Tesla Troubles, Software Weakness, Copper Hype, And Much More
Bryce here. I hosted Chat this week since Cody was on the road. Here’s the transcript from this week’s Live Q&A Chat:
Q. In general, are you still raising cash these days? Or have recent trims been only because of ‘normal’ trimming’?
A. We are not raising tons of cash right now. More just sitting tight and opportunistically trimming when stocks pop.
Q. What are your current top 5 positions?
A. Top five positions in alphabetical order: Meta (META), NVIDIA (NVDA), Tencent (TCEHY), Tesla (TSLA), and Uber (UBER).
Q. What would be your top buys and sells today?
A. The only things I might be interested in buying today would be Amazon (AMZN) and Toast (TOST), maybe some Uber (UBER). The top 4 trims (in alphabetical order) would be Apple
(AAPL), Meta (META), NVIDIA (NVDA), and TransMedics Group (TMDX).
Q. Regarding Tesla (TSLA): It’s hard to be bullish when prices of Teslas and other EVs are cratering. The valuation of TSLA is mainly based on cash flow, revenue and the promised revenue growth of EVs. EVs made up 84% of its revenue in 2023. If EV sales growth either stops or reverses to negative growth, and profit cratered or goes negative, would AI, and promised robotaxi (which might not even materialize for a couple of years) rescue $TSLA as a stock? Current data for Q2 points to lower EV sales (2023 comp), just like Q1.
A. I think it might be a stretch to say that Tesla prices are “cratering.” Tesla has cut prices for the Model Y, for example, somewhere around 27% from the peak in 2022. Meanwhile, production has ramped up significantly, thus Tesla’s margins have only declined from about 22% to 18%. Obviously, we need prices to stabilize but for the moment it seems like they have. More importantly, a lot of the value built into Tesla is based on the future promise of FSD, Robotaxi, Optimus Robots, Dojo AI, etc. We are seeing tons of progress in FSD and more developments on any/all of those projects should help the stock price. Tesla is the only mega-cap company I can think of that clearly has multiple multi-trillion dollar AI kickers, all they need to do is execute on the vision.
Q. As for Robotaxi, we should keep in mind how Elon has promised “next year” FSD every year for like the 10 past years, how do we know this is not another of his promises? Yeah, my personal experiences with FSD v12 was a leap forward but lately, there seemed to be steps back as I had to intervene a few times. Another worry is competition not only from Waymo, but also from multiple Chinese OEMs. I’ll be keeping my eyes out on how FSD fares in China. Can FSD best Chinese OEMs is the billion dollar question. I’m hedged.
A. Elon has famously bad timing. He even admits that. We would doubt the promise of FSD if we hadn’t seen such substantial improvements ourselves firsthand. The switch to an end-to-end neural network from a rules-based approach was obviously a game changer for the FSD software. In terms of competition, Waymo has less than 1/100th of the miles driven and they will probably never have enough data to train a neural network in the way Tesla has. Moreover, Waymo cars cost about 5 times as much as a standard Tesla. Chinese EV manufacturers are the only real competition for Tesla but I also think they are way behind on autonomy.
Q. Cathy Woods just issued a five-year target of price $2600. Your thoughts?
A. I appreciate Cathy’s analysis but it’s strange to me that nearly all of the gains in their model come from Robotaxi. They ascribe no real value to Optimus, standalone FSD, FSD licensing, Cybertruck, Dojo, Semi, batteries, solar, etc. I do think the Robotaxi model will be great, but really the Optimus Robot and Dojo AI have the most upside potential. We are clearly bullish on Tesla (we predicted it could get a $9,000 share price by 2040 in our book), and I think our approach is more well-rounded than ARK’s.
Q. Your thoughts on Elon diverting Nvidia’s H100 chips to his own xAI company? Why do you think Elon created his own AI company instead of forming an AI division in TSLA?
A. I think there could be any number of reasons that one or two orders of NVIDIA chips were sent to xAI instead of Tesla. That reporting came out of Reuters which is out and out anti-Elon and has done some really bad journalism regarding Tesla of late. Personally, I wish Elon had just built xAI within Tesla but his stated reason for making a startup was so that he could recruit top AI employees by giving them equity instead of $1mm+ signing bonuses, so I understand why he did it. At the end of the day, I’d bet that Tesla is a net beneficiary of xAI, not a net loser. Don’t forget that nobody has more of a vested interest in seeing Tesla succeed than Elon, especially after winning the options package today.
