Subscribe
Share
Share
Embed
TCW is a leading global asset management firm with over 50 years of investment experience and a broad range of products across fixed income, equities, emerging markets, and alternative investments. In each episode of TCW Investment Perspectives, professionals from the firm share their insights on global trends and events impacting markets and the investment landscape.
Over the summer, the Magnificent Seven continued to hit
records, largely driven by the vast potential of generative
AI, dragging the rest of the market along for the ride
However, tech giants like NVIDIA, Apple, and Alphabet experienced
their fair share of volatility, resulting in many headlines about
the AI trade unwinding, and investor uncertainty about the sector
Welcome to the TCW Investment Perspectives Podcast, where
we give you an inside look at what's driving the markets
I'm Michael Reilly, Chief Investment Officer for Equities, and
today I'm joined by Bo Fifer and Evan Feagans, co-portfolio
managers for the TCW Artificial Intelligence ETF
In today's episode, we'll dive into where the AI sector
is heading and where we're finding opportunities
So Bo and Evan, AI stocks had been riding high most of this
year, and then we witnessed a substantial sell-off this summer
Can you elaborate on what drove that sell-off, and where do we go from here?
Yeah, I think it helps to start with what was driving the
strong performance over 2023 and the first half of 2024
Essentially, we saw what was basically an arms race amongst
the largest technology companies to build out their AI
infrastructure, and this involves investing a lot in data centers
If you look at the hyperscalers, collectively they were spending hundreds
of billions of dollars on AI infrastructure, and that was boosting the
fundamentals of companies like NVIDIA and other suppliers to data centers
I think investors, after a strong run, which was justified by fundamentals,
I think they started to get a bit skittish and started to wonder
what their return on investment was for all of the spending
I don't think that there was a consumer-facing killer app, so to speak,
and I think the bar just got a lot higher and expectations got to
a point that it was just going to be tough to beat at some point
Fast forward to the second quarter earnings, and I
actually think the fundamentals were pretty good
There were a handful of blow-ups, but for the most part, I
think the average company, the fundamentals held up well
Estimates didn't go down on average, and it was actually
pretty positive from an AI fundamental perspective
But I think these elevated expectations, and then
there were also some macro fears that crept in
I think the risk of recession started to creep up in investors'
minds, and generally the margin for error was just incredibly low
We saw a significant trade-off first at the
beginning of August, and then again in September
Were the concerns about monetization of their AI investment
or a deterioration in fundamentals, what was actually
the nature of the concern on the part of investors?
I think the concern was that even though it was driven by fundamentals thus far
by all of this CapEx from the hyperscale cloud companies, that they wouldn't be
able to continue to invest indefinitely if they weren't getting an ROI on this
So there was a sense, okay, we're almost two years since the launch of
ChatGPT, we should be seeing more AI products in our day-to-day life
I think there was a perception that things like Microsoft Copilot would be used
by office workers across the world, and that's just not really the case yet
We think it's a little bit early to be measuring it
Again, it's so early
I like to use the comparison to the smartphone cycle, and at this
point of the smartphone cycle, we didn't even have an app store
So I think it's just too early to really be looking
for ROIs, and we need to focus on the green shoots and
development and the promise that we have going forward
Sure
So can you give us a sense of how does this sell off
compared to what we've seen in the past technology cycle?
I know we've had actually a bit of a recovery from
the sell-off, but can you put it into perspective?
