The Best AI Stocks to Buy Now — And One to Avoid

Joining us now to preview a big week of tech earnings is Angelo Zino, equity analyst senior vice president at CFRA. Angelo, welcome back. So good to have you. Great. Thanks for having me Caroline. So of course Nvidia is the highlight this week. But we'll also hear from names like Salesforce Marvell. Quite a few tech names. Before we get into those though, AI has already made investors a lot of money. Angelo, is there still easy money to be made here? So, Caroline, I wouldn't say there's easy money to be made here, but I would say there's still money to be made. And I think it all depends kind of at how long this kind of cycle really goes. And I do think kind of the narrative has now shifted away from, hey, listen, all the focus on higher CapEx to more of the focus now turning to the AI monetization strategy and kind of making sure that we're seeing, you know, better kind of data points out there that point to AI monetization really ramping up. And we actually did see some really good indications here during the early parts of Q2 earnings season, whether it had been from the hyperscalers, you know, in terms of posting some great cloud numbers and pointing to acceleration, even some of the Nio cloud players out, you know, when you think about it, and they get a lot of the kind of the residual demand out there away from the hyperscalers, but they're, you know, they've seen some really good pricing, uplift on their end. They've seen some really good, bookings momentum as well. So, we're seeing really some really good AI monetization stories start to play out. We need to obviously see that continue, and get stronger as we go into 2027. Dig into that a little bit more, though. What's the best evidence that all of this spending is actually paying off? Yeah, I mean, I do think right now it is some of the cloud numbers. I mean, you get the cloud numbers. What happens is, you know, you know, clearly you get the the greater capacity growth that's coming out there from the hyperscalers. You want to make sure that, that capacity is being absorbed and taking taken out by their customers. Right now, a lot of it clearly coming from the entropic open eyes of the world. But we need to see, you know, more of these enterprise companies really kind of absorb a lot of this stuff. You're you're now seeing better stories out there, I think, from, some of the software companies as well. It doesn't seem like it's all doom and gloom from the software companies. They're creating some great AI tools and we are seeing momentum there as well. I think this, again, is going to be an important week from the, you know, from the some of those SaaS players, like a Salesforce, for instance. So, you know, as long as we again continue to get those, you know, those positive stories out there and indications of the enterprise, purchases, I think, again, that, that helps. And of course, you want to see that token usage continue to grow, which, you know, all indications are we are seeing. Okay. So Nvidia reports Wednesday you have a strong buy on Nvidia. Still if I don't own it already would I buy it before earnings. I think you need to own some of it. So if you don't own it ahead of earnings I think you you should buy some of it. And you know, I would say when you think about Nvidia here over the last couple of weeks, you know, you've clearly seen a lot of the momentum names really. I don't want to necessarily say fall apart, but they've, you know, fallen clearly, you know, a good distance away from their their 52 week highs. But when you look at Nvidia on a relative basis has actually started to perform better than the broader, semiconductor industry average. And the big reason for that is clearly they haven't been kind of, you know, thrown into this momentum trade in the first half of the year. As we go into this print, all eyes are going to be on to on, Vera Rubin. Vera Rubin, is that new? You know, that new kind of, server that they are ramping up here that we're going to see, revenue in the October quarter. So if we stay stronger than expected demand for Vera Rubin, I think that could bode well here for for the stock. Looking ahead. Angelo, on that note, Nvidia is reportedly raising prices more than 15% on some of its next gen, chips servers. You know, do you look at that as pricing power or a reason for customers to look elsewhere? No, I mean, I actually think, one, it's pricing power. And, you know, clearly component costs are going up across, you know, the AI ecosystem. And a lot of those higher component costs are, you know, naturally, Nvidia is not going to absorb, but they're going to look to pass on to their customers, especially the hyperscalers. The other thing I would say is, as you go into the some of these next gen, servers, the whole focus for Nvidia is to continue to improve the economics, right? It's kind of trying to lower the total cost of ownership out there for their, for their customers. So, from that perspective, it continues to make an extreme amount of sense for these customers to continue to migrate towards these next servers. So I think it's the combination of the fact that, hey, listen, you've got to pass on these higher costs. And I think as long as they continue to successfully pass those higher costs on, they'll continue to do it. And second, it is the improving economics and what that means to the hyperscalers. And that, again, kind of allows them to have that pricing power. We know, though, with a lot of these companies that the bar is pretty high, right? We've gotten used to these beat and raised quarters. So the stock might not react favorably favorably, even if it's good news. What would actually disappoint you from Nvidia and what would Nvidia have to do wrong? I guess for you to change