Field notes from this week's moat hunt
Stream of consciousness on deposition tools, diesel engines, and more
What I Worked On/Thought About This Week
Each day, I spend 5–30 minutes writing a stream-of-consciousness “brain dump” about whatever I worked on or thought about. I do a quick edit (or sometimes none at all) and post those daily reflections to X (@stockthoughts81). At the end of each week, I compile them here in one place.
The goal is to clarify my own thinking, keep a record of my day-to-day process, and hopefully spark conversations with others tackling similar problems. These entries are closer to raw journaling than polished essays—maybe I’ll flesh them out more at some point, but for now, I keep it simple.
If you enjoy this kind of journal-like content and want to keep following along, feel free to sign up below. And please don't hesitate to reach out with thoughts on any of the companies or topics discussed.
2/3/25
-Moats and risks in WFE
-Tariffs
-LinkedIn cold outreach for semi co and VRTS
Small semi co today. The moat? Same thing as all the other WFE players. Technical know-how, customer relationships, switching costs, etc. Once you spec them in, kind of risky to try someone new. Small cost of the overall cap equipment spending and op costs for a fab. Don’t fix it if ain’t broken. Evidence of the moat? Strong ROICs/ROTAs. Strong market share (70% in batch ALD). AMAT tried to buy them in 2019. Couldn’t get Chinese regulatory approval in a timely manner so they ended up terminating it in 2021 and paying a $150m fee. Now AMAT owns a 15% stake in them. But what’s different here from e.g. AMAT LRCX MKSI etc. is that this company plays in one very specific part of the market. Their real strength is batch ALD, for the most part. This presents more technological risk. Is there a way to make single wafer more efficient? Spatial ALD? Plus - it is harder to switch WFE suppliers for multiple steps at once. 30% of AMAT’s sales in 2023 were of systems that integrated to 2 or more adjacent steps (20% in 2019). I guess small semi co kind of has this with treatment + deposition? But deposition share only 10-20%. I don’t know. It doesn’t feel as certain. But it’s also really interesting how closely AMAT works with them and how badly they wanted that deal. Obviously there was a build vs buy decision - why didn’t they want to build? They remain close partners to this day, with an AMAT exec actually doing an industry overview at their investor day. But still - i don’t know how solid the moat is. Is this just to be handled via position sizing?
Reviewed tariffs. Not much of note. PYPL not that great position. KMX maybe benefits in short term. CMI negative. MKSI maybe exposed via price increases on end devices.
Call with a company dealing with some activist involvement. They didn’t seem to be aware that they would likely lose a proxy war. That surprised me. Would they not want to take the time to figure that out? Or are shareholders not being honest with them? They didn’t even ask us how we would vote. I’d be surprised if they get the votes.
Doing some LinkedIn cold outreach. One is looking for people who used to work at VRTS. The stock seems so cheap, I don’t know what we could be missing. Basically want to ask them - how is morale? Is this falling apart on the inside? And then another project focused on the semi co. Basically searching for people that mention “ALD” in their profile and are in my area. Will ask to buy them lunch I guess? And IDK the incentive for the VRTS formers to talk. IDK. Maybe both of these should be Tegus projects. Just adding people for now, will send messages later based on who adds back. Maybe I should be using Dux-Soup for this, I don’t know.
2/4/25
-$CMI earnings review
-Does anyone know how to boost X engagement (esp. for blogs)?
-Why is $PYPL down so much?
-Metarant about DeepSeek + Tariff expert rants
1) $CMI reported. Pretty solid Q. The CEO said the quiet part out loud, “Frankly, the slower and messier [the renewable energy transition] goes, my view is, the better it is strategically for Cummins.”
Heavy duty softness by Medium duty and Distro strength, Power Systems still killing it, some price and operating efficiencies. Strong guide for 2025. Expect HD to pick up in 2H with some pre-buy. Important to monitor the risk of whether
$312m restructuring charge for Accelera. Not on track to breakeven by 2027, but still on track for total company goals (2030) laid out at Investor Day. Potential risk to monitor is whether they are actually just losing in this market vs the market itself is losing. I don’t see any evidence of that. Not sure how to diligence it (not really even aware of any big notable competitors) but haven’t seen anything that concerns me.
Expect total China rev incl. JV to be +5% in 2025. On a more normal run rate, China is 15-20% of earnings. And on-highway typically over half of that. Any surprise to upside will be high incremental margins.
Increasing Power Systems capacity. Doesn’t seem very recurring. AI capex risk?
Expect 2027 EPA regulations to stay in place despite Trump. Everyone already invested for it. 2030+ regulations an open question.
HELM - goal is 8% nat gas penetration. Reality is it will still be a lot of diesel. Nat gas adoption depends on diesel fuel prices, regulation, customer CO2 goals, etc.
I am looking at a midpoint fair value of $393.
2) Worked on everyone’s favorite Batch ALD company. Report mostly done. Misc. reading and skimming and calls to try to fill in the gaps now.
3) Increasingly frustrated at Twitter algo. Can’t seem to get it to work for either this, or even worse, the blog. Please comment or DM if you know a workaround.
4) I find myself on alternating sides of this, but I don’t know about the whole Dunning-Kruger “one week you’re a DeepSeek expert, the next it’s tariffs” meme. Whatever happened to mental models and all the praise for Charlie Munger’s approach to the world of being multidisciplinary? Are we supposed to just ignore all this stuff? I get that I am not going to be able to be an expert on everything. And a lot of the DeepSeek stuff out there is above my head. But can’t we selectively try to find the info that is the most useful and understandable to us, and form opinions (however loosely held based on that)? And try to learn from discussing these topics with others? Obviously the issue is overconfidence. Unfortunately I think a lot of that is driven by the way the algo works because people know confident posts drive more engagement.
