Market Commentary and Fund Performance

Performance data quoted represents past performance; past performance does not guarantee future results. The investment return and principal value of an investment will fluctuate so that an investor’s shares, when redeemed, may be worth more or less than their original cost. Current performance of the fund may be lower or higher than the performance quoted. Performance data current to the most recent month end, and standardized performance can be obtained by viewing the fact sheet or by clicking here.

Market Highlights

In April 2026, the Japanese equity market staged a strong rebound, with the TOPIX, a representative index of Japan’s stock market, rising 8.19% month-on-month.

Market conditions were volatile at the beginning of the month, reflecting persistent concerns surrounding geopolitical tensions in the Middle East. However, sentiment improved as expectations for progress in ceasefire negotiations increased, allowing the market to stabilize and shift into an upward trend.

From mid-month onward, investor focus rotated away from geopolitical risks toward corporate earnings and structural growth themes, particularly in artificial intelligence (AI) related demand. Supported by strength in U.S. technology stocks and the Philadelphia Semiconductor Index (SOX), capital flows within Japan increasingly concentrated in semiconductor and AI-related names. As a result, the Nikkei 225—driven by high-priced, large-cap constituents—continued to trend higher and, notably, surpassed the 60,000 level at the close for the first time in its history.

In contrast, the broader TOPIX index lagged from mid-April onward, reflecting more subdued performance outside of select large-cap growth sectors. This divergence led to a further expansion in the performance gap between indices, with the “NT ratio” (Nikkei 225 divided by the TOPIX) reaching historically elevated levels.

Toward month-end, the market faced renewed headwinds. Rising interest rates—both domestically, amid a more hawkish outlook from the Bank of Japan (BOJ), and globally, as the Federal Reserve maintained a cautious stance on rate cuts—combined with a rebound in crude oil prices to weigh on equity markets. Profit-taking became evident, particularly in semiconductor-related stocks that had led the rally. Although there were intermittent rotations into domestic demand-oriented sectors, these were insufficient to meaningfully narrow the elevated performance gap between the Nikkei 225 and TOPIX. Nonetheless, Japanese equities closed the period with substantial gains.

The Fund’s Performance

During the month, the Fund (HJPIX) returned 6.22%, underperforming its benchmark, the Russell/Nomura Total Market™ Index, which returned 7.32%. The month's positive performer among the Global Industry Classification Standard (GICS) sectors included shares of Information Technology, Financials, and Industrials while Consumer Staples, Real Estate, and Health Care detracted from the Fund’s performance.

Among the best performers were our investments in SoftBank Group Corp., the telecom and Internet conglomerate, ORIX Corporation, Japan’s largest non-bank comprehensive financial services company, and Tokyo Electron Limited, one of the world's largest manufacturers of semiconductor production equipment.

As for the laggards, Seven & i Holdings Co., Ltd., a Japanese diversified retail group and operator of Seven-Eleven convenience stores, Mitsubishi Corporation, the largest trading company in Japan, and Sompo Holdings, Inc., one of the three largest general insurance company in Japan.

April Commentary

This month’s update focuses on five holdings—Tokio Marine, Shin-Etsu Chemical, ORIX, Seven & i and SoftBank Group—where recent developments have reinforced or refined our investment view.

Tokio Marine and Shin Etsu Chemical continue to demonstrate strong fundamentals supported by resilient earnings, cost advantages and structural growth drivers. At ORIX and Seven & i, recent initiatives point to improving capital efficiency and a stronger shareholder focus.

Against a rapidly evolving AI landscape, SoftBank Group reflects this shift clearly: while we have moderated our position amid rising competition, exposure to Arm and OpenAI may continue to offer potential for meaningful long term upside.

Tokio Marine Holdings

This month, we met with Tokio Marine Holdings, one of the major holdings in our portfolio, to discuss in greater detail its strategic partnership with Berkshire Hathaway, announced on March 23.

Our primary question was: "Why would Berkshire Hathaway—a renowned acquirer with vast financial resources—choose to collaborate with Tokio Marine on merger and acquisition (M&A) activities?"

Following our discussion, we interpreted the rationale for the partnership as follows:

Benefits for Tokio Marine

Potential for Berkshire to become a long-term major shareholder: Tokio Marine stated, "Berkshire Hathaway has agreed not to acquire more than 9.9% of our shares without prior approval from our Board of Directors." This mirrors the language used in Berkshire’s August 2020 announcement regarding its investment in Japan’s major trading houses, made when its stake crossed the 5% threshold required for a regulatory large shareholding report.

