1) We've all been bombarded with stories about how artificial intelligence (AI) is "the next big thing," so of course every company is trying to tie itself to it to juice its stock...
My friend and market veteran Doug Kass of Seabreeze Partners shares my skepticism that the run-ups on most AI-related stocks is warranted, as he explains in this recent essay:
Don't Attend the AI Church That Is Happening Now
- The Dot.com era had only one dot – the A.I. mania has two dots!
- Roger McNamee chimes in on CNBC yesterday with some similar observations to mine
- Learn from history
- Avoid emotion and the breathless commentators on FinTV who know everything about price but very little about the subject of AI
- Why Nvidia (NVDA) wins – though it may already be discounted – but MSFT, GOOGL, AMZN and META may not
You know it is a mania again when, suddenly, almost everything with a ticker symbol is an AI Play – just like the amount of stocks that became block chain plays or VR plays or metaverse plays or COVID plays or Cathie Wood's "disruptive stocks" – no matter how little AI impacts or even hurts their P&L.
There is a difference between the AI mania and other manias – the AI mania started after a big rip up in the stocks and without rates being at a bottom.
I cannot tell you what inning we are in, and how many games are in the series. Certainly, the AI exchange-traded funds are sure to come. Cathie Wood doesn't own Nvidia. The funniest part of the whole thing, and of course she is out panning AI since she doesn't own NVDA. When does she roll over and start buying it?
From my perch NVDA is a bona fide AI play, as a large percentage of their revenue, and more importantly incremental profits, comes from AI.
As Doug explains, AI might simply be a huge expense for tech giants Alphabet (GOOGL) and Microsoft (MSFT)...
Are you paying anything for ChatGPT? But MSFT and GOOGL are spending enormous amounts of money producing ChatGPT (and Bard). These NVDA H100 products sell for $270,000 each – I'm not kidding, I have never seen anything like it. This is another impediment to growth, as only a few companies can spend at scale on this stuff. The 8 GPU baseboard alone is $195,000. They make about $190,000 of gross profit per H100 sold.
Should China and the U.S. go to war, it wouldn't be good for China or the U.S... But it would be good for whoever is selling them the weapons. This is not to say MSFT can't be a good stock, and I guess they will get some sheen from AI, but look at their P&L and where earnings are coming from. It is not this. AI is purely a huge expense for them.
I don't want to hear anything about 5-10 years from now because nobody knows what will be going on 5-10 years from now. The same people that make that argument are the same ones who trade the stock on every single data point and never look past a quarter, Nobody does because if they did, explain why Peloton (PTON) ripped up so much when it was obvious what was going to happen!
Not to mention, the theoretical 5-10 years from now does nothing for the economy today. Nor can you forecast the impact of these unknowns for the economy either. Again, this is not to say MSFT can't be a good stock, which is a separate issue. Clearly, with regard to their actual real business where they generate actual money, they have been executing well, especially in the cloud, where they are now gaining share.
Is Amazon (AMZN) somehow an AI play now? I think not. It's a retailer and cloud service provider. There is less than meets the eye to their retail business. Its growth has slowed substantially, and they struggle to make money consistently. Their higher growth and more valuable cloud business is now losing share to MSFT, and growth has meaningfully slowed. I think they guided to only single digit Y-Y growth for cloud. But the AMZN multiple remains high.
Meta Platforms (META)? In addition to the name, their business model seems to change every other week based on what people think will make the stock work better. Facebook (FB) is an advertising business. That is where their economics come from. Lucky for them they own Instagram, because core FB is tired. It might help to explain why the company tried the Metaverse stunt. Anyway FB has somehow gone from a mobile play to a VR/Metaverse play, to a cost savings play that was legit, and now, somehow, an AI play. World record. Maybe FB wants to sell ads where ChatGPT directs people to a fake business that sells oranges instead of the apples the customers are told it sells?
I can go on and on. Eventually the rubber meets the road. It will be interesting to follow the revenue and P&Ls of these companies and determine what is really driving the bus in terms of real reported numbers, as opposed to hype. And what the rates of growth actually are.
