Alex Bowles from our offshore partner, Orbis, discusses how the Orbis Optimal SA Fund is finding opportunities in overlooked Japanese companies as AI-driven excitement continues to concentrate investor attention on a handful of stocks.
The Orbis Optimal SA Fund is built to seek returns above cash by owning carefully selected stocks, while hedging away the bulk of its stock market exposure. Returns are therefore driven less by the market's direction and more by how the businesses we own fare relative to the regional markets we hedge against.
Japanese equities have fared spectacularly. Over the last 12 months, the TOPIX is up 43%, and the Nikkei 225 has climbed by 76% as several forces have converged. Corporate governance reforms have attracted growing interest from foreign investors. After decades of deflation, inflation has returned, encouraging companies to deploy idle capital more productively. Enthusiasm for artificial intelligence (AI) has spread beyond Silicon Valley and into Japan.
Even as Japan's stock market has enjoyed one of its best starts to the year in decades, only a quarter of stocks in the TOPIX and one-fifth of stocks in the Nikkei 225 have delivered index-beating returns year to date. Last quarter, we wrote about placing two different bets when buying the S&P 500. Applying a similar lens to Japan, when buying the Nikkei 225 index three years ago, roughly 40% of your capital went to 10 companies trading at a weighted average of 33 times forward earnings, with the remaining 60% spread across 215 companies trading at 22 times. Today, that same 40% exposure sits in just four companies trading at 48 times, while the other 60% still trades at a much lower 26 times. The advance has created an unusually narrow yet powerful market, propelled by a small cluster of stocks, many of which are tied directly or indirectly to AI. The poster child is Kioxia – a memory chip manufacturer that was spun out of Japanese industrial giant Toshiba – which has overtaken Toyota as the country's most valuable listed company, up a remarkable 55 times since its initial public offering in late 2024.
We have seen environments like this before. Investors become excited about a powerful theme, capital flows towards a relatively small group of perceived winners, and an increasingly large part of the market is left behind. The winners continue to win, attracting more capital and reinforcing the cycle. The stocks outside these favoured areas struggle to keep pace. Many high-quality Japanese businesses now trade at valuations that imply they are AI losers, or at best, irrelevant to the technology-driven future that investors are trying to price. In our view, this framing is often too simplistic.
Our contrarian philosophy means that we are comfortable being early, so long as we are paying the right price. What we are not comfortable with is paying too much or simply following the crowd. We do not dispute the transformative potential of AI, but when expectations rise faster than earnings power, we do question whether investors are being adequately compensated for the risk of being wrong, or merely less right than the market expects. In this environment, we are finding more attractive ideas in higher-quality Japanese businesses that the market has temporarily fallen out of love with – where we believe short-term noise is obscuring the long-term signal.
Few companies' products are as universally loved as those of Nintendo, a company that has been handed the "AI loser" label, with more dramatic consequences. Its share price has halved since peaking in August. We have used the sell-off to build Nintendo into the largest Japanese holding in the Fund. The market's reaction has centred on AI, with investors worrying both about near-term cost pressure and longer-term disruption to the gaming industry. In our view, both concerns are overblown.
The more immediate pressure comes from memory prices. As hyperscalers absorb supply to build out AI infrastructure, Nintendo's input costs have risen, and near-term hardware margins are likely to be squeezed. The pressure is real and will show up in the numbers. But what the market seems to be treating as a structural shift in Nintendo's economics looks to us more like a cyclical squeeze with management retaining multiple levers to pull over time.
The broader worry is that AI will democratise game development, lowering barriers to entry, flooding the market with compelling content and eroding Nintendo's competitive advantage. We think this view misunderstands where Nintendo's moat lies. Mario, Zelda, and Pokémon are well-made games, but they are also cultural institutions, built over decades of creative iteration, with multi-generational fan bases and genuine emotional attachment. Nintendo's advantages stretch beyond game development to its intellectual property, its characters and the trust it has earned through half a century of delivering exceptional experiences.
Nor is this the first time technology has been declared an existential threat to Nintendo. Piracy, smartphones and subscription streaming have each been seen as potential disruptors. Each time, Nintendo was written off. Each time, it was premature, as the company adapted and innovated.
While quick to price in risks, the market seems to be overlooking the opportunities. We believe Nintendo's intellectual property remains significantly undermonetised beyond gaming. The Super Mario Bros. Movie, which generated over US$1.3bn at the global box office, showed the commercial power of its characters outside the console. Theme parks, film, television and licensing remain meaningful, underpenetrated earnings streams. Meanwhile, the Nintendo Switch 2 cycle is still in its early innings, with a rich software pipeline ahead.
Nintendo illustrates a broader pattern of Japanese businesses with strong fundamentals, proven track records and clear long-term potential trading at valuations we see as too good to ignore. Against the extraordinary moves in AI-related stocks, some of our Japanese holdings can look comparatively unexciting in the short term. But share price momentum and business value are not the same thing.