"Berkshire Hathaway’s intention is to hold its Japanese investments for the long term. Depending on price, Berkshire Hathaway may increase its holdings up to a maximum of 9.9% in any of the five investments. […] The company will make no purchases beyond that point unless given specific approval by the investee’s board of directors."

This suggests that Berkshire may be signalling an intention to purchase a significant stake in Tokio Marine, depending on its stock price.

Berkshire has already acquired roughly 2.5% of the outstanding shares, and there are potential signs of further accumulation. Specifically, we are monitoring its issuance of yen-denominated bonds. When investing in Japanese trading houses, Berkshire utilized proceeds from so-called Samurai bonds, allowing it to secure low-cost funding while eliminating currency risk. The total capital raised between January 2019 and November 2025 (2,160bn yen/$13.6mn) broadly matched its cumulative investment in trading house stocks (1,900bn yen/$11.9mn, per SPARX estimates). A further 272.3bn yen ($1.7bn) was raised in April 2026, and we believe these funds—together with the residual capital from preceding issuances— have been, or will be, deployed towards the purchase of Tokio Marine shares.

Accelerated growth through joint M&As: Since the late 2000s, Tokio Marine has expanded into a leading global non-life insurer through overseas acquisitions. Going forward, by sharing the financial burden with Berkshire, it gains the ability to pursue larger transactions previously beyond reach.

While Tokio Marine has traditionally targeted annual profit growth of 7–8%,¹ there is a possibility for this to be revised upwards in the forthcoming 10-year management plan scheduled for release in May.

Berkshire Hathaway’s reputation as a permanent owner of businesses: Berkshire’s reputation as a permanent owner enhances the partnership’s appeal to potential sellers—a meaningful competitive advantage. Typically, acquirers fall into two categories: strategic buyers and financial buyers. The former—often industry competitors—seek synergies and generally aim for long-term ownership, whereas financial buyers, such as buyout funds, usually look to exit after a certain period. The Berkshire–Tokio Marine combination bridges this divide: it combines long-term capital with substantial financial firepower. As a result, the partnership can compete effectively for large-scale transactions, where its scale may deter rival bidders and facilitate negotiations.

Tokio Marine already has the strongest balance sheet among Japan’s major non-life insurers. We believe that with scope to increase leverage, it is capable of pursuing deals in excess of 1tn yen ($6.3bn), positioning the partnership as a potential consolidator in the global insurance sector.

Enhanced shareholder returns: By co-investing in acquisitions, Tokio Marine reduces its individual capital commitment, potentially creating capacity to enhance shareholder returns—for example, through a higher dividend payout.

Benefits for Berkshire Hathaway

Long-term investment returns: We believe Berkshire can enjoy long-term equity returns by investing in Tokio Marine as its strategic partner.

Enhanced access to high-quality insurance float: The two groups have maintained a longstanding reinsurance relationship. We believe the partnership strengthens Berkshire’s access to high-quality “float”²— funds held between premium collection and claims payment—while providing geographic diversification into the Japanese market.

Leveraging Tokio Marine's M&A expertise: Berkshire’s insurance division (National Indemnity) operates with a remarkably lean structure,³ limiting internal M&A origination capacity. Partnering with Tokio Marine is intended to offer access to its established deal-making expertise.

Optimized capital allocation: By sharing investment commitments in insurance acquisitions, Berkshire could have increased flexibility to deploy capital into non-insurance businesses—a domain where Warren Buffett has historically demonstrated exceptional expertise.

The timing of the partnership is notable. The global non-life insurance industry has just emerged from an approximately eight-year “hard” pricing cycle,⁴ characterized by rising premiums. Early signs now point to a “softening” phase, in which pricing weakens—a typical feature of the industry’s cyclical pattern.

Historically, softer underwriting conditions weigh on incumbent insurers but create favorable conditions for acquirers, as valuations decline. Tokio Marine’s management had previously noted a scarcity of attractively priced opportunities during the recent cycle, whereas peers such as Sompo moved earlier, acquiring Aspen in the U.S. last autumn. These contrasting actions reflect the shifting underwriting cycle.

We expect the Tokio Marine–Berkshire alliance to focus on acquisitions of non-life insurers outside Japan and the U.S. Insurance is fundamentally a risk-underwriting business; it is therefore logical for the partnership to target regions where its current presence is more limited. Europe, in particular, stands out as attractive: it is home to insurers with long operating histories and deep expertise in niche lines—characteristics that align closely with Berkshire’s investment approach.

