Morgan Stanley Chief US Equity Strategist Mike Wilson Talks AI Research on Wall Street

**SPEAKER_1** (0:02)
Bloomberg Audio Studios, Podcasts, Radio News.

**Paul Sweeney** (0:07)
Mike Wilson joins us here, Chief US Equity Strategist and CIO at a little shop called Morgan Stanley. I started, yesterday was my 40th anniversary on Global Wall Street, and throughout my entire career, Morgan Stanley has, in my opinion, the best tech research on Global Wall Street. What do you think about Mary Meeker, Frank Quattrone on the banking side, and all that kind of stuff? Mike, what are you guys saying about AI these days?
That's the theme, arguably, of our career. More than the Internet. What are you guys saying about that?

**Mike Wilson** (0:40)
Yeah, I mean, it's more of the same. So, I mean, a little known fact, I guess, is I started the, I was the original tech sales, tech, you know, desk analyst in the 90s. And so I saw front row seat to the whole boom and bust of that.

**Matt Miller** (0:55)
Henry Blodgett, Frank Quatrone.

**Mike Wilson** (0:57)
Yeah, I mean, so all these names and everything else. So anyways, the point is that this is just another CapEx, this is just another, this is the next tech cycle, okay? And so there are a lot of similarities, there are a lot of differences, but we're in the mix, we're in the middle of it right now. And that's what we've been saying, and that continues. And the reason why I feel confident it's not at the end of the road yet is because the capital markets are still funding this and the companies are being rewarded for it. Meaning, when we look at this carefully, capex to sales ratio is a factor, is being rewarded in the marketplace. Okay, in other words, if you're spending more, the stock goes up. Now eventually that will start to be a headwind and we'll figure out when that is. And when that starts to happen, you gotta worry. The other thing I would be paying attention to is credit spreads, right? And so for some companies, their credit spreads have blown out or CDSs have gone up because at the margin we're seeing, at the margin some companies are not being rewarded for higher capbacks now because they're viewed as a loser, okay? But that, broadly speaking, that's not the case yet. So there's usually early warning signs, there's canaries in the coal mine, I tell you, okay, they're overspending.
I'll say this, this is probably not a house for you, but it's my view. There's 100% chance there's gonna be malinvestment here, okay? Just like every major capback step. But that's not a crazy statement, okay? It's no different in fracking, it's no different in railroads, no different in electricity cycles. So, but we're probably not there yet, we don't know. And the other thing to keep in mind is that the marginal cost to compute, the marginal cost of the product itself has to go to zero for it to work. It will go to zero. There'll be all you could eat planned. So, but we're in that transition period. You know, there's going to be, I think we're now in the transition from the picks and shovels to the adapters. We're seeing like actual adapters of the technology are becoming more efficient, you know, driving revenue, not just cutting costs by hiring fewer people. That transition could last probably several years. And then ultimately, you know, we'll have the other side of it.

**Matt Miller** (2:41)
By the way, Jim Chainos was on with Tom and Scarlett last Friday on Bloomberg Money, their new program every Friday at noon. And of course, he's a short seller and he's talking his book to some extent. But he did say that, you know, during the.com era at the tail end, earnings were up 30 percent and the next year they were down 40 because people were putting in these massive orders due to a huge capex ramp up and then all of a sudden they disappeared. How do you know when we're gonna get there? Because I know you look at forward earnings and earnings revisions that are still, are they still to the upside?

**Mike Wilson** (3:17)
Yeah, so we are seeing a peak rate of change though in the revision breadth in a lot of different areas. And, you know, one of those is semiconductors where we haven't seen it roll over, but it's at such a high level, it can't sustain. So think about this, since ChatGPT was announced in, I guess, November 22 now, and we've had three decent corrections in this bull cycle of capex at three different times. We've seen these stocks go down 30 to 50 percent. So we can have a 30 to 50 percent drawdown and still have the long cycle in the act. And that's what I expect to see.