The S&P500 is trading at new highs this morning after the Nvidia earnings release last night. While personally I am pretty damned tired of all the hype around AI I will agree that some of technology is pretty fascinating–in fact if I am writing on my desktop computer I use AI to scan my writings–taking advantage of basic features such as spell check but also taking advantage of letting it restructure sentences. While I am using just free versions of AI programs I can see the time soon where I will step up to pay versions with many more features. Unlike all the featured headlines I see on Yahoo Finance which have story after story of worthless articles which are trying to sell me new Motley Fool newsletters (as an example) I don’t see me ever allowing AI generated articles. I can guarantee you that some of the drivel I see on Seeking Alpha is computer generated and is worthless.
Yesterday we had the Fed FOMC meeting minutes released and in my opinion they contained just what I thought they would—we need data and if that data is hot rates could go higher, but mostly they need data to consider rate cuts sometime later in 2024. Really no new news if one is paying attention, but since most of what we see on the boob tube is folks simply recapping Fed statements without any value being added–simply they need eyeballs to be able to sell advertising. I watch CNBC from 5 a.m. until 8 a.m. each morning then off it goes–no need to watch during the day.
In spite of Fed FOMC minutes–in spite of NVIDIA earnings the 10 year treasury is trading flattish again–around 4.42% this morning. The S&P500 is way up of course since the S&P500 has become very mated to the NASDAQ and tech stocks. Obviously NASDAQ is up sharply this morning as every AI related company is up.
NewTekOne (NEWT) didn’t price their new baby bond yet–at least I can’t find a SEC filing. I have an interest in maybe taking a nibble in the new issue if the terms and coupon meet my ‘high yield’ target (maybe 7.75% or more). On the other hand NEWT has a baby bond with an 8% coupon trading right at $25/share already (NEWTI) which would make a reasonable addition to my portfolio. We’ll see what happens here.
NOTE—NEWT has priced their new baby bond with a tasty 8.50% coupon with a maturity date in 2029.
This (NVDA) is my darling for my trading/risk account. For nigh on 10 months now, I have been legging into weekly NVDA strangles – I usually get $300-400 strike spreads and net $150-$200 per contract at 1-6 contracts per week. The volatility allows me to make the goal posts very far apart for a 4-day hold.
I have averaged $500-600/week profits. Only once was I put the stock (at $800) and immediately sold deep ITM calls against the position the next week and they were called away. Once, I felt the need to bail on the call side when we had an out-sized move higher (when it first went to intraday $974 when I was short the 960s) – turns out I bailed on a position that would have expired good but my risk mgmt strategy got me out.
This has been a money machine for me but, it may come to a close with the 10-to-1 split coming next week – I’ll have to see what the options do post-split. Fun stuff, and profitable! This week, I am short the 1000/1300 spread at $297/contract and added yesterday the 1045/1240 for $181 per.
Their stock is up 35x in price over six years, their pe remains the same!!!
Anybody who says it was an easy ride lived thru 2022 when in was down 75% from it’s high. My friends position dropped from over 100k to 30k. Yes it’s up 10x from there but it did take strong conviction to hang in there!
I’ve been an avid investor now for well over 50 + years. Yes, Iam truly ancient. LOL. I’ve been to this party before several times. But I will say this atleast right now NVDA is truly hitting on all cylinders. Like a finely tuned FERRARI. Here’s what you’re going to see happen down the road: Nothing grows forever & we all know that. Remember that NVDA has real earnings (EPS) so its not just some hype that you are reading about. Its a real company, with real in demand products and enjoying massive revenues and massive revenue gains along with massive increases in EPS. BUT, BUT, BUT when all that slows down in a meaningful way this stock (like many others) will get repriced in a southerly direction. The trick is how does one give off the “momentum train” before that happens. I have read almost countless research reports on NVDA and all the analysts are very GA GA over the company right now. I’ve seen EPS projections go out to 2029 with estimates well over $45 a share. I myself do own 300 shares but at some point I too will jump off the Ga Ga Bandwagon.
Other chip related stocks are also benefiting from the AI and Nvidia euphoria. TSM, MU, AMAT, LRCX are a few. It’s time to look at where they are headed in the next few months.
I don’t see NVDA commanding a $4-6 TRILLION dollar market cap at this point of their development but that’s me and they need to justify that to reach $2K a share. Right now, it’s almost pure mo-mo based upon a temporary “only game in town” stature. I lived through the 1999-2000 timeframe and heard just about the same justifications then as I am seeing with NVDA now. That said, the irrational exuberance can last a lot longer than we think (as evidenced back then as well).
