A Japanese "Shinkansen" (or bullet train), traveling through the Tokyo cityscape at dusk.

Japan is a country that remains steeped in tradition and ritual, even as it embraces and leads advanced technologies such as factory automation, semiconductor production equipment and high-speed trains. Staid practices such as invoicing expenses via fax machines, saving data on floppy disks and even signing documents with physical ink stamps continued unabated until the pandemic forced a wholesale rethink.

Historical context: System integrators

Japanese companies’ approach towards IT infrastructure differs fundamentally from their American and European counterparts. In the 1960s, the Japanese government was concerned about IT competitiveness against American behemoths, IBM and Intel. Therefore, the government funded the development of IT national champion NTT, as well as three other IT groups, Fujitsu and Hitachi, NEC and Toshiba, as well as Mitsubishi Electric and Oki. It also awarded these groups public projects over the decades since. By the 1980s, the private sector saw the spinoff of consulting subsidiaries, specializing in the IT needs of service sectors such as e-commerce and finance. These consultants became known as System Integrators (SIs).

Competitive advantage: Talent monopsony

SIs coordinate software vendors, hyperscalers, subcontractors and non-tech companies’ IT departments to meet their clients’ IT needs. SIs’ customer stickiness is strong because clients desire customization but can’t secure the top IT talent because of customers’ comparatively low salaries. Local companies’ IT workers are generalists who don’t know how to effectively procure hardware, manage software or even develop an IT strategy. Gartner found that 67% of such companies blamed « talent scarcity » as a major obstacle to IT upgrades (vs. 38% globally). With growing IT labour shortages and few students pursuing tech degrees, SIs’ core role between the key parties is what leads to higher margins, enabling companies to hire top SI talent.

Industry outlook: Long growth runway

IDC estimated that Japan’s $180 billion in annual general IT spend would grow at 6.4% CAGR into 2029E. Mordor Intelligence estimated that cloud spending would grow significantly faster than general at 17% CAGR into 2031E. According to Gartner, in 2021 31% of Japanese companies stored data on the cloud, with cloud comprising only 4.3% of total IT spend (vs. 14.4% North America, 9.7% Europe, 6.4% China). As of 2023, according to the Information Technology Promotion Agency, large Japanese firms with more than 1,000 employees had already drawn even with large American firms with ~63% of them noting that they had dedicated digital transformation (DX) departments (vs. ~64% for large American firms). In contrast, smaller Japanese firms with fewer than 1,000 employees were lagging behind with just ~12-41% reporting dedicated DX departments (vs. ~39–66% for smaller American firms). Smaller capitalization SIs serve small customers.

Gen-AI: More opportunity than threat

While software-as-a-service (SaaS) company stocks have sold off across America, Europe and Japan year to date, we expect strong demand for cybersecurity and infrastructure to continue, benefiting SIs. This is because declining software development costs amid AI-led coding and fiercer price competition against AI agents reduce overall software package costs. While lower prices hurt SaaS supplier margins, they boost customers demand.

SIs are crucial to the integration of software packages with hardware and networks, all safeguarded by cybersecurity. Japanese companies’ core IT systems were built by the SIs themselves in complex layers based on evolving business needs and characteristics. This makes it hard to standardize processes, a necessary precursor to an AI-first automated approach. Rather, our SIs will even benefit from rising demand for limited IT system standardization as companies seek to deploy agentic AI. Admittedly, agentic AI has the potential to replace end-user applications in enterprise resource planning, but we believe that SIs will retain their crucial role in maintaining infrastructure by offering cybersecurity.

DX favours smaller SIs

Mentioned above, DX refers to the implementation of digitalization through efforts such as transitioning data to the cloud to avoid reliance on onsite physical data storage and, more recently, rolling out gen-AI models to boost productivity. The term captures the shift in approach from treating IT as peripheral toward recognizing its centrality. As IT competitiveness and DX continue in Japan, the next leg of growth should be led by DX service providers that focus on smaller firms.

