced by profound technological change, diversification is paramount. We have measured investments in several of the companies at the heart of the AI ecosystem, but we also are focused on finding overlooked gems in other corners of the market. We believe, the seeds we are planting today will bear fruit that we will be harvesting for years. In a world where change is increasingly constant, our steadfast approach to diversified portfolio management should serve us well.

The ClearBridge Dividend Strategy underperformed its S&P 500 Index benchmark during the second quarter. On an absolute basis, the Strategy saw positive contributions from nine of 11 sectors: the IT and financials sectors were the main contributors, while the energy and utilities sectors detracted.

On a relative basis, stock selection and sector allocation detracted. In particular, an underweight to IT and overweights to energy, materials, consumer staples and financials detracted; stock selection in industrials, IT and communication services also weighed on relative results. Conversely, stock selection in energy, consumer staples and materials along with an underweight to consumer discretionary proved beneficial.

On an individual stock basis, the main positive contributors to relative returns were Texas Instruments, CVS Health, Taiwan Semiconductor Manufacturing and not owning Netflix and Chevron. The main detractors were not owning Micron Technology, Advanced Micro Devices and Intel, and positions in Exxon Mobil and T-Mobile.

In addition to the transactions mentioned above, we received shares of Honeywell Aerospace in the industrials sector following its spinoff from holding Honeywell International. We exited Diageo in consumer staples, DTE Energy in utilities, L3Harris Technologies in industrials, UnitedHealth in health care and Freeport-McMoRan in materials. We also exited the small position in Waters we had received following its spinoff from Becton Dickinson in the first quarter.