We had another incredibly strong month in Japan. But I have to admit, I don’t know why … well, I have suspicions as to why, tariffs and FX. Visibility with regards to tariffs has, somewhat, improved (before the time of writing, it was still a little unclear, but the executive order has since been executed). But whatever the number, at least the tariff is now fixed (until it isn’t …), and I guess that’s a good thing at least from what the market seems to be suggesting. During our post-Jun Q company visits with manufacturers, they still seemed cautious due to the possible impact on demand from tariffs and, at least at that time, hadn’t made concrete capex plans except for those that have been on-going. As mentioned last month, this quarter is the (fundamentally) least important quarter. But among the top 1,000 names in Japan, there were 89 upgrades to full year guidance vs 50 downgrades with an average change of -3.6% (10%-trimmed mean of +0.3%). At the same time last year, the numbers were 136 and 31 respectively (average change of +14.7%, 10%-trimmed mean of +14.1%). So, I wouldn’t say earnings were particularly strong. While I haven’t gone through all 89 upgrades in detail, about 1/3 were due to conservative FX and/or tariff impact forecasts at the start of the fiscal year, and another 12% were due to non-operating factors. The remaining half were mixed, but they leaned toward areas such as inbound traffic/Expo2025 demand, commodities and utilities, consumer staples, and IT services.
 
Admittedly, full year EPS estimates for the fiscal year did rise during the quarter by about +1.2% for Topix (and +2.3% for the exporter-heavy Nikkei 225 which benefits from the weak yen), while the trailing 12M EPS fell QoQ for 2 consecutive quarters (i.e. the last 2 quarters were down YoY). The last time this happened was in 2022 during which the markets were flat. By TSE sector, the largest sector “Electric Appliances” index, which is essentially your tech-related names and accounts for about 18% of TOPIX, saw EPS forecasts rise +2.1%. This was likely driven by FX as the yen has stayed persistently weak. However, the big revisions were in Banks (accounting for 9% of the index) whose EPS estimates grew +16.2% from the previous quarter, presumably as interest rate expectations continue to rise (the top 3 banks, which accounts for 70% of this index, all exceeded consensus, and EPS rose over the month). And the third largest sector, Telecommunications (8% of the index), also rose +4.7%, though I suspect this is mostly Softbank (which is over 20% of the index) and I’m guessing investors don’t trade Softbank on earnings. The next 5 sectors which cumulatively account for about the same percentage of the index as the top 3, all saw EPS estimates fall (Wholesalers, Transportation (=Autos), Machinery, Chemicals, and Retailers).
 
So, it’s probably not earnings that drove the market. So that might leave flows. There has been lots of news that foreigners have been driving the Japanese market. And the data agrees. Since Liberation Day, foreigners were net buyers of +5.4 tln yen of Japanese equities. This was followed by a very close second of +5.2 tln yen by corporate Japan (i.e. buybacks). The net sellers were retail investors, selling -5.1 bln, and financials who sold -4.1 tln, roughly 50:50 between cross shareholding unwinds and pension plans. The trouble with this argument is that foreign activity was actually flat in August. In fact, they were net buyers only in the second week and have been net sellers otherwise. The only investor group that had been consistent buyers were the corporates. While they’ve almost always been net buyers, the pace has picked up from mid-2024, I’m pretty sure thanks to the JPX. But they typically are not the driver of market direction.
 
I’ve exhausted my hypotheses for why multiples increased last month. I’d like to say that it was driven by the large caps but, in actuality, small-caps outperformed mid-caps who outperformed large-caps while EPS revisions were about the same (+1.2% large-caps, +0.8% mid-caps, and +1.3% small-caps). The only anomaly I notice is that volumes are, again very abnormal. August has been consistently one of the slowest months of the year, generally ranking #11 or #12th of the year in terms of monthly volumes. The exceptions were the flash crash in Aug 2015 which is the first time since data is available when August was the most active month, and the second time was August 2024 from the surprise BOJ rate hike. This August has again exceeded the volumes from the last 12 months. I’d note that this phenomenon did not occur in the US or Europe, and August was relatively quiet as usual. Whether this forebodes something on the horizon, I have no idea … but I’m guessing it might, though in which direction is anyone’s guess.

