The macro story for the month (and so far in September) was, of course, the continuing rise in global bond yields with the US 10yr at 4.8%, approaching the closely watched 5.0% level. The 10yr Gilts have already exceeded 5.0% while the 10yr Bunds and OATs have reached levels not seen since the financial crisis as have the Nordic countries. And, most notably, Japan has made new records with the 10yr JGB yield crossing the 3.0% threshold and marking a 30-year high. Yet, the equity markets have been very strong. The S&P 500 and the Dow Jones Industrials Average continued to notch all-time highs as did the TOPIX index. What was notable was that the Nasdaq and the tech-heavy Nikkei 225 did not, although both did recover from the technology sell-off which continued until Situational Awareness’ bailout at the end of July. This is noteworthy as the market leaders that have been driving the indices to all-time highs have stopped doing so and is being replaced by different factors and industries. The bulls would say that the strong equity markets, initially driven by the AI capex frenzy, is now fueling the rest of the economy. Recent headline figures out of the US and Japan would seem to support that view. The bears, on the other hand, might note that previous bubbles stopped when the handful of industry leaders failed to break out to new highs; of course, the latter implies that the AI-driven technology rally was or is indeed a bubble. And, for the Dow Theorists out there, I’d note that the Dow Jones Transports have not being able to exceed the April 2026 high, though that pop was driven by a short squeeze on Avis. Still, the ex-Avis Dow Jones Transportation Average hit a high in July 2026 but had failed to break out in August when the Industrial did so.
 

Now I’m not trying to side with the bears, particularly as we still have significant tech exposure. In our concentrated portfolio, while we eliminated all but one technology hardware name by the end of the month (our largest position in the portfolio), half of our book is still software-related names depending on how one defines “software” (in our case, online B2B or B2C services). In our diversified portfolio, while we similarly have eliminated or drastically shrunk most of our technology hardware names, technology hardware+software accounts for over 50% of this portfolio as well (I include stocks that are not traditionally considered technology but have, more recently, moved similarly to tech). As such, rising discount rates are a headwind for the valuations of many of our companies (though most of our software companies are mature and already generating high ROIC and FCF, yet still growing topline at double digits).

But history has shown that an environment like our current one tends to be unsustainable. While I’ve been concerned about tech valuations for a while, all past bubbles popped from a credit event but, at the time, I couldn’t see what that might be. Sure, there has been plenty of anecdotal evidence like the Norinchukin “whale” bond losses on their CLOs, Tricolor and First Brands bankruptcies, regional bank loan losses, private credit liquidity restrictions and defaults, CMBS delinquencies, and the circular AI financing worries. Also, the macro numbers indicate some credit stress whether it be auto loan, credit card, or mortgage delinquencies. In Japan, we’ve seen consecutive months of rising business bankruptcies with personnel costs and tax-related bankruptcies being the largest driver; in the past month, bankruptcies reached the highest August level in 14 years. Separately, I recently read that nearly 40% of new housing loans are terms of over 35 years versus 20% just a year ago due to both rising interest rates and home prices. According to PayPay Bank (a major online bank), 71% of new home loans by those in their 20s have selected terms at 40 or 50 years; in an earthquake-prone country with a population decline, 50 years feels like an awfully long time.

Of course, none of these have or will cause a credit event itself. Furthermore, the argument goes that, unlike the dot-com bubble where companies like Pets.com were loss-making, the AI companies and hyperscalers who are driving the economy are highly cash flow generative. This was true, but they are no longer free cash flow generative and relying on debt to finance their insatiable capex. Another argument is that the circular financing in AI or the asset-backed financing of data centers is nothing like the CDOs (or the toxic CDO-squared that created successive layers of loss subordination). But a recent FT article quotes an executive who said, “You can capitalise the GPUs in different layers of risk, sort of like a CLO.” And let’s not forget this financing is backed by a fast-depreciating asset; as Michael Dell once said, [semiconductor] “inventory has the shelf life of lettuce.”
 
I believe in the AI story and what it can do for the world. More importantly, I believe it is absolutely vital for an economy like Japan that is seeing a shrinking workforce and fast rising wages. I am confident that automation and AI will more than offset that decline; we mustn’t forget that Japan’s GDP per capita is the lowest among developed countries on a nominal or PPP-adjusted basis and therefore has abundant room for labor productivity improvement.
 
