Hi Multis
I hope you had a fine week, despite the market drop.
These drops never feel great, but without them, the market wouldn’t be healthy. It’s like your energy level. Your energy drops when you exercise, but over the long term, you will have more energy if you work out.
Of course, more pain could be coming. We don’t know. But I do know that he who persists will eventually win. So, just stay calm, stay the course, grab a cup of coffee, and read this Overview Of The Week.
Articles In The Past Week
This is the fourth article this week. Let’s look back at the three previous ones.
The first article this week was completely exclusive for paying members. So if you are, you can find it in your email inbox.
On the 15th of the month, you get the Best Buys Now. This month, all six stocks are Potential Multibaggers, which is pretty exceptional. You can find out which stocks I see as the Best Buys Now here.
On Saturday, I published the article about Datadog. Revenue grew 36% YoY in Q2 and the growth keeps accelerating. But after the sell-off of this week, the stock is down 12.8% from its pre-earnings level. So, is it a buy now? You can find the answer in this article.
Memes Of The Week
No memes this week.
Interesting Podcasts Or Books
I listened to the two podcast episodes I talked about last week, but nothing else worth sharing.
I got this question in a DM from one of the Multis this week: “What investing books would you recommend?” And of course, I answered it. You may be interested as well, so I share it here.
The markets in the past week
I think most of you will know that the market was red this week. Let’s see how big the damage was.
As you can see, the Nasdaq was the biggest loser this week, down 4.48%. The Russell 2000 was down 3.05% and the S&P 500 dropped 1.95%. Pretty big moves for the indexes.
The Greed & Fear Index dropped from NEUTRAL last week to EXTREME FEAR.
Quick Facts
1. A Bear Market Below The Surface
Semiconductors had a rough week. Or rather, a rough two weeks. The SOXX semiconductor ETF fell 9 days in a row, from October 6 to October 16. That is the longest losing streak in over 20 years, with a cumulative loss of 16.31%.
Rumors are circulating on Wall Street that the VARA AI fund (Value Aligned Research Advisors) blew up. This fund has $26.4B in assets under management. If the rumors are true (and often they are), then this could definitely be the reason for the sell-off or at least speed it up. There are also rumors that Leopold Aschenbrenner’s fund, Situational Awareness, blew up for the second time. If that’s true, it would accelerate the sell-off even more.
VARA has positions in Nebius, Bloom Energy, Lumentum, CoreWeave, Iren, bitcoin miners and other AI infrastructure plays and Situational Awareness in SanDisk, Micron, Bloom Energy, Nebius, CoreWeave and IREN, among others.
This Goldman Sachs chart focuses on the valuation of semiconductors. The forward P/E of the PHLX Semiconductor Index peaked at 38x two years ago. Now it’s at 18x, the lowest it has been in more than three years.
And what about memory? Michael Dell says Dell has $95 billion in backorders for AI server racks. The real bottleneck is getting enough memory chips to deliver them. And next year, “It’s gonna get worse.”
The different elements look confusing and may suggest memory is undervalued. Maybe a part of the answer is in another Goldman Sachs chart. It estimates how much hyperscaler capex contributes to S&P 500 EPS growth: 2 percentage points in 2024, 5 in 2025 and 11 in 2026. After that, it goes: 7 points in 2027 and -1 point in 2028, as depreciation catches up with capex.
So, semiconductor stocks fell 16% in two weeks, and memory even more, while the biggest customers say they can’t get enough chips. It’s confusing, especially when you add Goldman Sachs data suggesting capex growth from hyperscalers will disappear.
I also saw this.
So, maybe the market already anticipates the next quarters, where growth is still strong but declining. Of course, if you look at the past quarters, the actual results blew the expectations out of the water. I think that’s what the market will need. The earnings season and the midterms coincide in a few weeks (mostly early November) and the two together will determine the short-term trajectory of the market.
It would be a mistake to think that this is just semis. The indices still look OK, more than 5,000 stocks are down 20% or more from their 52-week high. Nearly 4,000 are down 40% or more.
And if you look at the biggest stocks, those in the S&P 500, it doesn’t look great either.
Last week, we saw that 134 stocks in the S&P 500 are down 40% or more from their all-time highs. This is the same picture, but broader. And it’s not just the US. The Indonesian stock market had its worst crash in 28 years.
It looks almost like a black hole. Ongoing fires in Indonesia and the government’s denial of a disaster make investors bearish. People have been breathing in the smoke for months now.
How about valuations? The Nasdaq 100 trades at 23x forward earnings, above its 20-year median of around 20x. But the equal-weight S&P 500 is at 15x, right at its median, and the S&P Midcap 400 at 15x is even below it.
And then there’s the inevitable bearish prediction. Paul Tudor Jones said on CNBC that “2027 is going to be really, really challenging for the stock market“, with the most negative flow picture since 2002.
Take those predictions with a big grain of salt, though. Jones was also on CNBC in October 2023, when he said a recession was likely in the first quarter of 2024 and that stocks would probably fall about 12%. Instead, the US economy grew 2.8% in 2024 and the S&P 500 gained 23.3%.
This game is all about ‘Told ya.’ 12 wrong predictions don’t count for these people if they have one right, they will tell you: “I told you so, but you didn’t want to listen.” And the weird thing is that many media outlets celebrate this as if these people have something special, almost god-like. Well, I can predict that there will be a next bear market. If I repeat it enough, I can always point at it and say: “Told ya.” But that’s fake.
Below: the AI debt is quietly cracking, the one neocloud I’d pick if I had to choose, the a16z founder calling a frontier model company “one of the most negligent we’ve ever seen,” why a cloud giant just kicked one of my picks out of its own conference (and why I think that’s a gift), and much more. Don’t want to miss anything?