Finally, while it may seem far-fetched at this stage, it is worth considering—however speculative—a scenario in which Berkshire ultimately acquires Tokio Marine outright. Although Berkshire is often viewed as an investment vehicle, given its high-profile holdings in companies such as Apple and Japan’s trading houses, it remains, at its core, an insurance group. Over the past six decades, it has created substantial value by investing the “float” generated by its underwriting operations.

Greg Abel, who succeeded Warren Buffett as CEO, reinforced this point in his 2025 shareholder letter:

"Berkshire is a unique conglomerate, intentionally designed to allocate capital rationally and efficiently. Insurance is our core, and we also hold substantial investments in businesses across many other sectors."

Expanding float—the foundation of Berkshire’s investment capacity—requires both organic growth in underwriting and scale through acquisition. The group achieved rapid, non-linear growth in its insurance operations up to the 1990s via major transactions, including National Indemnity and GEICO. However, since acquiring the reinsurer General Re in 1998, Berkshire has largely avoided large-scale deals in & the sector.

While the General Re acquisition significantly expanded the scale of Berkshire’s insurance operations, it was followed by a period where lax underwriting standards surfaced, leading to a sharp decline in profitability and several years of restructuring. Although General Re has since become a core pillar of the group—now the world’s largest reinsurer—that experience may help explain Berkshire’s long-standing caution towards sizeable insurance acquisitions. Against this backdrop, it is plausible that Ajit Jain, who heads the insurance division, initiated the present partnership in the post-Buffett era as a more measured route to expansion.

In terms of scale, Tokio Marine is sufficiently large to have a meaningful impact on Berkshire’s insurance operations. Berkshire generated $84bn of insurance premiums in 2025 (across its three principal entities), compared with Tokio Marine’s net written premiums of 5.3tn yen (approximately $33bn) for the fiscal year ended March 2025.

Notably, while Japan imposes foreign ownership restrictions in sectors such as broadcasting, telecommunications, aviation and energy infrastructure on national security grounds, current legislation permits full foreign ownership of property and casualty insurers. If Berkshire’s long-term objective is indeed to deepen access to Tokio Marine’s float, the possibility of a full acquisition cannot be entirely dismissed.

Shin-Etsu Chemical

We initiated a position in Shin-Etsu Chemical in 2023. While it remained a mid-sized holding over the subsequent three years, we have increased our exposure this year following a reassessment of the long-term outlook for its core silicon wafer and polyvinyl chloride (PVC) businesses, in both of which it is the global leader by market share. In addition, its smaller segments may also offer attractive growth characteristics.

The rise in AI-driven wafer demand

Key growth drivers of wafers: The global surge in investment in AI infrastructure has driven strong growth in graphics processing units (GPUs), memory chips and fiber optics, with wafer demand now following.

Silicon wafers—the foundational substrate for semiconductors—are critical to end-product quality. Even minor surface imperfections render a wafer unusable during the lithography process. For leading-edge semiconductors, such as GPUs used in AI data centers, wafers must achieve “eleven-nines” purity (99.999999999%).

The explosion of AI demand is driving two positive shifts in the wafer market.

First, volumes are increasing. AI data center chips are trending towards larger sizes in order to maximize performance—in contrast to semiconductors used in consumer devices such as smartphones. Larger chips are more susceptible to defects, requiring a greater number of wafers to maintain acceptable yields. At the same time, increasing use of stacking and multi-layering in memory semiconductors is driving further growth in wafer consumption.

Second, the mix is shifting towards higher value-added products. Advanced applications increasingly require “epitaxial wafers,” which undergo specialized processing. Japanese manufacturers, with longstanding expertise in materials science, dominate the reliable supply of such products. Shin-Etsu Chemical and SUMCO together command a substantial share of this segment.

This reflects a broader pattern in Japanese materials manufacturing. High-value wafers require precise control of temperature, pressure and rotational speed of ingot pulling at the nanometre level. Similar strengths are evident in other sectors, such as high-grade automotive steel, where Japanese producers maintain a leading global position. More broadly, the country’s track record in fundamental chemistry—including numerous Nobel laureates—underscores the depth of its materials science capabilities. Breakthrough discoveries in materials science are often born from serendipity through trial and error with persistence and perseverance. Fundamental scientific research is hardly glamorous but Japanese people’s studious ethos is well-suited to this.

