Memory stocks are still “early innings” — here’s the ETF to invest in
Transcript:Caroline Woods:Memory has become one of the hottest trades in AI. But after a recent pullback, investors are asking, is this a buying opportunity or a sign the trade has gone too far? Joining me now is Thomas DiFazio, ETF Strategist at Roundhill Investments. Thomas, great to have you on. ...
Overview
Transcript:
Caroline Woods:
Memory has become one of the hottest trades in AI. But after a recent pullback, investors are asking, is this a buying opportunity or a sign the trade has gone too far? Joining me now is Thomas DiFazio, ETF Strategist at Roundhill Investments. Thomas, great to have you on.
Thomas DiFazio:
Thanks, Caroline. It's great to be here. Appreciate you having me.
Caroline Woods:
And well Thomas your Roundhill DRAM ETF has been getting a lot of attention lately. I've been seeing it mentioned in the comment section from our viewers. Before we get into the trade and the fund, let's start bigger picture. Why has memory suddenly become such an important part of the AI story?
Thomas DiFazio:
Yeah, well, what we've seen and I the AI revolution has evolved significantly just in the past couple of years. And that is a byproduct of the technology, itself. But AI is no longer just a chip story. It's about the whole machine. And as investors have realized, it's not just about the semiconductors, it is about memory, which has really become a constraint and bottleneck in the AI revolution.
Thomas DiFazio:
The best AI chips are going to underperform if it doesn't have enough memory to, service all the requests that it's gotten. And we constantly are seeing, as we have the next, the next cloud model come out, the next chatty, cheap ChatGPT model come out. Is it? The queries are only getting more energy intensive. And when we add on thinking and verifying all of this information, you need more memory to satisfy those requests and to do them properly.
Thomas DiFazio:
And that's really where, that's really where the enthusiasm for the trade has originated and where this insatiable demand throughout the industry has come about.
Caroline Woods:
The only issue is we've seen this really strong run in the memory trade already. We are seeing some profit taking now. Micron, SanDisk, Western digital names like that are all off the highs. So do you think that the memory trade has gotten ahead of itself? Where do you think we're still in the early innings here. And for those investors who aren't in, do they still have a chance?
Thomas DiFazio:
I think memory overall, and this is the round tables from U2, is that we are in a long term structural change for the memory cycle. This isn't just a couple months story now since Dram has launched in early April. We have seen that memory stocks have been very much embraced by the overall market. And right now we, it's hard to argue that when you see such eye popping price action, it is it is easy to think that, wow, maybe, maybe the trade has gone ahead of itself.
Thomas DiFazio:
But we do cite a couple other important factors. The first one being that of all the largest market caps in the world, semiconductors are among the cheapest. Samsung and SK Hynix traded around seven times earnings, and micron trades at around ten times forward earnings. That is extraordinarily cheap relative to the broader technology sector as well as, relative to, broader equity markets.
Thomas DiFazio:
That's an important figure to keep in mind. It does showcase that even with the extraordinary price action, you have relatively in-line valuations. But the other part of it that's important to keep in mind is that and we saw this with microns earnings last week as well. There are still fundamental underpinnings that are supporting this price action. This isn't just multiple expansion or even from suppressed multiple levels.
Thomas DiFazio:
So now we're really encouraged by what we've seen so far. And micron, Samsung, SK Hynix and the rest of the memory trade have delivered. Going forward, I think we are entering a new era of base discovery for the memory stocks in our conversations with clients. We we do get those questions about, well, is this the memory of old?
Thomas DiFazio:
And for for for those that are curious about the history, memory used to be very much a boom and bust industry where you are capped by demand of device upgrade cycles. We look at windows 95 in the mid 1990s, and then you could look at the iPhone seven, in the mid 20 tens, where eventually that insatiable demand for memory was short lived.
Thomas DiFazio:
And then you saw memory prices greater as supply started to catch up. We're now in a new era where we don't believe that there is a device cap on memory demand. As I continues to grow and be integrated into the world economy. And frankly, we think that AI is industry agnostic. It's applicable everywhere. There's only going to be a need for more memory, even if the devices get more efficient.