Simplex Holdings

Simplex Holdings Inc. (4373 JP) was founded in 1997. In 2001, it began offering banks with solutions like IT consulting, systems development, and operations and maintenance. Over the decades, it expanded into FX brokerages, equity, futures, options platforms, insurers and crypto. In 2013, it conducted a $211 million buyout with Carlyle. Carlyle later sold its equity stake upon Simplex’s September 2021 relisting on the Tokyo Stock Exchange. We feel that Simplex is well positioned to benefit from this trend.

* all dollar amounts referenced in this article are in USD.

The historical Fed would be shifting to a tightening bias in response to recent economic news, according to a simple model.

To recap, the model classifies the Fed as being in tightening or easing mode depending on whether a probability estimate is above or below 0.5. The estimate is based on currently reported and lagged values of annual core PCE inflation, the unemployment rate and the ISM manufacturing delivery delays index. Despite the small number of inputs, the model does a satisfactory job of “explaining” the Fed’s past actions.

The model reading moved below 0.5 in August ahead of the September / October rate cuts, falling further in December, when the Fed delivered another reduction while signalling an expectation of additional moves in 2026 – see chart 1.

Chart 1

US Fed Funds Rate & Fed Policy Direction Probability Indicator

The reading, however, rebounded to around neutral in January and has climbed above 0.7 in February. The turnaround has been driven by a combination of a fall in the unemployment rate from 4.54% in November to 4.28% in January, a rebound in the ISM deliveries index from a November low and slightly firmer annual core PCE inflation (3.0% in December).

Additional data points for all three series will be available before the March FOMC meeting.

The January model shift is consistent with minutes of last month’s meeting, showing a strong consensus in favour of a hold with “several” participants viewing interest rate risks as two-sided.

A high-performance personal computer displaying a modern video game in a room illuminated by futuristic neon lighting.

The rapid repricing of global gaming equities year to date reflects a sharp narrative pivot in the market, hitting the stocks of portfolios holding Tencent, as well as other leading players such as Nintendo and Roblox. Only months ago, consensus held that AI would be an operational tailwind for game developers through cost reduction and faster content generation.

AI enhancing content production and experience
Image illustrating the different ways that AI can enhance gaming content production and the in-game experience.
Source: Tencent Investor Relations 2026

Following new AI model releases such as Anthropic’s Claude Code and Google’s Project Genie, the prevailing fear is that AI will disrupt traditional game development entirely.

The question for investors is whether Tencent, as the world’s largest gaming company by revenue, is positioned to benefit from or be impaired by this shift.

AI as a development tool

Tencent’s core strength is scale – both financial scale and model training scale. In discussions with management late last year, they emphasised that AI is already deeply embedded in their workflow: procedural content generation, NPC behavioural modelling, art and animation tooling and faster iteration cycles. These capabilities are not theoretical; Tencent purchases more AI compute and silicon than nearly any other company in Asia, outside hyperscalers.

Small studios will indeed be empowered by AI, lowering entry barriers and enabling “one hit wonder” creators, much like YouTube transformed video production.

However, distribution, marketing and IP longevity remain durable moats. Tencent excels in all three. Owning evergreen franchises – over 80% of its portfolio – means that even if development costs fall, the value of recognised IP rises.

Timeline showing the years different evergreen game properties of Tencent were introduced. Logos of Tencent owned studios, invested external studios and external partners are displayed to show Tencent's network.
Source: Tencent Investor Relations 2026

AI makes content easier to create, but not easier to distribute at scale, monetise efficiently or ensure regulatory compliance – areas where Tencent’s ecosystem advantage is overwhelming.

Golden age of movie studios gives way to more atomised content creators – parallels?

Consider the shift from studio dominance in Hollywood to a more atomised creator economy. AI could indeed enable a long tail of nimble game creators, just as digital tools transformed music and film production. If so, Tencent’s role may shift toward that of a global distributor and platform – akin to Netflix in video or Spotify in music.

