Recently, when asked to describe the current market, I didn't hesitate: an elastic band stretched so far beyond its natural length that the only question is not whether it snaps back, but when — and who is holding it when it does.
In the last two months only 10 of 33 sectors in Japan beat the market. Leadership was concentrated in AI infrastructure — Taiyo Yuden, Murata, Kioxia, Furukawa Electric, SUMCO — names we have avoided on valuation grounds. Many of these stocks pay little or no dividend: Kioxia pays none at all; SUMCO's dividend of ¥10 per share represented a yield of just 0.1% at its recent lows. For a fund built around income, capital discipline, and low volatility, they are not candidates at the moment. The Japanese market behaved, in effect, as a geared expression of US semiconductor capital expenditure. That hurt.
It is worth pausing to set Japan in its wider context, because what has happened here has happened everywhere. The largest seven companies in the United States now account for roughly a third of the S&P 500. Widen the lens to the world index and the picture does not so much diversify as repeat itself: the US is close to two-thirds of the MSCI All-Country World Index, and the same handful of names that dominate Wall Street sit at the top of the global index too. An investor seeking broad exposure finds themselves, in practice, making a very concentrated bet — and Tokyo’s recent sprint was simply the local chapter of the same story, written in yen. This is the context in which our lower beta and our deliberately different portfolio should be understood: not as a drag on returns in normal times, but as a considered refusal to load up on the single most crowded trade in global markets.
We are up +9% to mid-June — which, set against our three-year return in sterling of just under +70%, is no cause for alarm. What this reflects is eight weeks of discomfort — and eight weeks in a forty-year career is not cause for concern, but for explanation — which we hope this letter provides. With an ex-ante beta of 0.86, we will lag in narrow, momentum-driven rallies of this kind, as the past two months have demonstrated, but are positioned to preserve capital and recover that ground, with interest, when those rallies unwind, as every comparable episode over the past twenty-five years has shown.
Now let me tell you why I am very relaxed. I have been here before — many times.
The Valuation Picture
The valuations now being ascribed to the winners are not just full, they are extraordinary. More importantly, they are all telling the same story: that AI infrastructure spend is infinite and immune to the business cycle, with no margin for disappointment. The AI infrastructure segment of the Japanese market now trades at price-to-book multiples that would have been unthinkable a decade ago: as an example, Kioxia trades at 42x book value; Murata, which twenty-three years ago traded at 2x book, now trades at 8x and Advantest’s book value stands at 28x, well above its 10 year median of 5.4x. These are not growth multiples; they are perfection multiples. Whether one looks at price to book, price to earnings, or implied growth expectations, the conclusion is the same: they are ultimately priced for perfection.
The Macro Backdrop
Investing in Japan today is not simply about stock selection, it is also about the backdrop. Fiscal and monetary settings remain unusually loose relative to inflation. Strip out intra-government holdings and the underlying debt burden is meaningfully lower than generally presented; net debt is closer to 80–100%. The Bank of Japan is behind the curve in a way that is reminiscent of 1989 — and that comparison does not end well. Markets are pricing a terminal rate of 150bps. I believe it should be 250–300bps. Net yen shorts are back to August 2024 levels. When the yen corrects — and it will — the most geared plays on offshore earnings and overseas semiconductor demand will correct the hardest. Every prior episode of yen strength has rewarded quality names. That was 2000. That was August 2024. That will be this cycle too.
Where we are Winning — and where we will
The companies we own are boring in the best possible sense: cash-generative, reliable, well-run, high market share, cash-rich — and trading at fractions of the valuations being accorded to momentum names.
The numbers that describe our portfolio today are, I think, quietly extraordinary. We trade at around 1.0x book, against a market closer to 1.8x — and in a market where parts of the index trade at 8x, 28x, and even 40x book, that gap is not a rounding error. On earnings, we are at roughly 13x forward, a discount of nearly a third to the market’s 18x — yet these are not cheap businesses in any pejorative sense. Several are growing earnings meaningfully; the market has simply not noticed, because there is no analyst coverage, no momentum narrative, no tickertape. Then there is the balance sheet. Latent assets — investment securities, land, net cash — represent around 55% of the portfolio’s market capitalisation. For every £1 invested, roughly 55p is already backed by identifiable assets before anyone has placed a value on the operating business at all. These are compounding engines that need no external validation from a macro theme. The elastic band, in other words, is being stretched from both ends: the index is reaching in one direction, and our companies — quietly, without ceremony — are building value in the other.
The Setup
In 2000, the broader market was the place nobody wanted to be. If it wasn't Silicon Valley, it wasn't interesting. Then the Nasdaq fell. Then tech fell. Then the yen strengthened. The elastic band snapped back, and the boring companies generating cash, trading below book, overlooked by every sell-side desk in Tokyo — delivered.
The setup today is remarkably similar.
I am not asking for patience based on hope. I am asking for patience based on history, valuation, and a macro backdrop that is — beneath a very serene exterior — building toward a significant inflection. The yen is at the epicentre. When it moves, the trade that has hurt us will reverse sharply, and the companies we own will be reappraised.
The elastic band always snaps back. It just needs time — and something to snap back to. We have that.
250626.281