Personally, I see NVDA more like a Sun Microsystems rather than an AAPL since their business is very commodity-like and not broad based services. But, even AAPL, who has that broad base, dropped 80-90% in value after the internet bubble burst – we know where SUNW and others like it are now. I would not be surprised to see the same thing happen to NVDA. The thing is, will that 80% drop be from $1200, or $2000, or even $4000? Time will tell because the AI hype is very similar to the internet hype then.
Ride that euphoria but be ready to pull the trigger very, very quickly when the music stops.
Just to show how crazy the numbers are around NVDA now – the company has a run rate of roughly $110B annually in sales now, and likely $25/share in annual EPS. At $1040, trades for 41X EPS.
But with 2.46B shares outstanding, it has a market cap of nearly $2.6 Trillion. That puts it at nearly 24X Sales.
I now must bring up one of the greatest quotes of all time that was said back in 2002 by Sun Microsystems CEO Scott McNealy when his company once traded for 10X sales:
‘At 10 times revenues, to give you a 10-year payback, I have to pay you 100% of revenues for 10 straight years in dividends. That assumes I can get that by my shareholders. That assumes I have zero cost of goods sold, which is very hard for a computer company.
That assumes zero expenses, which is really hard with 39,000 employees. That assumes I pay no taxes, which is very hard. And that assumes you pay no taxes on your dividends, which is kind of illegal. And that assumes with zero R&D for the next 10 years, I can maintain the current revenue run rate.
Now, having done that, would any of you like to buy my stock at $64? Do you realize how ridiculous those basic assumptions are? You don’t need any transparency. You don’t need any footnotes. What were you thinking?’—
Scott McNealy, Business Week, 2002
At some point, the hyperscaler companies that are spending the mind boggling amounts of money in cap-ex for GenAI right now (Alphabet $50B in 2024, Meta $40B in 2024, MSFT $50B in 2024, Amazon, etc.) will have to find significant new revenue streams to get some kind of return on their massive investments. What are the truly revenue generating uses for GenAI? Summarizing meeting notes and computer coding obviously wont be enough. MSFT is at least trying by making companies pay per user for their CoPilot software. But how long will that last?
So there is a very strong chance that this GenAI cap ex spending frenzy ends up as a massive capacity overbuild of computer servers, storage, etc.
Basically the same thing that happened with the dot coms back in the early 2000s.
My guess is those brave enough to buy long term put options on NVDA will eventually do extremely well. I haven’t had the courage…..yet.
Here is a counter argument… I work in the industry for a multi $B firm that sells engineering software into the Aerospace, Automotive, High Tech and other industries. Our software is used to deisgn every car, airplane and high tech device that is manufactured today.
NVDA has a monopoly in a core technology which is being rapidly adopted by ever major firm on the planet. When have you last seen a monopoly like that?
NVDA is selling a portfolio of expensive products into this demand with a very clear and (we are told) backward compatible upgrade path. Which other company has done this in this way?
NVDA is executing flawlessly. How many other companies that you know if in history have executed this well?
NVDA has a very broad eco system that spans software developers, chip manufacturers, industrial clients. It is very broad. Name another company like this?
What multiple should one put on that? What terminal growth rate do you put in your DCF valuation model?
Intel prior to 2000
hey Charles,
Certainly are similarities, but it’s not exactly comparable because of the speed of adoption and the competitive landscape. AI is being adopted now and AI applications are being rolled out quickly.
Tesla puirchased 35K NVDA chips to put into cars today. I dont think Ford/GM did that with INTC back in 2000 – at least I don’t remember.
As I recall INTC had competition. There were also Motorola processessors and competing models for computers such as mainframes, and Unix workstations. I recall Apple had a large workstation. Sun had a SPARC processor as I recall. There were lots of computer models that were not on the INTC platform is the point.
McNealy had Silicon Graphics & HP (for example) as competition. I worked in the software business and our software ran on at least 10 different Unix systems that competed with Sun and none of them had INTC processors.
Today NVDA has literally no competition. It is an actual monoploy.
Lastly – every single CEO on the planet knows that he/she needs an AI strategy and they need it now.
Speed of adoption is not comparable to 2000 and competitive landscape is not comparable.
Would I buy NVDA today?No.
Am I selling it today?No.
Is NVDA going to last forever? No.
Will you make more money selling NVDA puts than buying them? Most likely.
August:
“Tesla purchased 35K NVDA chips to put into cars today.”
That is wonderful for NVDA. But please tell me how TSLA will generate increased revenues on this? Last I checked they were lowering prices on all of their models and laying off thousands of employees. They are lining up at the trough of Congress begging the Administration not to allow Chinese competitor BYD into the US market (and it is working).