Competitive landscape: The wafer industry has undergone significant consolidation, evolving from a fragmented structure into an effective oligopoly. Today, the top five players account for more than 90% of global supply. This consolidation reflects both intense historical price competition and the scale of capital investment required for next-generation technologies, which forced weaker players to exit or merge. Within this structure, Shin-Etsu Chemical holds a leading position, with roughly 30% market share and the highest profitability in the sector.

Until about two years ago, we had been cautious on the potential for increased competition from emerging Chinese wafer manufacturers. However, that view has shifted. Taiwan’s trade data—a useful proxy given the presence of TSMC, the world’s largest semiconductor foundry—suggests that the growth of Chinese wafer shipments has stalled. It is likely that TSMC remains reluctant to adopt Chinese wafers for leading-edge processes. A similar pattern can be observed in SUMCO’s results: while sales to China declined sharply between 2022 and 2023, they have since stabilized. This indicates that the phase of import substitution by domestic Chinese producers may have largely played out.

Long-term agreements as a key characteristic of the business: A defining feature of the wafer industry is the widespread use of long-term agreements (LTAs). In response to past cyclicality—often referred to as the “silicon cycle”—the industry has moved towards contractual frameworks that aim to support both volumes and pricing, potentially helping producers to recover substantial capital expenditure. Shin-Etsu Chemical strictly adheres to a policy of "no investment without LTA backing." Such contractual stability is rare in other electronic components sectors (i.e. memory chips like DRAM, NAND, and ceramic capacitors like MLCC), where prices often fluctuate wildly based on supply and demand, significantly impacting earnings.

Management indicates that the company serves more than 30 customers globally in 300mm wafers, with the capability to supply approximately 4,000 different product specifications. This degree of customization represents a significant barrier to entry. As supply tightens in response to AI-driven demand, forthcoming contract negotiations should enable the company to pass through higher capital costs—driven by both volume expansion and increasing technological complexity potentially strengthening its earnings base.

Although silicon wafers account for only about 2% of total semiconductor manufacturing costs, their quality is the primary determinant of manufacturing yield. As a result, they are mission-critical inputs. With its technological edge, dominant market position, and stable contract-based model, Shin-Etsu Chemical is an essential player in the AI era.

Competitive edge of the PVC business amid geopolitical risk

Shin-Etsu Chemical’s PVC business is supported by high barriers to entry, most notably its manufacturing footprint in the United States.

Many global competitors rely on naphtha, a petroleum-derived feedstock. By contrast, Shin-Etsu uses low-cost ethylene sourced from U.S. shale gas. The company has also built a vertically integrated production structure, manufacturing both ethylene and its downstream intermediate, vinyl chloride monomer (VCM), in-house. This reduces transport costs and limits exposure to supply-chain disruption. The recently announced expansion of VCM capacity further reinforces this integrated value chain and strengthens cost competitiveness.

The rise in crude oil prices following the U.S.–Iran conflict in late February has, counter-intuitively, highlighted Shin-Etsu’s relative advantage. Heightened concerns over oil supply, driven by tensions around the Strait of Hormuz, have forced naphtha-based producers to cut output or raise prices. PVC prices in Asia have consequently risen sharply, providing a tailwind for Shin-Etsu.

Higher oil prices may also benefit the company from a cost perspective. Elevated crude prices tend to encourage increased oil drilling in the U.S., which boosts the supply of associated natural gas. While crude oil is readily exportable, natural gas requires liquefaction at very low temperatures, constraining export capacity. As a result, most U.S. natural gas is consumed domestically and inventories remain ample. Since the outbreak of the conflict, U.S. natural gas prices have declined. In contrast to the usual inflationary impact of Middle Eastern geopolitical risk on energy inputs, Shin-Etsu appears positioned to capture both higher realised prices and lower feedstock costs. There is therefore a reasonable prospect that earnings exceed current market expectations.

PVC is a staple industrial material, primarily used in housing construction such as water pipes and window frames. This exposure aligns well with the homebuilder investments currently in our portfolio (Sekisui House, Sumitomo Forestry, and Daiwa House). We view the current slump in the U.S. housing market as a "contrarian" buying opportunity. Given the shortage of housing stock and the projected growth in household formation, we believe there is significant upside potential for earnings as the market environment recovers.

Additionally, PVC demand is increasing as a construction material for AI data centers. Potential shifts in Chinese government policy that may erode the export competitiveness of Chinese-made PVC represent another positive catalyst, supporting Shin-Etsu’s global market share expansion.