But there is a fair value for everything. And with risk-free rates reaching attractive levels, I worry about the earnings yield of tech stocks. But within technology, I feel more comfortable owning an asset-light business, growing double-digits at 20x earnings or less vs a capital-intensive one that may be growing similarly at double-digits, but at 30~40x. And Japan has lagged far behind most other developed markets in terms of digital infrastructure and, therefore, I believe that the risk of AI displacement is comparatively low.

​

We shall see.
 
Masaki Gotoh
 
*** We will be co-hosting the inaugural SOHN Tokyo Investment Conference on Oct. 16. We hope you can all attend and help support a great cause with the Karen Leung Foundation *** ​​​

​

===

​

​​“Let me warn you, Icarus, to take the middle way, in case the moisture weighs down your wings, if you fly too low, or if you go too high, the sun scorches them. Travel between the extremes.’” – Translation of Ovid’s Metamorphoses, Book VIII.
 
I remember back during my early days at Goldman where I met a highly quantitative candidate that had a music background. She said that music and math were very similar; and I immediately related. After I gave up my youthful dream to become a musician, I used to have a hobby of listening to songs and creating the sheet music in my mind and sometimes on paper. The songs would flow onto the sheet music, and I found it very satisfying to visually express sound. I always had a knack for “seeing” patterns whether it be objects or numbers. Even now, when my mind is turned off, my eyes instantly connect dots on the wall or markers along the street or other random objects, unconsciously looking for patterns. I find mathematics to be, well, beautiful whether it be figures like phi (the golden ratio) or sequences such as Fibonacci or equations like Euler’s identity or topics like cryptography. I remember feeling excited when I took my first options class in business school and learned how combining stock price action from Geometric Brownian Motion and the hedging process of an option with some other simple formulas and lemmas lead to Black-Scholes. And I remember feeling even more satisfied when I learned that the price of options being traded in the markets were nearly equivalent to those computed by the Black-Scholes formula well before the formula was discovered.
 

When I began my career in finance, I liked watching the markets because the so-called “random walk” didn’t seem so random, and stocks, too, seemed to flow like music with different genres playing during different times whether it be minimalist classical, jazz, or hard rock. And like music, the changes between verses could be abrupt, much like symphonies. Of course, it is these inflection points that are hard to preempt. At times, the change is gradual and hard to notice, perhaps like a well-made mash-up or a DJ mix. But it is this variety that makes music so enjoyable.

But as I moved away from the quantitative side of the markets and steadily toward the qualitative, I also began to lose my sense of “flow”. It was alive and well during my period in proprietary trading, but I later worked for a fundamental long-short fund which took me away from my Bloomberg addiction. Still, I was fortunate that my PM was extremely price-action aware, though I had a lot of trouble trying to reconcile my new learning in fundamentals and short-term stock price action; before fundamental investing, I didn’t know or care “why” a stocked moved the way it did. Now, I needed to have a view. I moved even further away as my career moved to activism, particularly since most of my other partners didn’t even have a finance background (nor a musical one).
 
Here at TriVista, while it took me a bit of time to find my bearings again, I have attempted to reconcile these two differing approaches. The “Quality” variable helps define our universe with varying degrees of Quality. We believe the Quality will ultimately manifest itself in the stock price because of its superior fundamentals which, if our analysis is correct, should overcome any non-structural headwinds while excelling during tailwinds. If we were only following the Quality variable, we could just own our entire universe, sizing positions based on the magnitude of quality. But not only would that be a rather dull job, it would almost certainly perform badly. The music changes all the time and some chords fit the current verse better than others. So, we use “Value” (to be clear, we mean “undervalued”-ness) as a guide to help adjust weightings. One way I like to think about it is that we actually own a 100+ name portfolio, but most of them are 0% weight because most are not cheap.
 
But I think most fundamental analysis stops there. And that’s probably because that’s as far as fundamentals can get you. We can hyper-analyze the industry dynamics to become absolutely certain that a company is the best in the industry and that they will continue to take market share to the end of days, thus securing a higher growth rate than its peers. And, if management runs it right, the company will have an even higher earnings growth rate, thus achieving our double-digit EPS growth targets during our investment horizon. We can also calculate how cheap it is versus its peers by taking into account its excellent business and figure out how much the company is worth, albeit based on a highly subjective estimate of future cash flows and discount rates.
 