your Strong Buy rating. Yeah, I mean, for us, I mean, the valuation is extremely enticing going into this print, which is why we continue to have the strong buy as long as the market continues to act rationally. From that perspective, I think it makes sense to continue to own the stock. As you go into this print, you know clearly what could go wrong of what, you know, could make this, you know, make us change our minds. Clearly, if the the Vera Rubin brand is not as, as positive as we anticipate, you know, Street's looking at about 104 billion here for the October quarter. You should suspect at least a 1 to 2 billion, beat from that perspective. But as you kind of, you know, you want to see that trajectory continue to play itself out. Well. And if it's not, you want to know why, right? Is it some something on the supply side? We do think the supply chain is much better positioned with Vera Rubin than they were with Blackwell. So we don't think it's going to be a supply issue here early on. So, if there is a demand problem, I would say, or if Jensen kind of provides guidance, into 2027, that may not be as positive as investors out there would hope. That's something that, you know, would make you kind of revisit your, your thought process on the stock. But at this point in time, there shouldn't be any reason for that to be the case. If Nvidia does stumble, does that help Marvell, which also reports this week, or a Broadcom which reports next week? Or do you think it hurts the the whole AI trade? It probably hurts the entire AI trade. I mean, the entire AI ecosystem is going to hinge on Nvidia. And I think when you look at the other names that are reporting here over the next couple of days and weeks, there are a lot of off peak names, whether it be Marvell, Broadcom or even micron next month, which, you know, all eyes are now on. As far as the, the the memory trade is concerned, it really all stems towards, Nvidia because they continue to have the lion's share of the market here. And a lot of those companies are essentially all those companies to some degree, do rely on in on, in video, maybe Broadcom could be a little bit of an outlier. But I would say nonetheless, I mean, all of those companies need to rely on Nvidia as well. So you want to make sure as far as customer spending is concerned and their visibility that that looks good. For the kind of the rest of the ecosystem to remain healthy. Let's talk about Marvell really quickly because it's down about 4% today. But if you take a look, it's a more impressive chart year to date and for the past one year than Nvidia. It's a 167% year to date, more than 200% over the past one year. How are you looking at valuations on a marvell? Would you buy the weakness that we're seeing today? Yeah, I mean listen, we we like the Marvell fundamental story a lot. And we do think it has pulled back enough from its peak, where a lot of that, you know, maybe a lot of that fluff out there has kind of been taken out of the stock. I think this is one of those names you definitely have to keep an eye on in a name that we do like a lot. Going into the print here, I mean, all eyes are going to be on a recent Google announcement that was made last week. We think it has the potential to drive significant, upside to expectations. So, we would own this stock. And, I think what dictates the, the move on Marvell has almost nothing to do with the actual numbers itself and everything to do with how they, articulate what that upside potential, and what reality for that contract is going to look like here over the next couple of years, because it really does have the potential to drive significant upside to some of our numbers. Okay, let's move on to Sales Force. It's actually higher today and it's well off the lows but still down about 20% year to date. What does Salesforce need to prove this week? Yeah I mean Salesforce is a tough one. I will say you know what they've got going for them obviously right now is the fact that the the software trade up here is like, it's getting some momentum here over the last couple of weeks. And a lot of that has to do with the fact that these multiples have gotten so depressed, and the fact that it doesn't look like, at least at this point in time, you know, we've seen any sort of massive destruction in terms of the fundamentals on the software side of things. So for Salesforce, I would say what you want to say is you want to see the core business, their subscription oriented business continue to remain healthy. We think that will be the case. There could be some weakness. Or they could disappoint slightly on that side of things. And then more importantly is what they have to say in terms of the momentum for Agent Force, what they're seeing in terms of that inflection here in the second half of this year, because that is a big part of this story. So if they see a really good inflection, if they start seeing a lot of kind of pilot use cases, you know, move to, full production and we see some really good growth numbers out of their eye oriented business. That might be enough to kind of convince some investors, to continue to buy up the shares and get a little bit more optimistic going into 2027. On Salesforce. I mentioned it's well off the lows but still down on the year. Do you look at it as an eye bargain then, or is this a value trap for us? We do look at it as a value trade. At this point in time. I wouldn't necessarily call it a value trap because we do think the core business remains intact. We do think there are signs, in ways that they will continue to improve the margin trajectory of the, of their overall business. And then at this point in time, almost kind of hey, listen, if that AI business really starts to swing in favor, and, you know, it does what it needs to