​

Masaki Gotoh​​

​

===

​​

“Woe to the makers of literal translations, who by rendering every word weaken the meaning! It is indeed by so doing that we can say the letter kills and the spirit gives life.” – Voltaire, philosopher and writer.

 

I had mentioned in a past monthly (Dec 2023) how the non-verbal communications, while important in any society, is particularly so in Japan and one of the reasons why we choose, where possible, to have face-to-face meetings. And, while it was painful to the point of abusive to have to travel during this summer’s heat wave where temperatures reached above 40 degrees Celsius (about 104 degrees Fahrenheit) in abominably humid Japan, there were several times that reaffirmed for me why we choose to stay in this high tax, high cost-of-living, and physically straining city.
 
For example, I met a comp to a company we have been researching. Our potential company is a leader in a highly fragmented industry whose consolidation is accelerating. I had already met the #3 and #4, both of whom were not serious threats. And, therefore, I was looking forward to meeting the #2 (we’ll call them Company XYZ) who had been one of the most aggressive with M&A and trading at the highest premium within the industry. Furthermore, XYZ is the only firm that has a JV with one of the largest peers globally (whose revenue is 2x that of the top 4 Japanese combined), so they seemed like a formidable adversary. While I had questions about the synergies from several of their recent targets, and, furthermore, they had reorganized the firm several times, often a sign that requires caution, it is understandable given their frequent M&A. While they had the lowest margins among major peers, the industry is structurally a thin margin business and I had assumed that they were making accretive acquisitions of smaller targets whose margins were lower, thus allowing room for margin expansion through economies of scale and operational leverage.
 
The first thing I noticed, after making that terrible trek from the train station to their office, was that their headquarter building was huge and they occupied the entire building with the name “(Company XYZ) Building”. I had already seen that they had sold this HQ and leased it back several years ago, but I was still surprised at its size. Upon entering, I noticed the long list of companies, all subsidiaries, in the building directory. Their filings list 58 consolidated subsidiaries and one equity method affiliate. To put that in perspective, the average number of subsidiaries for companies of similar size (+/-20% in market cap) is 15 subs of which roughly half of these 193 companies have 10 or less. My first thought was, how could they manage so many companies? While I was cooling down in their large foyer, I noticed 2 people also waiting for their appointment. I noticed that they both were wearing Sony badges. I already knew that Sony was a large client, supplier, and shareholder. One of the key points about this industry was that there were two types of companies, independent firms and those affiliated with a large manufacturer. XYZ played down the Sony connection according to some sell side reports, insisting that they were independent, but they had the characteristics of an affiliation such as lower margins, large percentage of revenue, and capital ties. While it was most probably a coincidence, the only people waiting to meet the company were the people from Sony and myself.
 
After the meeting, I came to the conclusion that their M&A strategy was not very organized and that they did, indeed, have little grip over their subsidiaries which could be a significant governance and/or operational risk in the future. In a sense, they were the perfect comp, one who was consolidating the industry which would be good for all of the leading companies while staying ineffective at becoming a serious competitor. Admittedly, I would’ve probably come to the same conclusion via a Zoom call. But peer meetings, in particular, are best suited for physical meetings. I can ask the tough, probing questions with a smile on my face and gesturing positivity which often helps to extract honest, less biased information.
 
There was another example that, while it is not an example of the benefits of physical location, shows how it helps to be Japanese. I was in a group meeting with an analyst from one of the leading US bulge bracket securities firms along with both Japanese and non-Japanese investors. As such, the analyst would ask the question in Japanese, then translate it in English, receive the answer in Japanese, and translate the answer again into English before going on to the next question. So, I, of course, get to hear the answer in both languages.
 
