芯片股延续跌势，零跑汽车财报受关注

[music]

Bloomberg Audio Studios. Podcasts, radio, news.

[music]

Welcome to the Daybreak Asia podcast. I'm Doug Krizner. [music] There was a cautious tone in trading of US equities in front of those earnings from Nvidia that are due Wednesday after the bell. Now, you may recall last week we learned Nvidia had warned some big customers of price increases on servers containing Nvidia's AI chips. Now, on the US hunt Monday, Nvidia shares were down for a seventh straight session losing 2.9%.

We also had the Philadelphia Semiconductor Index dropping about 2.7%. This morning in South Korea, we're seeing declines in both SK Hynix and Samsung shares. For a closer look at the price action, I'm joined by Bloomberg's Mark Cudmore. He is global leader of our Markets Live Strategist team. Mark joining from the Lion City of Singapore. Thank you for being here. Talk to me about what you're seeing in this AI trade, particularly as it relates to the chip stocks.

The AI trade is getting a lot more complicated. And what I mean by that is that for a macro tourist like me, it's becoming much more important to do your your kind of bottom-up work and to to really differentiate between different sectors within the tech industry. In fact, in the same way that like we took about 15 years to realize that the term EM, you know, emerging markets as an asset class no longer worked because the divergences were too big.

I think we're we're we're at the start of that same process in tech where we kind of still refer to the tech sector. And and in reality, you know, the tech sector diverges massively. And I think yesterday's session was a it was a perfect example of that. You know, the really the the biggest underperformers and out performers were were different sectors of the the tech band overall. And and so I think that really applies to what's happening in the AI trade.

So that the people who are doing best in this market, which is why partially why retailers are doing so well this year, because retailers are doing the bottom-up work and they're getting excited about the stories at different stages of the AI theme, rather than just being blanket going, "It's an AI trade. I'm bullish, bearish." And that's not the way you can look at it anymore. So, I think that's probably the biggest takeaway at the moment.

I I I I think the backdrop picture is I am still in that we are on an incredibly massive AI capex bubble. We're still in the inflation stage that bubble, which is still bullish for the broader market cuz it's pumping money into the system. I still believe the AI capex bubble will collapse at some point and that's going to have severe ramifications across global markets and economic ramifications. It's important to emphasize that the bubble is in capex.

It's not a bubble in stocks. In fact, the stock market the macro environment of stocks is still very bullish overall. Many stock markets around the world are not expensive. So, it's specifically a capex bubble, but certain stocks will suffer horrendously when the capex bubble ends. That's the backdrop picture and then we've now gone to the more volatile kind of end stage. However long that lasts, I don't know. Many multiple months is is what's expected.

And it within that, it's become a micro game of bottom-up differentiation.

Does that mean that the hyperscalers in your view are most at risk right now in the cycle that we happen to be in?

So, I think the hyperscalers are are probably some at risk from the AI capex bubble, which is why people associate that, "Oh my god, when you say AI capex bubble," they immediately mistakenly think that's a bearish call. It's not. First of all, if you find that one of the basic rules of markets is if you find a bubble, run towards it and get in board. Now, you obviously hope you get out as soon as it pops or early on when it pops.

But the point is is that the whole point of bubbles is that they see extraordinary gains. So, first of all, uh, the the the hyperscalers are associated with the capex bubble, which means that, uh, I think people are kind of going, "Oh my god, at some point the capex bubble is going to burst." Now, when it bursts, the hyperscalers, uh, some of the hyperscalers are very vulnerable because they're the ones that are pouring money, they're the ones into this capex.

They're the ones who are like have gone from from the the these kind of cash-generating uh, machines with so much free cash flow to like just spending so much on building these data centers. And they're borrowing more and more money to do it. So they've completely changed their fundamentals. So they are the most vulnerable. However, the reason I wasn't kind of asserting that immediately, they're the most vulnerable to the AI capex bubble collapsing is that's kind of priced in a lot of these names.

Like people have been been negative hyperscalers for a while. Hyperscalers have underperformed much of the market for a long period of time. So it depends what your your trading horizon is. You know, if you're like me and you don't think the AI capex bubble collapses soon, then I I think people have got too bearish hyperscalers relative to the rest of the market too early. And I think that actually potentially they can have a period of outperformance.

But if you're a long-term investor, yeah, sure, you don't want to you don't want to try trading this end game.

That's a nice segue into the story on Alibaba that we were discussing on Monday in the Asia region. Alibaba is looking to raise a little more than 10 billion US from a secondary offering, and apparently these proceeds will be used to invest in full-stack AI capabilities. Right now, this company is ratcheting up capex to nearly 10 billion dollars. I understand how competitive this space is and that you want to really fiercely defend your position, but is Baba at risk of maybe getting a little over its skis here?

I mean, it's the same kind of story as the US hyperscalers, it's just on a much lower scale. And actually, I find again, uh, the the the Chinese names are probably they're they're they're less uh, you know, committed to the quite the scale of capex and and they're already much more discounted. So I'm I'm less worried about them. I actually think on a longer term basis, you know, the these these companies could do fine.

I don't I don't have any tactical view on kind of single names, but I think that the Chinese AI story is much better priced for the longer term and therefore in a longer term where I think it's a much better story.

So I want to get your views on what we've been seeing play out in the US Treasury market. Particularly since last week Treasury Secretary Besson unveiled a plan to try to get longer term government bond yields at lower levels. And he's going to do this. He's going to attempt to do this through expanded buybacks of longer dated government securities. Monday we heard from Besson. He said the Treasury has yet to buy a single bond and he kind of I think refrained from sending signals about revamping US debt management.

What do you see playing out in the bond market right now and could that be problematic going forward?

So I think his kind of manipulation of the bond market at the moment is to kind of squeeze some kind of short positioning out there. It's you know, some people have called it an operation twist redux. It's quite pretty marginal in terms of amounts and as as you kind of pointed out there, you know, the first buyback I think is September 9th. So it's still a couple of weeks away. Um Ultimately, I think how this has been executed in in my view is that this was if you're going to intervene in the market, do it properly.

Like don't do kind of a a job that gets questioned and you have to follow up with multiple interviews to kind of complain that the market isn't getting the message. If the market didn't get the message, then you didn't do a good enough job. So I kind of feel that n