Thomas DiFazio:
And that's where we tend to not get too concerned about. Okay, perhaps we're seeing a downturn from last week's incredible price action or last week's difficult price action. Because when you look in the longer run, this is about finding a brand new baseline for memory stocks. And that is going to take some time. And that does come with volatility.
Thomas DiFazio:
And we're cognizant of that. So for the shorter term tactical traders Caroline I would say that yeah you're going to have to play. You're going to have to play a little bit more thoughtfully for now. But for those longer term investors that are looking to, to, to allocate to a part portion of their portfolio, this is where we would say that it gets a little bit more difficult to start sweating the techs and to start thinking about the longer term.
Caroline Woods:
So this is an ETF spotlight. The the fund is Deram I was taking a look at some of the top holdings, some international plays here Samsung SK Hynix, micron SanDisk, Western Digital, Seagate. If someone already owns a broad semiconductor ETF, though, like the SMA, are they already getting memory exposure or what are they getting with this?
Thomas DiFazio:
It's a great question. That was actually when our product development team was hard at work at this. That was where we saw the gap. If you were holding in an estimate, just as a great example here, you're still getting broad semiconductor exposure. And depending on the day, you're getting mid-single digit exposure to micron. Which if that's what you're looking for, that's that's perfect.
Thomas DiFazio:
That being said, the SMA is not what you're looking for. If you're looking to have an outsized or, excuse me, a targeted that on memory stocks. You do get you get micron. But you could just on the single stock in that regard. But you are also missing two of the most important names in the memory space, which are Samsung and SK Hynix.
Thomas DiFazio:
SK Hynix actually leads, the overall the global market for high bandwidth memory share controls around 55 to 60% of global market share, depending on the quarter. Now, I've we've also seen, that in the lead up to the launch of Dram that we, South Korea at the South Korea ETF, we saw a significant amount of inflows.
Thomas DiFazio:
And that was because investors were looking to get exposure to SK Hynix and Samsung. But you were also getting anywhere from 45 to 50% exposure to a bunch of South Korean names that you may not necessarily look for that exposure to. So when Dram came about, we wanted to give investors a tool to get targeted exposure to just those memory stocks and and that product set.
Thomas DiFazio:
And that construction has really resonated with our investors.
Caroline Woods:
But if I can buy micron or even SK Hynix, I think next month right on the Nasdaq through address, why pay an ETF fee at all? Why not just buy the individual names?
Thomas DiFazio:
Yeah, it's a great question. It still is about it still is in access play. The ADR for Sky NEX is, I would say, welcome news for the market. Up until this point, investors have not had that ability. It'll be interesting to see, how the how liquidity is and how it trades. We suspect that it could it could trade well and be well embraced by the market.
Thomas DiFazio:
But ultimately, memory is not just a US store. And that's an important distinction. While the Skynet ADR is really important, we still are missing one for Samsung. And we do also hold names, within Taiwan, and, and other global equity markets. And that is where investors do lack, certain access points. And so that's where we believe Dram continues to have a big value add for investors, in their portfolios.
Caroline Woods:
Talk to us a bit more about what role Dram actually plays in a retail investors portfolio. Should this be a satellite position around broader semiconductor exposure or something different?
Thomas DiFazio:
Yeah. No, it's it's a great question. We've seen arguments that it can replace core technology positions. That's not where we sit. Frankly. I think that is more for those investors that are really further out on the risk curve because as you've seen, memory just trades, in a very volatile fashion right now. The way we see it, we think, Dram could be a wonderful satellite position where you're looking to have targeted that's in certain pockets of either the eye bottleneck or just a certain express bet within technology that you're looking for.
Thomas DiFazio:
But it doesn't need to replace your core technology position. Those are for, that it would still lack some of that. It was still lack, I would say, the diversification that a broader technology fund would have. This is one part of, of the eye trade that some investors want, only that. But we think that's why it still should stay as a targeted position.
