00:00It was great to get you on the show today because I saw yesterday you were quoted in our stocks wrap, which is very widely read, saying essentially that if you don't have a decent allocation to these AI stocks now or to these MAG7 stocks, it's still not too late to get in. Is that the case?
00:16Yeah, we believe, in fact, that AI is the most profound innovation in human history and also one of the biggest investing opportunities. But I think one thing is also clear that AI is the new frontier, even between countries, between sectors and companies.
00:32And I think even the new divide within the MAG7, if you look at the performance this year, we now have three of the MAG7 companies underperforming the NASDAQ 100.
00:42And we believe that there is going to be a new group emerging, which may be the AI7 that is an enabler of AI and also can profit from AI at scale.
00:52So I think that is that is one of the key questions in the public, but also in the private markets.
00:56For these stocks, just discerning who is benefiting from AI and who's just, in the words of Dan I, is putting together a good presentation that makes it sound like they're AI players.
01:04It's complicated and it's difficult. How much of a specialist do you need to be?
01:08If you are an equity investor, how deeply and intimately do you need to be able to understand what's happening in AI?
01:14Now, if you think about AI having driven two thirds of the returns in the S&P 500 since the launch of Chad2PT, you need to have an AI focus to understand where the equity markets are going.
01:26I think the one key question to ask is, what is OpenAI going to do?
01:30That's the big elephant in the room. It's amassed now 800 million weekly average users, 15 minutes a day.
01:37And that has captured one of the scarcest resources that we have, which is consumer attention.
01:42And I think the next back question is, how is OpenAI going to monetize that user base?
01:47And is this necessarily a zero-sum game where some of it is going to shift away from other digital consumer platforms?
01:54Yeah, I mean, earlier this month we had Paul Tudor Jones on. You're familiar with his work.
02:00And he was saying, like, this October, to some extent, looks like October of 1999.
02:05Now, that isn't necessarily bad because we had a 92% run-up in the NASDAQ into March of 2000.
02:11But it was nearing the end of that run-up in valuations.
02:15And it took us then years to recover from that burst.
02:19Are you at all worried about a bubble burst?
02:21We are certainly in an environment of elevated valuations. That's clear.
02:26Whether it's a bubble or a high-valuation environment, for me, is largely definitional.
02:31I think one thing is clear. The investment implications have to change.
02:36Risk management has to be tighter.
02:38The higher the valuations, the more scrutiny on the fundamentals.
02:42Those markets don't correct just because of high valuations.
02:45They correct because fundamentals disappoint.
02:47And that's what we saw in 1999-2000. In the early 2000s, Lucent had a profit warning.
02:53And the dot-com companies were running out of cash. The Fed was hiking.
02:58Those were the catalyst for the implosion of the bubble, in my mind.
03:02And that's, for us, the biggest investment takeaway.
03:03That's huge. At a time when everybody is piling in to say, this is the time where you need to have more due diligence and extra scrutiny.
03:13Does that mean that that's not happening then, Ulrika?
03:15And then there are some concerns that issues are getting papered over?
03:19Well, I think, look, right now we are not too concerned about the AI trade.
03:23The ROI that we have seen on enterprise use cases is huge.
03:28If you just think about the coding use case, $100 of GitHub co-pilot license, you're getting 56% more productive.
03:36We're not talking about 20%, 30% improvements in efficiency.
03:40That's more than 100x ROI.
03:42So we have enough data points to know that we can monetize AI.
03:46The question is one of speed and how does it line up with the CapEx.
03:51And, yeah, that is the ratio that we are continuing to monitor.
03:54But we also know this technology is very compute intense, whether it's training, whether it's inferencing, with now the reasoning models, the agentic AI framework.
04:03We estimate that we could need more than five times the current install base of compute by 2030.
04:09So, again, monetization potential is there.
04:12It's compute intense.
04:13For us right now, we still see that ratio in line with our expectations.
04:18So, and you've done very well with these stocks, right?
04:21If I look at the industry group breakdown of the S&P year to date, IT is up 30%.
04:27Communication services, which is the hyperscaler, is up 25%.
04:31Utilities and industrials have to play on the energy they'll need and the Caterpillar equipment to build these data centers.
04:36What else should investors be looking at outside of the AI play or what else that we're maybe not thinking about as AI?
04:45Yeah, one sector that we like as a diversifier to the AI trade, but yet it's also a sector that we think is long-term a huge beneficiary of AI, is the healthcare sector.
04:55Now we have some of the headwinds that we're facing that sector out of the way with the Pfizer deal showing us the path forward on most favorite nation pricing and also tariffs.
05:03And then we have upside from AI, both in the drug discovery side, but also all the way through clinical trials.
05:11We think AI could long-term reduce the cost curve of bringing a drug to market.
05:15And that's now together also with the de-stocking finally coming to an end from COVID with valuations relatively benign.
05:22We think that's a really attractive sector.
05:24And we have seen yesterday it's a sector that actually is up when the AI trade is down.
05:28So a very nice sector to hold as part of an AI portfolio.
05:32We notably haven't talked about the other event risk that we got through this week, namely the Fed and conversations between Presidents Trump and Xi.
05:41With those out of the way, Nomura's Charlie McElligot makes the argument that macro risk is on the back burner now, that now we're concerned about the right tail of AI and the melt-up in equities.
05:50Is he right? Can you put aside macro as a driver of this equity market for the time being?
05:56Yeah, we actually think that macro could be a driver for more upside.
06:00If you think about the last 12 months, we've come to all-time new highs despite a challenging macro.
06:05We think the macro looking also into 2026 is actually improving.
06:10We have powerful fiscal and monetary tailwinds.
06:13We have, of course, defense and industrial spend on the agenda.
06:17We also have the refunds from the OBBBA bill.
06:20And then we have now the monetary stimulus through the rate cuts and, as you said, still strong structural trends.
06:28So that's a very unique setup, very rare in the history of financial markets.
06:32So that makes us very bullish on U.S. equities in particular for this and next year.
Be the first to comment