5) I cannot figure out why $PYPL is down this much. Branded TPV slower vs e.g. Visa? AMZN? Unbranded TPV decel to 2%? What is the real concern here? My big issue here is that where they had the widest moat and thus were the most profitable (branded) - that moat is eroding. Not good. Still seems cheap.
6) $MTCH new CEO Spencer Rascoff. Former CEO of $Z and Hotwire (sold to $EXPE). About time.
2/5/25
More semi stuff. Ran into a strange situation where an IR deck shows a Gartner market size number and accompanying share for the company. And that math leads to a much larger % of total revenue for the company vs what the co itself discloses. Weird.
Factor modeling webinar. Factor modeling is still mostly above my head but seems interesting.
Everything in research seems to take longer than expected. Hopefully wrapping up first draft of semi stuff either tonight or tomorrow but who knows. Lot to review for the meeting this week.
2/6/25
$EXPE $MKTX
Expedia is crushing it. Strong accel in 4Q. Something very strange- the incoming CFO was on the call even though he technically doesn’t start until the 10-K is out and was just appointed 12/19. Strange! Wonder what the scuttlebutt is on what happened with Julie. Vrbo accelerated from 3Q. HCOM return to growth. Buybacks slowed down a lot. Dividend reinstated.
Finished up first draft of semico. Idk. Not my best work but I wanted to get the first draft out ahead of time for team to review ahead of expert calls. This is a tough one. Normally I would’ve wanted to read more but I think the base is there. I skimmed the semi-analysis piece. It felt quite technical. Like, going into the details of how deposition is done in NAND, DRAM, Logic. IDK. It’s good, I think? I have no way of judging that. Over my head. All I know is - they are good in batch. I certainly don’t like that as much as a company like AMAT or LRCX where the moat is more in the customer relationships, breadth of offering, service quality, etc. Some of that is here. But you’re just so dependent on that little niche. I think there is enough there for me to be comfortable potentially investing, but position sizing comes into question. Cap at 5%? 10%? 15%? Idk what the right answer is. And I am still doing more work. Kind of skimmed some of the existing Tegus and read Bernstein/MS with varying levels of depth. You also have export control risk. Competition from legit players like Tokyo Electron and ASML Intl. But also like NAURA in China. Batch ALD dominates 3D NAND. And they are just going to keep stacking layers so idk why that would change. DRAM I think is next strongest, and then Logic in some ways is kind of TBD? But it’s weird because they also say that will be the biggest portion of revenue in medium term guide. Idk. I haven’t worked on many (any?) Japanese cos before and there are just some weird things that pop up. It’s IFRS, for one. And then - just- disclosures and what not, and how they give details and numbers (or don’t) on the calls. Like - the 50% number for Batch ALD is as % of Equipment revenue or % of total revenue? Idk. And like I mentioned the other day - Treatment number doesn’t line up with Gartner. So I don’t know. I am like on the very edge of my circle of competence/comfort here. Doesn’t help that there is no one to talk to about it. Hardly anyone has tweeted about it beyond headlines. Excited for the Tegus calls, though. But also wondering if it’s just going to be over my head because these people know so, so much more than me. Need to keep doing the work though. And even if I never quite get there on this one, I do think this has helped me understand the rest of WFE more. And made me like MKSI even more.
Reviewed for research meeting. Could be a long one tomorrow.
Some MKTX stuff. It seems like they might be losing share to Trumid??? Not expected. Mgmt says they basically just do new issues when you ask them. And I am not sure, that may have been where they started and what they are known for but I think they are taking share. Put together their ADVs from their PRs and you see market share spikes on new issuance (exactly when MKTX goes down) but you also see the overall amount steadily climbing over a couple years. From like $1b to $5b or something. Seems like a real third player now.
My balance of working vs thinking has not been good lately. Too busy. In work and out of work. Need to figure something out. Need to get through research meeting tomorrow and then afternoon I can hopefully zoom out and catch my breath.
2/7/25
I am back to being unsure how to feel about the hyperscaler capex. The ROIC case is just becoming more and more important, and isn’t any less fuzzy than it has been before. I lean toward “probably fine” for now. The risk is they are just doing this because they have to in order to defend the moat and it will bring down returns for everyone.
The Deutsche Bank note on China is really interesting (or at least what I saw of it on X).
PYPL - not sure how to feel. May do more work here. Maybe attend investor day.
2/8/25
Been thinking about whether US based investors should have China exposure. I tend to agree with the DB note put out last week - many of the widest moat companies are turning out to be in China. The market sentiment is palpably negative. Is it out of my circle of competence? I don’t know. It probably requires more maintenance than other things. But this is sort of the case for any generalist who wants to try e.g. a new industry. Sure, I am not on the ground in China. I am also not on the ground in Safran’s factory figuring out how engines are made. That is what experts and sell-side are for (not that you have to know how engines are made to invest in SAF but you get the point). It’s just exaggerated here, having to e.g. follow a few newsletter you otherwise wouldn’t so you have some sense for how things work there. I don’t know. Open to thoughts. I think part of what might make this seem more difficult than it is, is that people get their china news and views from Eg the WSJ. Kind of like the blind/biased leading the blind/biased.
Disclosure: none of this is investment advice. I/we may have positions in securities mentioned.