Q. Hertz (HTZ) is selling 2023 Teslas for about $23k.
A. I’m not sure what Hertz — a $1bb market cap car rental company — does with its Tesla fleet is necessarily indicative of the market for Teslas. Hertz had somewhere around 35,000 Teslas at the end of last year, and Tesla sold 1.8mm Teslas last year. Also, Hertz regularly sells their used vehicles off so I wouldn’t reach too much into this. I see more and more Teslas on the road every day.
Q. Would you buy some Uber (UBER) here? Are you and Cody not as worried as you were when you sold 1/3?
A. If you don’t own any you might pick some up here. Don’t forget we were pounding the table on Uber early last year when the stock was around $30. We think the stock probably overshot to the downside because of Robotaxi which may still be a year or two a way from release, and another five years or so after that before it goes truly mainstream. I don’t know exactly where Uber will fit in the Robotaxi world, but my guess is that they will work with Tesla to bring Robotaxi to market. Uber is also starting to turn very profitable and if Robotaxi doesn’t destroy the business in the next five years, its a good investment here.
Q. Seems like Transmedics Group (TMDX) is going on a roll. I’ve seen big hospitals in Florida advertising the number of transplants they do. Could this 10x in 3 or 4 years?
A. If TMDX keeps growing at this pace, it certainly could be a 10-bagger from here. The stock is not cheap, but if they end up being able to transplant tens of thousands of organs in the US and eventually around the world, the stock will be much higher than here.
Q: Good afternoon, cybersecurity seems to be a strong long-term trend, any buys such as Palantir (PLTR)?
A. A lot of enterprise software companies like Snowflake (SNOW), Zscaler (ZS), Palo Alto Networks (PANW), MongoDB (MDB), and Cloudflare (NET) have sold off recently. Crowdstrike (CRWD) is probably the only one that has hung tough. I think large companies are shifting their IT budgets to focus on AI-specific software, and are placing a “pause” on adding to their budgets for traditional software. If you are not an AI-native application, companies may be wondering if they will even need your services in a year or so. If within a year or so we don’t see any killer new AI apps come out, then maybe enterprises start spending more on traditional “legacy” software again. Regarding Palantir, I’ll just say that I still don’t really know what Palantir does (does anybody?). I think PLTR is probably helping companies figure out what to do in AI, sort of like a consultant/IT service company like Booz Allen Hamilton (BAH), Cognizant (CTSH), or IBM (IBM) even. Maybe PLTR has something proprietary that those other companies don’t, but I cannot figure out what that is. Do they make their own LLMs? No. Do they own a bunch of NVIDIA GPUs? No. Do they have the best AI for the physical world? No. So . . . what do they have?
Q. How does it feel to not have any cybersecurity stocks in the portfolio as we see trends being strong and AI will likely only contribute to their need exponentially more? Any regrets selling Crowdstrike (CRWD)?
A. Fear of missing out is one of the worst reasons to own stocks. Clearly, there will be secular growth for cybersecurity and cloud software for many years, but I’m not sure that justifies paying 10x, 15x, or 25x sales (not earnings!) today. I rerun the software companies through my model all of the time and they are still not terribly cheap. That said, there has been a lot of pain in the software space lately, and semis are on fire. The semi valuations are almost as bad as software. If I had to make a trade, I might go long some of my favorite software names (including CRWD)and short some semis.
Q. One of the former picks in the space revolution, Redwire (RDW), has been on quite a run of late. Lots of volume too. Any insight there? Do you guys still follow those picks closely?
A. Yes, I wish we had stuck with RDW longer. RDW makes a lot of satellite components and that business is really picking up lately. It’s a good business but they still aren’t profitable and probably won’t be for a few more years at least. They will probably also need to raise cash between now and then. I like the company but I’m worried there is significant dilution risk down the road. Maybe next time the stock crashes we’ll put it back on the sheets.
Q. Over the past few months, I’ve heard many people talk about how copper was going to be the next big winner. Do you guys like copper over the next couple of years?
A. I think copper got hyped up with AI and this whole idea that we will be building billions of dollars worth of new data centers and electrical facilities over the next decade. That might be true, but that doesn’t mean that copper should have rallied 60% in two months like it did earlier this year. Prices have already come down some but I think there is probably more room to the downside, more so with the copper miners than the commodity itself. There is no reason for a copper miner to be trading at a 20-30 P/E. We actually mentioned on May 29 that Southern Copper Company (SCCO) was a good hedge and that stock is down about 11% since then.
That’s it for now folks! Till next week.