do. You know, we do think you can kind of maybe see, you know, a rerating to the upsides at least slightly on the multiples. And that could drive, you know, easily another 20, 25% upside here in the shares. But I would say here it when you kind of think about just the broader software trade, what you do have to be mindful about is, hey, listen, there's going to be you've got the anthropic IPO coming out probably in a couple of weeks. You've got OpenAI probably in the first half of next year. When they actually do IPO, they're going to be able to generate a ton of proceeds. They're going to be able to reinvest in that business, you know, get a lot more, capacity built there. And it's going to allow them to really announce a number of new, offerings on that side of things. So I wouldn't necessarily say you're out of the woods at this point in time. As far as, the uncertainty in the software side is concerned. What's a software stock that you think will hold up and do? Well, regardless of what happens with AI and software? I mean, I don't know if I'm allowed to give this answer, but I would say Microsoft. I mean, that might be kind of a, you know, a, you know, a cop out. And but, you know, we continue to be extremely bullish on the Microsoft Play. And obviously they've got the cloud story and what have you. But it is the fact that we are still fairly bullish on the copilot momentum and what they're doing in terms of the application software side of things. So that's something we do still like. And I would say maybe if you're looking for pure play application software, ServiceNow is another name we like as well. Okay. So you kind of said how to think about the software trade overall. Still not out of the woods. How should investors, especially retail investors, be thinking about the overall tech trade at this point in the cycle? What do they need to consider? You know what? How should they be thinking in terms of market leadership? Yeah, I mean, I think that's a great question. I'd say right now, what I would say is you want to be a lot more balanced in nature across the, the, the tech, the overall tech space. And, you know, we continue to have an overweight weighting on the IT sector. We have since 2022. And what I would say is, you know, clearly the last three, three and a half years, it has all been on the the kind of the semi side of things, the AI infrastructure build. The comps obviously get a lot more difficult from here and nobody knows kind of how long in the long in the two. This cycle really goes. Right. But all kind of indications are what you're going to see on this moving forward is likely multiples continue to continue to impress. We know the the the metals on the AI infrastructure build probably hold up and remain very strong through 2027, just based on our forecast for, the hyperscale or CapEx spend. What it looks like in 28 and beyond is what the market is trying to figure out. So as long as you know what we tell investors, as long as the market stays rational, and that's and that means discount those risks, compress the multiples a little bit. But see that good growth earnings growth trajectory I think that's why you will remain invested overall. But you stay balanced across software semis as well as big tech because big tech should be able to provide some nice, you know, downside support if kind of if we see some volatility there. Does balance mean 33.3% in tech, 33.3% and big tech and the rest in chips are how how does that balance look? Yeah. I mean, what I would probably say is, I mean, I would probably be a little bit more, bullish right now on more of the still on a semi as well as the, let's call it the, the, the hyperscalers side of things. So maybe 40, 40 and then 20 on software. And if we continue to see some better monetization stories on the software side of things, and maybe the cycle continues to get longer in the tooth, there may be some uncertainty on the semi side. Then you increase your your exposure on software, but maybe be a 40, 40, 20 mix. If I already own plenty of big tech through the S&P 500. So you know, I kind of have that covered. And I'm thinking about where my next dollar goes. Where should that go first. Yeah I mean I think that's a great question. I think, you know, we continue to be I'd say more more focused on growth at a reasonable price. So, I would say continue to, to invest in areas of growth, within tech, maybe kind of a growth oriented, you know, tech ETF over value. But nonetheless, I mean, you know, you still want that, that value exposure but still focus on on growth within the the IT sector is what I would say. Okay. So before we even get to rapid fire, a quick answer here. Best growth stock that comes at a reasonable price is Microsoft. In our view, I think I think that's kind of the way to look at this. I mean, you kind of look at the, and, and this is if you're a long term investor, right? We look three, five, seven years out. And there is a lot of uncertainty in this market still across the, the I infrastructure trade. But Microsoft is a name that has been a very diligent in terms of how they spend their money, being very rational with it. And that's a company that we just trust a little bit more when it comes to the management team relative to some of the others out there. Okay. All right. I think this is a great time to pivot to our rapid fire, to officially pivot to our rapid fire game of this or that you've played before. Quick questions, quick answers. No hedging if you can help it. Are you ready, Angela? Let's go. I boom early middle or late innings. I'd say middle innings. But also be mindful of the fact that we are likely going to get some choppiness along the way. So there will be fear out there that maybe we are approaching an end, but our view is probably middle