This analyst has a BUY on the name and the second highest target price on the Street. Obviously, they are biased toward the upside. Now, I, too, am long-term bullish on the name which has both cyclical and structural tailwinds, but we only have a small position because it is not very cheap. I find sell-side-led group meetings to be highly efficient as I can get the short-term laundry list completed while also getting a feel for what others think of the stock. However, the downside is that the progression of questions generally leads toward the analysts’ bias which, in this case, was very bullish. Furthermore, they generally have a checklist of questions that they need to complete in order to write their post-earnings notes, so they don’t focus on those answers that were unexpected and warrant further investigation. If I heard a datapoint that is contrary to what I heard elsewhere or one that contradicts what the speaker said himself or what is in the presentation, I’d stop my checklist and drill down. This is where one often uncovers the most interesting facts, whether it be about the company or the industry or perhaps even leads to a derivative theme. In fact, unlike sell side analysts (and many buy side analysts as far as I can tell from these group meetings) that often asks a question looking for affirmation, I generally ask the same topic in the opposite direction, hoping that they would dispute it, something that Japanese people tend not to like, which is why this method is so powerful. If something looks good, I’d focus on the downside of it and see what they say. If something looks bad, I’d praise the positives from it and see if they try to deemphasize them (and, if they don’t, I’d go back to the negative and see how they reconcile). This is similar to null hypothesis testing in statistics whereby, if I can reject the null hypothesis, I can conclude that the alternative hypothesis is true. Of course, one has to frame the hypotheses in wording that wouldn’t be offensive or too direct, a very Japanese colloquial etiquette.
 
And, so, we were going through his checklist, confirming the positives of the quarter. But I noticed that, when the IR speaker (whom I’ve known for over a decade and generally speaks unbiasedly) mentioned a word of caution as a caveat to the affirmation, the sell side analyst skipped over that part. Arguably, it deserved to be glossed over as it was often trivial and subject to interpretation. Given the limited time from all of the translations being done and the list of checkboxes that had to be completed, it’d be, at most, a footnote if this were a text. But after a few of these cases, I could no longer hold back and had to interrupt. IR was confirming that, as they had said at the large presentation, they believed the Dec quarter would see improvement QoQ. I suppose everyone was willing to leave it at that. But I’ve heard similarly from many firms in the sector without any backing. Furthermore, this company, whose IR I trusted more than others, rarely provides guidance that far out so I assumed they had something a little more concrete than “because I think so”. So, I asked which segments they are expecting improvement and what is driving that optimism? Are they seeing increased activity from those clients? Are you getting queries for 2026 deliveries earlier than usual? Or is it a capacity issue particular to you? Unfortunately, the answer wasn’t very concrete. Those that have been strong will likely continue while those that were weak looks like it’s bottoming, and it really can’t get much worse with inventories where they are. In essence, “because I think so”.
 
Now, we’re long the stock for long-term structural reasons. Whatever his answer, I wouldn’t have moved our exposure. And since I trust him more than other IR officers, his “I think so” means more to me than someone else’s. But, especially with valuations where they are, I wouldn’t be calling my clients to buy the stock today. But sure enough, the next day, the report came out that “new growth drivers … will drive earnings, leading to above-consensus profits … While [the] share price has corrected somewhat following its latest earnings release, we continue to see significant upside potential and maintain our Buy rating.” I’d note that there were just 2 questions out of well over a dozen regarding those new growth drivers (though, to be fair to the analyst, IR did confirm progression in two of the three new growth drivers often highlighted by industry experts, one of which seems not insignificant in the not-so-distant-future, although IR downplayed the third). Also, I noticed that the way this meeting was summarized in the analyst note is definitively biased to the upside. For the above comments about Q3, one could write (as I heard it) that “While [they] expect shipments to increase in the Dec Q, there is, at present, little visibility.” But they chose to write, “While the outlook for Q3 shipments … is still unclear, … [they] said it is monitoring demand trends for xxx, which are a leading indicator of a recovery in capex.” Both statements are true. But I’m guessing the difference in tone can be noticed (I wonder if AI could figure out the difference?)
 
Admittedly, all of this is marginal. But I continue to believe there is a benefit to our team being culturally Japanese and on the ground. The observations above were for short-term, trivial items that tend to be more detailed and, therefore, should be less prone to biases and yet, they are still subject to interpretation. Think of the differences that interpretation could bias if this were something as abstract as long-term competitive advantage.​​​

​