Doug concludes that while some companies will benefit from AI, far more will hype AI even though it's irrelevant to their bottom lines, writing...
GameStop (GME) might turn into an AI play, you can draw a linkage, those Reddit guys are creative. But eventually reality will hit the numbers.
The other thing I wonder is AI just a substitute for normal search? It might help MSFT gain market share from GOOGL. But in the case of Google, it seems more cannibalistic than incremental. It just substitutes for text search. Then the problem is how to insert ads into the results from ChatGPT? Seems somewhat challenging. Will Chatbots be further manipulated to favor who is paying the provider? Then from a customer's perspective they really lose utility because you cannot see a full set of results, thus the utility really drops.
Given the massive costs of AI delivery, MSFT and GOOGL will probably have to start charging money for it. Then what happens when it is no longer free? Once it's no longer is a toy, it had better have huge utility – and right now, for most cases, it really doesn't.
Frankly, besides being fun, it is probably productivity decreasing in a lot of cases because you need to double check everything it spits out, like this lawyer should have done: Lawyer apologizes for fake court citations from ChatGPT!
Finally, Roger McNamee was interviewed yesterday on CNBC. Here is the tape! McNamee views AI as a mania. He sees OpenAI as not having an attractive business model. It is a very good interview.
Thank you, Doug!
2) In Friday's e-mail, I shared some tweets with various economic indicators, most of them negative, so today I wanted to share two positive ones.
Here's great news about labor force participation from CNBC's Carl Quintanilla:
And here's great news about a huge increase in manufacturing construction spending:
3) Holy cow, this Bloomberg story is insane! Hedge Funds at War for Top Traders Dangle $120 Million Payouts. Excerpt:
When portfolio manager David Lipner said he was quitting billionaire Izzy Englander's Millennium Management to join a rival, the hedge fund countered with an unusual proposal: A one-year paid sabbatical and an incentive upon return if Lipner stayed.
And stay he did. For Millennium, the $58 billion industry giant known for ruthlessly cutting underperformers, the generous offer was seen as totally worth it. After all, Lipner had made money for the firm for more than a decade, longer than most hedge funds remain in business.
Such enticements are now becoming part of a growing array of expensive tools the world's biggest hedge funds are deploying to hire and retain traders. They show how a limited pool of talent and surging demand for steady returns in a volatile market are prompting firms to pull out all the stops to attract the best — with clients footing the bill.
The hunt is no different from the bidding war for Premier League or NBA players, one executive said. Last year, a senior portfolio manager was lured by a major New York fund with more than $120 million in guaranteed payouts, according to a headhunter who said he'd done several deals paying north of $50 million. Contracts worth $10 million to $15 million are increasingly becoming common for traders, said another.
Hedge funds have long been the land of eye-popping rewards, but a few recent trends are converging to take it to new levels. The stellar track record of several giant firms that spread money across teams of traders following multiple strategies has caused their assets to surge. That's prompted a hiring spree to add more traders and strategies so the existing ones aren't over-stretched.
The performance — and resulting wait lists of investors wanting in — has also given the firms leverage over clients to charge many times than the traditional 2% management fee and use that for recruitment and retention. And as the firms increase rewards and defer more of them over several years, it's taking even bigger offers to tempt traders into leaving.
4) I suspect my share of this will only be a few dollars, but it took me less than five minutes to fill out the application form (here), so I figured why not? How to Claim Your Share of Facebook's $725 Million Privacy Settlement. Excerpt:
If you used Facebook in the United States between May 2007 and December 2022, you can apply to claim your share of a $725 million settlement that Facebook's parent company agreed to pay to settle a class-action lawsuit, according to a claims website set up by a settlement administrator.
Users can enter their information on facebookuserprivacysettlement.com to get their payment through their bank account, Venmo or other methods. The size of payouts is likely to be small, and it will depend on the number of people who submit valid claims and the length each applicant was a Facebook user during the period covered by the suit. The payout will be divided among claimants, with more given to those who have used the site longer.
Best regards,
Whitney
P.S. I welcome your feedback at [email protected].