Yeah
So we've looked at this going back since the inception of the fund, and there
have been, depending on how you measure it exactly, almost a dozen cycles,
up and down cycles for the fund where performance has swung
around pretty significantly, both positively and negatively
And in that historical context, the current drawdown reached a peak of about 20%
in early August, which is only the fourth largest drawdown in the fund's history
And each of those drawdowns has been followed by a significant
recovery that averages around 40%, typically takes
anywhere from four to five months to realize that recovery
But it just makes sense in the context of an industry that moves and fits
and starts sometimes where investor interest definitely ebbs and flows
And related to everything that Evan just talked about in the first half of the
year and all the excitement that gave way to some concern both on the macro side
and whether all we were doing was working through an initial bolus of spending
for these AI systems would lead to a period of dormancy,
getting that all out of the way is very healthy
And the demand for these systems, the demand for the
infrastructure continues to run extremely hot, and we
think it's going to drive us into the next upturn starting
Hopefully it started in early August
We have recovered a little bit off of the trough, but still a ways to go
I want to ask about semiconductors in particular, because if you look at
the sell-off, it seems like the epicenter was in the semiconductor sector
I think from mid-July, semiconductors were down
maybe 20 to 25 percent until through early September
Can you talk about what was driving that volatility
in particular, and what is your view on the sector?
Yeah, there's a couple of things about semiconductors
that make them particularly interesting
It is probably the most cyclical industry within tech and has
certainly been among the most high-flying industries within tech
And so all of that euphoria in the first half of the year that, again, remember
there was an expectation at the beginning of the year that we would see
rate cuts start pretty early, and we would get a lot of them this year
That's good for growth stocks
It's good for tech stocks
And that kept getting pushed out and sentiment
started to weaken commensurate with that
And we reached this point in early July, which I think probably precipitated
the sharp sell-off in the SOX, Semiconductor Index and broader tech, where
the concern became that the Fed had now waited too long
to cut rates, and that would drive us into recession
That would be bad for everybody
The semiconductor companies most exposed
to AI have basically powered through that
And so the stock volatility has been almost purely a function of
investor sentiment because the fundamentals have been quite strong
As Evan pointed out, coming out of 2Q earnings, the median estimate
change for the stocks in AIFD actually went up for both 3Q24 and 2025
Individual stock price reactions were pretty severe in some cases and not
reflective of what we think is the strength of the underlying business
Outside of AI, the biggest end markets like
automotive are going through a correction
PCs seem to be stable, but there's this hope that
there'll be an AI-driven upgrade cycle around PCs and
smartphones, which is the other significant end market
And networking that's been actually quite positive, but that gets more toward AI
So we've had a bunch of cyclical factors that over the past five years have
been out of phase and created a very steady growth for semiconductors that just
kind of lined up negatively over the summer to drive sentiment very, very low
I would just add that we're still extremely constructive
on semiconductors broadly, especially in the near term
As you think about the AI opportunity, the models
themselves still aren't good enough yet, to be honest
There's still a lot of product development
that needs to happen and a lot of improvement
And to do that, companies need more and better chips and networking
and memory and everything that goes into that data center
So we still think that's where a lot of the
near-term investing opportunity is in AI
A question for you, Michael
I think beyond AI, there's increasing uncertainty with the US
economy
Can you maybe elaborate on your macro outlook
and where you see opportunity in equities?
Well, look, I mean, I think stock market investors
got very comfortable with the soft landing narrative
And what's happened is we've seen softness in the labor market
We've seen weakness in the manufacturing sector
And so it's called into question the soft landing narrative
And so, yes, the Fed is going to cut interest rates
But I think the question on the minds of investors is, is it just a
matter of slower growth or does this slower growth give way to recession?
And so far, it looks like slower growth because most
of the macro data isn't consistent with recession
And I think it's really important to keep in mind that in a slower
growth context, there's still plenty of investment opportunities,
but it probably does make sense to focus on secular growth trends
In other words, stocks that don't require a strong GDP
growth backdrop to be compelling investment opportunities
So maybe that's a good segue to talk about the opportunity right now for
AI investing, which is all about tapping into big secular growth trend
Maybe I'll put the question out there
It's clear many investors are currently concerned about
entry points into the equity markets given higher valuations
So how do you think about that valuations and then the
broader opportunity for AI investing in thematic equities?
