In my time investing (I’ll date myself here) my first ever market hype was fiber optics!!! FIBER. OPTICS. !!! It will change the world!
Next came things like…3D printing (it will CHANGE the world!), then came drones and now it seem AI is the all new hype.
The reason I started thinking it is the next tulip bubble to avoid is that we have soooo many conversations about AI on this board, and how our companies will die or fly with what they do about AI. The article is ranging and not totally related to the title, but there is this part:
"While AI could lead to some substantial efficiencies that help fight inflation, it’s likely no match for the earnings recession that we forecast for this year,” Wilson said.
The analyst said he believes the rally will prove to be a “head fake” as seen last summer, given the unattractive valuations, too-optimistic earnings estimates, investors’ unrealistic expectations regarding a Fed rate cut, and the markets turning a blind eye to key risks such as the banking crisis.
All the things in the past that were the next big thing, still exist, but I never made money on any of them. They were also still “stories”. The vision for AI is still being created and I would not invest in ANY company just for AI. If they are working on AI, like Nvidia, and have many other business lines already making money…cool. But the hype of AI has not made me buy any Google or Amazon or any others talking up their AI.
Sidenote and OT - UPST, yeah, I have history there and I do see evidence that using machine learning (not AI) to crunch algorithms and determine default risks can work. But this is where we have to be pedantic about AI or ML.
Probably better to change our terminology - the big wins this past week have all been based on datacenter retooling - instead of AI the software, it’s the hardware providers that are currently adding gobs of revenue to their forecasts and guidances. I mean, adding 50% QoQ growth to revenue at 7B quarterly run rate is ridiculous - Marvell had terrible results after close today but is up because they are projecting a doubling of revenue in their data center segment (which makes up 40% of revenue). I don’t see how NVDA can keep up 50% QoQ comps, but we could at the same time be overestimating the actual revenue impact of AI on SAAS companies while underestimating the revenue impact of AI demand on these “picks and shovels” providers to datacenters
OTOH, what about DVD by mail/streaming? What about cloud computing? What about electric vehicles? Those have/are changing the world.
Today, looking at how NetFlix, Amazon, and Tesla have done I’m not as quick to write off AI as the “next tulip bubble.”
Of course, investing in it could be like investing in solar energy, which has been disappointing for the past couple of decades, with companies like FSLR peaking 15 or so years ago and Solar City having to get bought out to prevent failure. Solar is around to stay, but there has been no clear winner in the stock market, and even today looking back I’m not sure what one could have done to really profit off of it.
So, even if you agree with me that AI is world changing technology, picking the winner companies doesn’t seem simple. The “picks and shovels” plays may either not be really known, or if really known have already had big associated stock price reactions. How AI will affect other businesses is not yet clear (at least to me). Is DDOG really looking at it only as more web sites to monitor and not incorporate AI into more intelligent monitoring products? Cloudflare seems likely to have some products using AI, but can they make a successful growing business out of them? Will some scrappy startups without legacy code be able to come up with whole new products that will revolutionize some industries?
This is where my opinion differs the most.
For those who bought SMCI like I did, last week- which one reason I wrote about explicitly buying it was to help ride the AI bubble trend - we are already up 28-30% in a single week. Isn’t that real money made?? We just have to be selective with the stocks chosen. Don’t buy purely on narrative hype. Buy those companies who also have real fundamental business growth and lower relative valuations to others. It gives you a safer floor. That means to play the AI ride, I think it’s safer to hold SMCI than, say, UPST; recall UPST’s claim of AI revolutionizing lending but they remain in hypernegative growth
I believe the key lesson I learned, particularly from the excesses of the 2021 peak bubble period and subsequent 2022 crash, is to embrace the inflating bubble, not to shy away from it. That is where excess returns can be made.
What went wrong was when I stuck my head in the sand and refused to believe it was even a bubble last year and clung to ever growing losses- that was a total mistake to not exit when the bubble starts to pop.
This whole part strikes me as too much market timing. There seems to be a whole lot of using hindsight to prove something that is extremely hard to repeat. You have to get in at the “right” time and you have to get out at the “right” time. That is trading and not investing.
I am just saying that to me AI seems to be the newest fad.
OH DANG! How did I forget the marijuana fad? GAH, I lived through that too. LOL. Anyway, we are probably about to get OT on this thread.
Yes. I agree with you. I would also argue everything done by Saul and others on this board is pretty much a form of ‘trading’ anyway. We can call it however we want as long as it works for us.
Market timing doesn’t have to be perfect though. You don’t need to exit at the tippy top. For example with UPST, we didn’t need to sell at $400 in 2021. Many of us got out on the terrible November 2021 earnings report- I believe I was out at $280 or so, and still reaped incredible profits.
And another example is Zoom. I was never in it but looked like Saul and others profited handsomely from that bubble.
It is certainly veering too off topic- hence why I have written things off of Saul’s board.
A well-timed post, I have been thinking similar thoughts. I’m firmly in the camp of “AI will change the world”, but these things never happen on a timescale anyone can truly predict. Some aspects of it (increased efficiencies and productivity in existing processes) will take years to play out, because use the people who will leverage these tools will either need training, or are just graduating from high school today. Other aspects will be very quick, like artists already licensing their voice or likeness for others to use in their own art. It’s the FOMO that pushes us to make rash decisions…some of us will only be correct in hindsight, but that’s not a game I want to play, and least, not more so than is already required to invest at all.
There’s the rub: only hindsight will tell you if it was still inflating when you bought/sold it. You can’t know in the middle of inflation that it isn’t about to pop.