At some point you would have to think all those spending these money piles are going to realize they are getting little to no return on their investment.
But no doubt NVDA is the clear winner in the AI frenzy/bubble for now.
Sure
You quoted me correctly, but my statment was not correct. I misunderstood the headline. To correct it – Tesla are using 35,000 NVDA GPUs in a compute cluster which is used to train self driving algorithms which are to be used in autonomous driving. Evidently Chinese EV manufactureres are buying NVDA chips as well https://www.investors.com/news/tesla-stock-nvidia-stock-jensen-huang/
According to NVDA CEO he sees Auto sector as his largest growth potential vertical segment.
How do these automotive OEMs make money off this? No clue. I don’t invest in auto oems becasue of the competitive landscape.
My concern is NVDA. It is pretty clear how NVDA make money from it, and it is also clear that there is no competition.
As you say – for now.
August, here is how bubbles like this always end. At some point companies start reporting that their AI investment produced too little income for the money invested. Everyone will have enough capacity already. Nvidia chip and software orders will be cancelled. Suddenly the explosive growth numbers will look to have peaked. It happens fast. But of course we cannot predict when this will happen.
This happened in Y2K, the dot com, and every other tech bubble. Suddenly orders dry up. Scott McNealy saw this 1st hand, when sun microsystems went from a shooting star to nothing fairly quickly.
The time to short or buy puts is AFTER the stock starts correcting, then retests the previous highs. If it makes new highs you stop out. Good luck out there.
Thanks for the update about Sun Microsystems and the technology cycle. I recall having one of the fisrt SPARC workstations and I also had the stock for a while.
I also had an SGI workstation, an HP, Digital Equipment all around the same time. In fact I can recall carrying SGI workstations on airplanes from Detroit to Dallas to give demos. Lots of expensive equipment to carry on a plane and lug through an airport. So… PCs happened and Windows 3.1. Then laptops. The Unix Workstation went obsolete. Sun could not transition and their cloud compute model was way to early for its time. So it could not take up the slack.
Now look at NVDA – where is the PC equivalent? Where is the Windows 3.1 equivalent? It does not exist, and is years away.
You seem to miss the point that NVDA is a monopoly.
Will it be a monopoly for ever? Probably not.
Is it a monopoly now? Yes.
Anyway feel free to wait for a spot to buy puts. I would only wonder if you managed to buy and sell puts in March? If so good for you. For my part I used to pullback to add to already long positions.
NVDA has done some really aggressive window dressing accounting maneuvers around their quarterly announcements in recent quarters.
(this was around their Graphics cards business not the AI chipsets)
The question is why did they bother?
This is a case where the people who make money off the stock falling in price will need to be more lucky than smart.
Tim, I am the one that has worked as a developer and we chatted on email. I am also annoyed about hearing about AI and also agree that lots of articles are junk, there are even YouTube videos that are being created that are almost undetectable AI but of no value.
I am writing this to say though that it is amazing and you can run open source models for FREE on your desktop. Many models have less restrictions and are trained on different data. You can use AI and create “agents” to just about any type of activity or data gathering. It’s mind blowing and exhausting at the same time.
yield hunter –
we had a business in China (started several years ago) that generated videos (based on generated scripts) on a lot of “basic” things. Amazing how good the avatars were – looked/sounded very human. we distributed them via Chinese social media. Some were fairly technical (like intro to blockchain). I thought they were pretty good because we had knowledgeable humans review the scripts to avoid some of the silliness AI sometimes generated.
People who wanted to get a basic understanding of a subject could learn by watching our videos. The business made money.
AI is great for those kinds of projects because it can gather and synthesize existing information (“backward analysis). What it can’t really do is forward analysis – coming up with that “epiphany” idea of what might happen next.
We sold that business. We could see that the huge amount of garbage being generated without any human oversight was going to swamp good content, and it seems to have done so.
“AI” has been around a long time…… I use MS Word for my writing and have long used spell check. It works quite well. As I write this post on a Win 10 machine I see my fat finger typos highlighted also. I’m a terrible typist. Not so great is the grammar correction / suggestions. I watch it very carefully. My Google smart phone uses spell correction and suggestion, do I have to watch that! I have had some whoppers get out to folks!. AI has long been in other applications also, it has just been somewhat subtle. It has recently spread to the hated (for me anyhow) automated telephone answer systems and now to the cold calls you get constantly about things like Medicare insurance supplements. I can spot them easily. Just respond with an off the wall unrelated comment or question and watch it struggle or not answer! More concerning is where AI has advanced to the point it can imitate me. I can see scammers taking huge advantage there. That is scary. Same goes for video you see. I am getting to the point I trust nothing…….