Yeah, I'll touch on valuations, but first I want to pick
up what you talked about with secular growth opportunities
And I just want to emphasize that AI theme is largely macro agnostic
These hyperscale cloud companies, these have tens of billions of dollars of cash
They have the ability to invest through the cycle and it would take
an extremely severe recession to cause them to change these plans
Being at the forefront of this new wave of
technologies is mission critical to them
I mean, and they can't afford to fall behind
It's not just offensive growth
It's also protecting their existing businesses
I think of Google's search business is potentially under
siege from things like chat GPT and they need to invest
to protect their own businesses, not just grow further
So maybe moving beyond that to devaluation, right now the S&P
tech sectors is currently about 25 times next year's earnings
That's a 25% premium to the overall S&P 500
That is a bit elevated
I will say that tech is typically trades a few turns
above the overall market because of higher growth
But if you really believe that AI is one of those once
in a decade or once in a generation waves in technology,
I really don't think that's too much of a premium
I compare this to the dot-com bubble when
the IT sector was pushing triple digits
So we're nowhere near that on valuations
I'm not worried about a bubble
I actually think the valuations are quite
reasonable given the opportunity that AI offers
I just want to emphasize a point there
This growth that we're seeing in AI is not being viewed
as optional by the companies that are deploying it
This is in many cases deemed to be existential
And as Evan points out, the big guys that are investing in it right now
clearly have the power to invest whether we dip into a recession or not
And they will
The only question is whether and if they throttle that back a little bit
And this is one of the reasons we're so excited about AI and why
we think this recent sell-off has provided such a great entry
point, because so much of this growth is still in front of us
These models are getting exponentially larger
The compute that's required, the power that's required, the cooling that's
required to run these models is growing a thousandfold every couple of years
And the hyperscale cloud service providers are collectively
spending hundreds of billions of dollars in the
biggest arms race that we maybe have ever seen in tech
And that's just the first stage of growth
Beyond that, we've got enterprise
And I think this will look a little bit different where
it won't just be these large so-called frontier models
These are the big chat GPTs with trillions of parameters that run on hundreds of
thousands of GPUs or trained on hundreds of thousands or tens of thousands of
GPUs, but smaller models that are custom trained for a specific task, which
can be evaluating an x-ray or a CT scan or generating a monthly statement
They don't have to do everything
They can be smaller models trained on company-specific and
customer-specific data that are therefore a lot less expensive to implement
And I think that's what's going to drive enterprise into this game in a big way
And we're talking about tens of thousands of large companies across
the country, let alone the world, that will bring their spending to
bear as these applications get developed, proven out
business cases and revenue models get put into place
There is so much growth in front of us
And as one CEO said at a conference this week, anybody with a five
to 10 year investment horizon should not be worried about whether
we have built out all of the training capacity that we need
And now it's just all about inferencing
All of this is still in front of us
All good thoughts
Well, that's it for this edition of TCW Investment Perspectives
I'm Michael Reilly, and I want to thank our guests, Bo Fifer and Evan Feagans
Join us for more expert analysis and in-depth
conversations on the trends shaping financial markets
For more information on our strategies, visit tcwcom
Thank you for joining us today on TCW Investment Insights
For more insights from TCW, please visit tcwcom/insights
This material is for general information purposes only and does not
constitute an offer to sell or a solicitation of an offer to buy any security
TCW, its officers, directors, employees or clients may have positions
in securities or investments mentioned in this publication,
which positions may change at any time without notice
While the information and statistical data contained herein are based on
sources believed to be reliable, we do not represent that it is accurate and
should not be relied on as such or be the basis for an investment decision
The information contained herein may include preliminary information
and/or "forward-looking statements" Due to numerous factors,
actual events may differ substantially from those presented
TCW assumes no duty to update any forward-looking
statements or opinions in this document
Any opinions expressed herein are current only as of
the time made and are subject to change without notice
Past performance is no guarantee of future results