Caroline Woods:
I do want to ask you about risk, too, because we're already seeing higher memory costs ripple through the rest of the tech sector. We certainly saw that play out with Apple saying that they have to raise prices because of higher memory prices. So does that eventually become a headwind for AI spending, or is it still bullish for memory companies?
Details
Thomas DiFazio:
So it it could the I think the the what is really interesting right now what we've seen as a is through really two different charts. First one is when you look at free cash flow, for the Magnificent Seven, the hyperscalers, you're seeing that that has really taken a hit because of their strong CapEx spending.
Thomas DiFazio:
Meanwhile, free cash flow for the for the memory, for the major memory players has benefited extraordinarily from it. So you're really seeing a tale of two trades. The hyperscalers are really being used as a funding trade right now for the for the memory stocks. And I think, I think that's a really I think that's a really compelling, I think that's a really compelling argument that spending appears to be in place for, the intermediate term.
Thomas DiFazio:
And we think that the memory stocks are going to continue to benefit from, and we're continuing to see they are investing as well. They're doing a lot of long term, contractual agreements with their customers. Now, the the really interesting part is that means there's an even smaller pool of supply that needs to be dedicated and, for, for for investors to get exposure to and for companies to get that memory allocation to.
Thomas DiFazio:
So these memory names continue to grow, their production because they have to and they need to be able to satisfy the demand which they're seeing in place, till at least 2030. And it could go beyond that.
Caroline Woods:
Six months from now. What do you think investors will wish they had understood about the memory trade today?
Thomas DiFazio:
So I think that investors are going to wish, that they understood more about where the memory sector had been so that they can start to, so they can, have a more informed understanding of where it could possibly be going. I'm a student of history, and my, my optimism from the memory trade came from looking back and where my memory had previously been and understanding what were those drivers that changed the cycles previously and then are those drivers still in place today?
Thomas DiFazio:
There's an argument to be made that memory cost rising is a big impact. Is a big impact, right. But we're I think if you continue to use, use history as your guide, you can you can then start to have a more, a thoughtful approach on where the trade could be going.
Caroline Woods:
Okay. Before we let you go, we like to do a rapid fire round of this or that. Quick questions, quick answers. We don't always do it for ETF spotlight, but I wanted to do it with you. A memory trade edition. Are you ready?
Thomas DiFazio:
Let's do it.
Caroline Woods:
Memory trade. Early middle or late innings.
Thomas DiFazio:
Early.
Caroline Woods:
Recent pullback. Buying opportunity or warning. Sign.
Thomas DiFazio:
Buying opportunity.
Caroline Woods:
Memory. Structural or cyclical? Structural AI hardware or AI software from here. If hardware biggest risk supply catches up for AI, demand slows.
Thomas DiFazio:
Supply catches up.
Caroline Woods:
More upside over the next 12 months. Magnificent seven or memory place?
Thomas DiFazio:
That's a good one.
Thomas DiFazio:
I'm gonna go make seven.
Caroline Woods:
One AI theme you'll be watching most closely over the next year. That's not memory.
Thomas DiFazio:
Good. I'm very interested in the photonics and optics stocks.
Caroline Woods:
Give us a few names. We should be watching.
Thomas DiFazio:
Momentum is is, is a big one for for me within that, within that space. It's been it it's been a winner in this tape. But I think it's another part of the I bottleneck that could be could be a really compelling opportunity for investors.
Caroline Woods:
Okay. And because I know Roundhill has a whole bunch of, ETFs not just around memory, but, throughout all of tech, if you could only own one AI theme for the next five years, which would it be? Yeah. Yeah.
Thomas DiFazio:
I would, I would stay with memory.
Caroline Woods:
Okay. We'll leave it there. Thomas DiFazio ETF strategist, Roundhill Investments. Thank you so much for playing along and for your insights. We really appreciate it.
Thomas DiFazio:
Thanks, Caroline. It was a pleasure to be here.
Caroline Woods:
If you enjoyed this ETF spotlight, check out our full conversation with Joe Grogan. He lays out how to diversify beyond U.S. stocks, in particular European defense stocks.
Source
Originally published at www.thestreet.com.