innings. By any dip, yes or no. So we. Yes. Eye spending accelerating or peaking. Peaking with semiconductor sales likely having peaked here mid-year. But nonetheless we do expect growth numbers on a year over year basis to remain attractive over the next 2 to 3 years. So next wave of I winners chips or software. We do think, chips are still a good place to be, but, the way you would play chips might be, different looking ahead. So it may not be memory like it has been over the last 12 months. It may be the shift towards more, networking oriented plays, potentially like a marvell, as we kind of look at that shift away from copper to laser, for instance. Chips or data centers. Or chips. Software I winner or I victim. It depends. But, we continue to be positive on the fact that you're going to see a number of eye winners. It's all about finding those winners. One eye winner. That's a software name. That's not Microsoft. I would say, ServiceNow. One AI software name that will likely be a victim. Adobe better value Salesforce or Nvidia. Still Nvidia hyperscalers or semis? Semis. Nvidia ahead of earnings. Buy or wait. It depends. We would we would buy again if you're if you're not invested at this moment in time we would buy and we do expect overall in video to perform well. It just may not see the reaction. Some investors want to see. Broadcom or Marvell. Marvell AI stocks stick with the leaders or bet on the laggards. Probably still, stick with the leaders because most of the leaders continue to see, at least on the semi side, continue to have some of the best content stories that are out there. And historically, the way to actually benefit or win across the semi ecosystem is to find the best content winners out there. Okay, it's time for fill in the blank. Finish the sentence. The AI leader I have the most conviction in is. Microsoft. The AI laggard most likely to become a leader is. And I don't know if this is a good answer, but I would say meta. They're not necessarily a laggard, but the stock is absolutely lagged here of the last 12 months. We do think that it's an opportunity. The best AI stock to buy today is. AMD. That is a company where we think the earnings power remains significantly underappreciated and we think has probably the longest tail among kind of the names within our AI semi ecosystem. The AI stock I'd be most nervous owning is. Probably. Probably Adobe if you consider it. Or I would say kind of some of those, you know, content names out there that, you know, could potentially be easily replaced by. What's a content name. Give us an example. Yeah. I mean, I think Adobe is a really good example. In terms of, you know, just if you're if you're a company that just relies on, you know, just, you know, if you don't own the actual data out there, it's maybe the best way to put it. The companies that do own the data, specifically on the software side, will be the biggest winners there. The first sign the AI boom is in trouble is. When NIO clouds start, start cutting CapEx or can't, you know, can I would say get the actual funding that they need to, you know, meet some demand expectations out there. The market is underestimating blank. We think meta, may might be the best way to put it. We think they are not properly, pricing in, or factoring in the, the potential AI monetization story. They're across a number of different levers that they have. Nvidia's biggest threat is. Probably hyperscalers, with their own custom, custom silicon chips. I would say, a close second would, would be a. And we do think AMD is doing a lot in terms of some of their software initiatives. If that kind of ramps up, and gets more formidable quicker than people anticipate, you know, that that could potentially be problematic for Nvidia. One of the biggest year to date winners, SanDisk, micron, Dell and Seagate. The one I'd still buy here is micron. We expect them by the end of this year to announce a, significant CapEx plan. And, we do think there's, there's fairly, there's a really good floor here for the stock, just given our view of the fact that, hey, you look at the free cash flow potential as well as the balance sheet. We we think that's it. Just a name that you continue to stick with. Of the biggest losers Oracle, IBM, Adobe, Zscaler, Salesforce I bet on. Salesforce. When all is said and done, the company that wins the I arms race is. The company that can continue to grow. But also, be prudent in terms of their spending plans. If I had to guess which company that is, it is. I would say, I would say Microsoft. Obviously. I would also say, I would also say Amazon as well. There are two names that we do like a lot. I would also say a name that we haven't brought in is Apple. As we eventually kind of shift towards, edge devices, we do expect them to be a big winner. So the one I bet I'd make for the next five years is. I would say, and followed by, but by Qualcomm, actually, it's a name that we, that really hasn't, had the move yet, but it's a name again, as we shift towards edge devices. And you look at the multiple there, as well as the fact that they will be a big beneficiary of six G towards the end of the decade. There are enough levers, whether be tied to AI or even not tied to AI, that should make them a huge winner, especially given the valuation there. All right. We managed to squeeze in so many names there. Angelo Zino, thank you so much for playing along and for for getting us ready for tech earnings this week. That's Angelo Zino Equity Analyst, Senior Vice President, CFRA. Always a pleasure. Thanks so much, Angelo. Great, thanks for having me. If you enjoyed this street talk, check out our full interview with Ahmed Riesgo. He has a 7800 price target on the S&P 500, and explains where he'd take profits and what he's rotating into.