What I think some people miss, myself included, is the timing. This requires a mix of fundamental macro analysis and, well, listening to the music. I am obsessed with current events. I subscribe to over 10 newsfeeds, and I glance through all of them at least once a day, and several at least twice. I subscribe to a broad array of periodicals with varying frequencies. And I’m checking my work email all the time; it’s the first thing I open when I wake up and the last thing I check before getting ready for bed (and sometimes even while in bed). I know I should use AI to help me consume it better, but I like reading the nuances of the texts so I generally read through the ones that pique my interest; the AI summaries do help me decide if I should read the full article or not, though. I actually miss having over a dozen sell-side brokerage coverage because I liked hearing lots of different views (we have to concentrate our brokerage coverage in order to pay a somewhat meaningful amount of commissions to each, given our low turnover).
 
Through all of this (and, more recently, with my daily chats with AI), I formulate a macro view. Now, of course, we aren’t day traders. It doesn’t change what we do day-to-day. But this helps me understand the tone of the music and whether the chords we own are in harmony with the current market music. And it helps me formulate a view on whether these chords will be consonant or dissonant with the market in the near future and, more importantly, in the distant future (if they are going to be dissonant for too long, we should probably review our weights). There’s a wonderful dialogue in “Margin Call” where Jeremy Irons (playing the fictional CEO of a major investment bank at the brink of the Global Financial Crisis) says “Do you care to know why I’m in this chair with you all? … I’m here for one reason and one reason alone. I’m here to guess what the music might do in a week, a month, a year from now. That’s it. Nothing more. And standing here tonight, I’m afraid that I don’t hear – a thing. Just .. silence ...”
 

I listen carefully on an individual stock level as well. When we decide that one of our most excellent companies is now very cheap, I certainly don’t rush to buy it. We first think why (fundamentally) it is cheap and why (fundamentally) that might change, or at least, why the stock might revert. But we need to go beyond that. What is the price action telling us? Are we so incredibly intuitive that we know something that the market does not (which is highly unlikely)? Most importantly, do we truly understand what the market is thinking? I make extensive use of my obsession with current events to try and understand what the market *might* be thinking about the macro environment surrounding the stock and the company itself by listening to stock prices. And I just don’t rely on our fundamental hypotheses as to why, but I try to think of second or third derivative reasons as to why the stocks *might* be doing what they are doing.

 

But ultimately, I let the music tell me. Maybe the rhythm starts to change such as an increase in volumes. Perhaps we get negative news in the industry or, more significantly, the company itself via yet another weak earnings result but the stock isn’t reacting. Or perhaps it does fall initially but reverts in a week or even intraday. The music has started to change. Such changes in tone prompt me to consider an investment in the name now vs when it first turned cheap.

 

Admittedly, what this means is that we are slow to react. Also, we generally want corroborating fundamentals, even if yet unrealized. I wouldn’t invest in a stock from the stock momentum or reversal alone, for example. Therefore, we especially have a difficult time investing in a stock whose valuation is rising unless the fundamentals are rising at least as “fast”, although how one measures the speed of fundamental change is difficult. Additionally, we need to also understand “why” it is changing and, more importantly, “is the reason sustainable”.

 

At the start of this year, our hardware technology stocks already weren’t cheap. Multiples were rising but so were earnings expectations. Therefore, while somewhat uncomfortable, we held on to them. But as we started to see less bullish price action with technology stocks in the US and South Korea, we became concerned. Stocks were falling after explosive results that were well above the top range of consensus. Commentary was strongly bullish, but the stocks weren’t corroborating. Maybe the stock would open strong but close down. I’d go to sleep while listening to the strong US earnings results, but I’d wake up in the morning and see Nasdaq close down for the day. The music was changing – the DJ was mixing a different song. It ultimately took us 6 months to get out in full, not because of liquidity but because the mix is still on-going. To be honest, I have no idea how long this could last. Or maybe we go back to the old song. Maybe we already have and I would’ve sold prematurely.

 

As I continue to say, I like the long-term fundamentals of AI. But I think it simply flew too high. Maybe I’ll get another opportunity should they start to fall back to earth. Admittedly, I would’ve preferred that they took the middle way with steady double-digit growth. But the chords are increasingly sounding dissonant right now and that is a painful sound to listen to.

 

Some people call this technical analysis. Some call it voodoo. I’m just listening to music.

 

PS: Goldman dinged the candidate I interviewed. I really wanted her for our team in Derivatives Research but, back then, you had to clear dozens of interviews across multiple departments in equities. She was an odd character so I can understand it would have been difficult to get everyone to like her. But it was precisely that oddity that I liked. I hope she found a rewarding profession and if she happens be in finance and is reading this, please reach out! I interviewed you in New York in 1999.

​