Focus on broader AI-related products

Beyond silicon wafers, Shin-Etsu Chemical offers a broad portfolio of materials that support the evolution of AI. AI-related products now account for approximately 15% of total company revenue. At a recent investor presentation on this segment, the company highlighted both its advanced technological capabilities and the long-term earnings potential of these businesses.

For instance, the Isesaki Plant, which commenced operations in April, is expanding the production of advanced lithography materials essential for semiconductor manufacturing, particularly "middle- and under-layer hardmask films." The company holds a significant market share in these films, which play a critical role in protecting wafers as circuit patterns become increasingly miniaturized with the adoption of EUV (Extreme Ultraviolet Lithography).

Quartz cloth, which minimizes signal transmission loss in data center switches and server boards, is expected to become a future growth driver. A key competitive advantage lies in the company’s vertically integrated production of raw quartz and yarn, enabling superior quality control and supply stability compared to competitors.

Moreover, the company is actively expanding its portfolio of materials designed to address challenges arising from higher server performance and density. Development is accelerating in areas such as heat dissipation and cooling materials for servers, components related to co-packaged optics (CPO) for high-speed data transmission, and optical fiber preforms.

Rare earth development off Minami-Torishima Island

In recent years, the discovery of substantial rare earth-rich mud deposits on the deep-sea floor off Japan’s Minami-Torishima Island has attracted growing attention. Shin-Etsu Chemical appears well placed to benefit should commercial development proceed.

According to expert estimates, heavy rare earths—critical for high-performance magnets and other advanced technologies—account for around 50% of the total rare earth content in the Minami-Torishima deposits, compared with roughly 25% in China’s land-based reserves. Unlike many terrestrial deposits, which often contain radioactive elements and present waste-management challenges, deep-sea rare earth mud is free from such contamination, making it a comparatively cleaner resource. Confirmed reserves are estimated to be sufficient to cover several hundred years of Japan’s annual consumption, with the potential to position Japan as the world’s third-largest producer. Trials have also demonstrated that deep-sea extraction using “air-lift” technology can be economically viable.

Shin-Etsu Chemical has historically demonstrated strong technology and supply capabilities in the rare earth magnet business. The company already operates integrated production systems in Japan and Vietnam that encompass the entire value chain, from raw material separation and refining to material processing and recycling.

China currently holds a monopolistic share of rare earth production, representing a significant Achilles’ heel for the global economy. There is increasing momentum behind the development of the Minami-Torishima deposits and the establishment of domestic refining capacity in Japan. Shin-Etsu Chemical is among the companies best positioned to benefit directly from such a shift.

ORIX

On April 27, ORIX announced the sale of its consolidated subsidiary, ORIX Bank, to Daiwa Securities Group. We view the news positively for two key reasons. First, the sale price of 370bn yen ($2.3bn) is highly favorable. It represents a material premium to the bank’s net assets (250.2bn yen/$1.6bn as of September 2025), implying a price-to-book (P/B) multiple well above 1.0x. The implied price-to-earnings (P/E) multiple, based on the most recent full-year net income of 20.9bn yen ($131.4mn), is also above our estimate of fair value. Second, and more importantly, the decision signals a renewed sense of urgency within management to improve capital efficiency.

ORIX Bank had long presented a strategic challenge for the group, given its relatively low capital efficiency compared with other portfolio businesses. While the bank lacked the strong deposit-gathering capabilities of Japan’s newly listed digital banks, it nonetheless made a meaningful contribution to group earnings, generating close to 30bn in pre-tax profit. Divesting such a subsidiary inevitably creates an “earnings gap” that must be filled elsewhere, underscoring the degree of resolve required to proceed.

If the decision was led by Mr. Takahashi, who became chief executive last year, it would point to a clear determination to raise return on equity. Mr. Takahashi has articulated ambitious targets for 2035, including an return on equity (ROE) of 15% and net income of 1tn yen ($6.3bn). Central to this strategy is the expansion of the asset management business. As noted in previous commentaries, this marks a shift away from a balance-sheet-heavy principal investment model towards a more asset-light, fee-based earnings structure.

Should ORIX succeed in increasing the share of stable, recurring fee income, the market is likely to place greater weight on future earnings potential rather than balance-sheet value. In practical terms, this would imply a shift in valuation from P/B or net asset value (NAV)-based metrics towards a P/E framework. At present, ORIX trades on a P/E multiple of around 13x, well below the broader market average. We expect both earnings per share growth and a gradual re-rating of the multiple over time, supporting long-term share price appreciation. We will continue to monitor developments closely.