But unlike movies, gaming economics rely heavily on ongoing monetisation: loot boxes, in-game economies, battle passes, skins and continuous seasonal content. Even if AI reduces production costs, developers with large user bases can simply retain the value by expanding monetisable content. Consumers rarely pay less – they typically pay more in more immersive and interactive environments. Tencent’s superior ability to drive retention and average revenue per user works in its favour.

Fear premium

Tencent today trades at ~16x PE with mid-teens EPS growth, and minimal risk to near-term earnings. This is historically inexpensive for a high-quality global IP and distribution engine. The derating reflects uncertainty over future industry economics – not current fundamentals.

The key debate is not whether Tencent gets disrupted this year (unlikely), but whether AI compresses long duration returns on capital for AAA studios globally.

Markets are trying to reprice the terminal value of moats like content creation and distribution.

Our view: Tencent is better positioned than most

AI will shift value around the gaming ecosystem. Some of that may move to consumers, some to new AI native studios and some to distributors. But scale matters. IP matters. Distribution matters. And Tencent is uniquely advantaged in all three.

The company may face multiple compression as investors debate the long-term competitive dynamics, but fundamentally, Tencent is more likely to be a beneficiary of AI than a casualty. The path will be volatile, but the structural advantages remain intact.

Trump’s trade doctrine: Opening the door to higher-return industrial champions

Last month, I spoke with Benefits and Pensions Monitor about the short-term noise generated by Trump tariff headlines. We explored whether investors should be looking through the noise based on the TACO (Trump Always Chickens Out) view that the US president will retreat in the face of market revolt.

My argument was that, while amusing, TACO risks obscuring the unmistakable direction of travel. US trade policy signals a shift to a multipolar world, defined by a US centric economic sphere and a China centric one, each with competing supply chains, industrial priorities and strategic alliances. Tariffs are signals of tectonic shifts in global trade.

Our view is that these shifts bring risks, but will also be a durable source of opportunity for EM investors.

When China is taken out of your supply chain, everyone makes money

Traditionally, sectors like shipbuilding, industrial machinery, energy logistics and specialty manufacturing have been deeply cyclical with limited pricing power. They lived and died by freight rates, commodity cycles and economic growth. But the combination of US reindustrialisation, reshoring and decoupling from China is transforming these industries.

Historically commoditised, price‑taking businesses are now at the heart of national security and industrial policy. Reindustrialisation and rebuilding supply chains have the potential to drive visibility, margins and returns on capital that would have been unthinkable a few years ago.

For example, the order books of Korean shipbuilders are increasingly less shaped by commercial shipping cycles, and increasingly by long‑cycle defence, LNG infrastructure and government‑aligned industrial programmes across the United States and its allies.

Major opportunities ahead for Korean shipbuilders in LNGC and naval vessels
Bar graphs illustrating Korean shipbuilders decreasingly affected by commercial shipping cycles.
Source: CLSA, Clarksons

US-China decoupling has effectively removed Chinese yards from security‑sensitive projects, structurally elevating demand for non‑Chinese capacity.

For countries aligned with US industrial and security priorities – Korea, Japan, India, parts of ASEAN – select industries have the potential to enjoy rerating as increasingly strategic rather than cyclical businesses.

This extends far beyond shipbuilding. We are seeing it in components for AI data centres, grid and power equipment, strategic metals, defence, electronics, energy infrastructure and advanced manufacturing.

These sectors are beginning to enjoy the boost of multi‑year, policy‑backed spending.

Many EM countries offer the scale, labour force depth and geopolitical neutrality that global supply chains now require. As the world bifurcates, EM manufacturers, suppliers and logistics operators are becoming essential nodes in both the US and China spheres. This creates a long pipeline of opportunities in markets that historically suffered from volatility and low returns.

In short, Trump’s trade doctrine accelerates a global realignment that raises the return potential of industries previously stuck in low‑margin cycles.