원문 보기
With us is Jeremy Siegel, professor emeritus of finance at the Wharton School of Business at the University of Pennsylvania and chief economist at Wisdom Tree. Jeremy, we meet again. I remember how many times you were optimistic against all odds, and it paid off. When you were last here , I think I said you were being pretty cautious, and the market continued to rise.
You're still... I think you were talking about inflation. Well, that's never good . You talked about rate hikes, about Iran, oil prices and all these things. And it made you think. Are you surprised by what the S& P has been up to recently, the last few sessions? How much did we have, 10 consecutive weeks of upward momentum?
Did that make you think that the moment of truth was approaching, or that it would continue to grow? Well, Joe, I don't think anyone was surprised. If you had told everyone, even Tom Lee, three months ago that oil would be between 90 and 100, that the straits would be blocked, that there was little prospect there— would the S&P have gone up 10-15%?
The answer would be: this is madness. Now there are many workarounds and reserves. China stopped importing 5 million barrels per day. We don't know if we've hit a wall where suddenly everyone will say, "Oh, the stock depletion is over," and that's why the market has perked up. If we're close...we all know Trump has said many times that we're close to it, but I thought the war was starting again a few days ago, but they both stopped.
So maybe we have a start, which is certainly very optimistic. I think, as you mentioned before the break, SpaceX will obviously take up the whole week, but next week is the Wars meeting. Wow. I do n't recall, I probably have to go back years, if not decades, to find something that would be more interesting and influential to me than what he's going to say.
I know, if it were me, I would say, you know, Mr. President, I, thinking, well, I don't... you know what I mean? It's too late to retreat. I think so, but I, um, don't envy what he has to do. I'm curious what you attribute this upward movement to, against all odds, Jim. And we, you know, I know you've read " Memoirs of a Stock Trader." That's the job of the market—to do it and try to go up when so few people believe in it .
I think they're in this, you know, "prove it to me" period right now, right? Show me what you're doing with this AI spending. Show me how you're going to monetize this, generate revenue from this . And I think it's an interesting tactic that they've chosen, perhaps replacing some of the spending that was allocated to Anthropic and OpenAI and focusing on their own AI models .
I think it's an interesting move, and we'll have to wait to find out a little more during the earnings report to understand what those costs will be. But yes, they are definitely being punished for being part of this software sector. The stock is down 20% year-to-date, and that's actually 33% below its 52- week high of around 550, well, not that long ago.
So, where we are now, this could provide a good entry point for investors who believe, as you mentioned, that if this is overblown and Microsoft is being undeservedly punished, then this is a good chance. So I'll leave it at that and get Kevin's opinion since his surgeries and mine are similar today.
So maybe we'll listen to what Kevin has to say. We can go through the agreements. Kevin, I'm also curious to know your thoughts. I think the question that investors are asking today on the trading floors and in hedge funds is: is IBM's problem a software problem , or is it just an IBM problem? Because the software was under pressure.
It has held and bounced back in recent days. We mentioned on this show the other day, Diana, remember, they were selling memory and buying software. So a lot of these software companies have bounced back. Not today. Microsoft, Oracle, Salesforce, Adobe, Palantir look like they're coming in positive, but Service Now, Adobe, they're all in the red today.
So, software is taking a hit again today. But you mentioned, you know, Microsoft just keeps moving forward, bringing more AI inside the company, laying off employees from Xbox, directing assets towards AI. They're very consistent and they have, remember, a product portfolio. So they don't all depend on AI or on those parts of the US economy that are currently suffering.
And so I mean, three price targets were changed this week. Argus, [ __ ] Capital and Wolf Research. All target prices are above $500. And that's why no one wants to lower their target prices for Microsoft. And so remember it hit $555, you know, it dropped significantly. It seems like it's starting to hold or bounce.
Today it dropped to 378, and it has already recovered from that low. So, one of the clearest messages from the latest round of tech company reports is that the building of AI infrastructure is far from complete. On the contrartarting to act even more aggressively. Every reporting season, investors ask the same question.
When will hyperscalers slow down their spending on AI? And so far, the response has been surprisingly consistent . Not yet. Alphabet continues to expand its AI infrastructure. Tesla is increasing its investment in computing, while cloud providers continue to report strong demand, and executives describe an environment where customer demand still exceeds available supply, with huge implications for the semiconductor industry.
For much of last year, investors worried that the AI trend would eventually collapse under its own weight, and the fear was always that companies were spending too much money too quickly. Eventually they will realize that there is no return. Capital spending will slow, demand for chips will weaken, and the AI investment boom will come to an end.
But recent comments paint a completely different picture. Instead of discussing cost reductions, executives continue to talk about supply constraints, and instead of describing excess capacity, they explain why they still need more infrastructure, and nowhere is this more evident than in the memory market.
Memory has quietly become one of the most valuable assets in the field of artificial intelligence. Without sufficient high-speed memory, even the most advanced AI processors cannot operate at full capacity. This is why companies compete so aggressively for suppliers. Tesla's earnings report perfectly highlighted this dynamic.
Elon Musk directly thanked Micron for helping secure future memory supplies, acknowledging that the supplier must carefully decide which customers will receive limited production. He also called current memory prices extremely inflated. These comments reveal something important. Memory is no longer treated as a mere commodity.
It is increasingly becoming a strategic resource. Companies don't wait until the last minute to purchase supplies. Instead, they negotiate years in advance. Multi-year agreements are becoming commonplace as customers simply cannot risk being without one of the most critical components in the AI supply chain, marking a significant shift from previous memory cycles.
Historically, manufacturers produced memory, distributors sold it, and prices fluctuated depending on normal supply and demand conditions. Today the market looks completely different. Customers try to guarantee access to products even before production begins. This changes the competitive environment.
Companies with guaranteed supplies gain an operational advantage. Companies without long-term agreements may find it difficult to obtain sufficient quantities of components during periods of shortage. This naturally benefits companies capable of producing high-quality memory for AI. Micron continues to expand its presence in high-bandwidth memory , while Samsung and SK Hynix remain the major suppliers in the industry.
Demand has become so intense that many analysts now believe that premium memory will remain in short supply for a long time to come. Of course, investors should avoid assuming that the deficit will last forever. Any deficit ultimately leads to additional investment. Manufacturers are building new factories, expanding production and product range. Supply gradually catches up with demand, and over time prices stabilize.
Comments 0