Seven & i Holdings

The company held its Investor Relations (IR) Day on April 23, where it outlined its medium-term management strategy and specific action plans. Despite this, the stock price has remained sluggish. However, we continue to take a more constructive view than the broader market and maintain a positive outlook. We believe the stock is undervalued and represents an attractive investment opportunity for the following reasons:

1.    Significant growth potential: The convenience store business remains a growing industry globally. The company has substantial room for expansion in the U.S. market, alongside further opportunities for new store openings in many untapped geographies.

2.    Recovery signs: There are emerging signs of a sales recovery in domestic existing stores, while a full-scale turnaround effort is now underway in the U.S. business.

3.    Heightened management urgency: Driven in part by activist pressure in recent years, the current management team displays a significantly higher level of urgency compared to the Isaka era (the previous CEO).

4.    Strong cash generation: Even during the past three years of stagnant operating performance, the company has generated approximately 500bn yen
($3.1bn) in free cash flow (FCF). On current market capitalization, this equates to a free cash flow yield of around 10%, which is exceptionally attractive.

5.    Compelling valuation: The stock is significantly undervalued based on forward P/E projections of only 11.5x (we believe it is appropriate to base our calculations on the company’s forecast EPS of 162.56 yen ($1.02), adjusted for goodwill amortization, which we regard as a more appropriate measure of the company's underlying cash earnings power).

6.    Enhanced shareholder focus: Since the Alimentation Couche-Tard acquisition proposal fell through, management has pivoted towards a more shareholder-centric approach and has been actively executing share buybacks. This not only boosts EPS but also provides a strong floor for the stock price through improved supply-demand dynamics.

7.    M&A potential: Although Couche-Tard has withdrawn, the possibility of a new suitor emerging remains. Given the current depressed valuation, the company may be considered attractive to financial buyers.

While the turnaround is taking longer than we initially anticipated, we believe the company is steadily moving in the right direction. In particular, the new executive compensation scheme announced on April 16 represents a meaningful improvement, better aligning management incentives with shareholder outcomes. For these reasons, we will maintain our current position.

SoftBank Group

SoftBank Group’s shares rebounded sharply this month. While we have not commented in detail on the company or developments in AI frontier models since our October 2025 commentary, the rapid shifts in the competitive landscape have prompted us to reduce our position. The environment has shifted beyond our initial expectations, and OpenAI’s prospects no longer appear unassailable as they did a year ago. Although the outlook remains mixed, it is clear that the gap between OpenAI and its closest competitors has narrowed since last summer, when we aggressively increased our exposure.

Despite this adjustment, we continue to hold a stake in SoftBank Group for two primary reasons. First, the growth outlook for Arm. SoftBank retains an approximate 90% stake in Arm, the UK-based semiconductor design company. Arm’s power-efficient, high-performance CPU architectures are widely utilized across fields ranging from smartphones to data centers. Demand is now being driven increasingly by the proliferation of AI agents. While GPUs have historically underpinned AI infrastructure, the growing importance of agents—where AI performs tasks on behalf of human users, ranging from reasoning and planning to web browsing, software tool operation, and decision-making—is likely to shift demand towards CPUs, which are better suited to sequential processing.

Second, the evolving competitive dynamics in AI models. Competition has intensified materially since the release of Google’s Gemini 3.0 last autumn, followed by rapid iteration in Anthropic’s Claude models. In some areas, OpenAI’s ChatGPT series no longer maintains a clear lead and, at times, appears to lag. Anthropic’s progress is particularly evident in coding and enterprise use cases. However, the increasing adoption of AI agents may alter the basis of competition once again. Historically, end-users selected models according to specific requirements. Going forward, agents are likely to select models dynamically based on task optimization. This implies that model providers must focus on being “selected” by agents rather than users directly—an approach analogous to search engine optimization (SEO)⁵ in the early internet era. OpenAI may yet regain an advantage under this framework.

Recent developments offer a more balanced picture. Initial reception to ChatGPT 5.5 has been positive, particularly in terms of functionality, and its availability on Amazon’s AWS Bedrock platform, alongside Microsoft Azure, should broaden distribution and support revenue growth. At the same time, OpenAI’s previously aggressive investment in computing capacity now appears better justified. By contrast, Anthropic’s more conservative approach has, in some instances, constrained its ability to meet demand despite rapid technical progress. Taken together, it remains premature to identify clear long-term winner(s) in this space. Nonetheless, we still believe that SoftBank Group's position as a major shareholder in OpenAI—a central player in cutting-edge AI—is a positive factor.

Click here for a full listing of Holdings.