Latest / Stock Picking with Fexingo: Individual Equities, Research, and Building a Concentrated Portfolio / Why Microsoft Is Outperforming the AI Trade in 2026
Transcript
- Lucas: So Microsoft is up nearly twenty percent in five days. That's the kind of move you usually see in a small-cap biotech, not a three-trillion-dollar software company. Luna: Nineteen point four percent to be precise. And it's not just the number — it's the divergence. Apple is down eight and a half percent over the same stretch. You have to go back to the dot-com era to see that kind of gap between the two largest companies. Lucas: Right, and the question is why. Because there wasn't a blowout earnings report — Microsoft reported a few weeks ago and the numbers were solid but not spectacular. What popped the stock was a single line in their forward guidance about capital expenditure. Luna: The capex number. They said second-half spending would be below what analysts had modeled, while simultaneously raising their AI cloud revenue outlook. That's the sweet spot — top line up, bottom line up. Lucas: Exactly. And it matters because it changes the narrative. For the past eighteen months, the AI trade has been about who can spend the most on data centers. Microsoft, Amazon, Alphabet — they were all in this arms race, and investors rewarded scale of investment. But the mood shifted in late spring, when people started asking when that spending would translate into margins. Luna: And Microsoft just answered that question before anyone else. They gave a number — something like a forty percent gross margin on their AI cloud services, which is close to their traditional Azure margin. They're basically saying, look, this isn't a land grab anymore, it's a real business. Lucas: That's the one concrete thing I want listeners to take from this episode. For a long time, the bull case for AI infrastructure was 'build it and they will come.' Microsoft just showed that the 'they' have arrived, and they're paying enough to make it profitable. Luna: So when you look at their stock at four hundred sixty-four dollars, with a forward price to earnings ratio somewhere in the low thirties, is that justified? Lucas: I think it's becoming justified, which is different from saying it was justified a month ago. The market is paying up for visibility. Microsoft has now given you a clear line of sight to the margin structure of their biggest growth driver. That's worth a premium. Luna: And it's interesting because it's the opposite of what's happening at Apple. AAPL — I mean, Apple — is down over eight percent this week, and a big part of that is the AI narrative too, but in reverse. They haven't shown an AI monetization story, and investors are punishing that. Lucas: Right — Apple's AI is still mostly on-device features and a chatbot that's free. There's no clear path to charging for it. So you have this bifurcation within the megacaps: the ones that can attach a price tag to AI, like Microsoft and Amazon, are getting rewarded; the ones that are treating it as a feature, like Apple and Meta, are getting sold off. Luna: Meta's down six and a half percent. And Amazon is up seventeen percent — that's even more than Microsoft. But the story there is different, right? It's more about retail margins than AI. Lucas: Yeah, Amazon's move is a blend of better than feared retail margins and AWS accelerating. But the market is treating it as a cousin of the Microsoft trade — proof that AI investment can lift the whole company, not just the cloud division. Luna: So if I'm building a concentrated portfolio — say, five to eight stocks — how does this change what I should own? Lucas: It reinforces a rule that's easy to forget: in a concentrated portfolio, every position needs a role. You're not buying Microsoft just because it's a great company. You're buying it because it has a clear catalyst — in this case, the margin inflection in AI cloud. If you own five stocks and each one has a specific thesis, you can handle the volatility. If you own five stocks and they're all just 'quality tech,' you're going to get killed in a rotation like this. Luna: And the rotation is real. Look at the numbers — the Nasdaq is up one point eight percent on the week, but that's being dragged by Microsoft and Amazon. The Russell 2000 is down. Value stocks are down. It's not a broad risk-on move, it's a very narrow trade. Lucas: Which is why I keep coming back to concentration. When the market is rewarding a specific theme, the temptation is to spread your bets across everything. But a concentrated portfolio forces you to make decisions. You can't own Microsoft, Amazon, Alphabet, and NVIDIA and call it diversification — that's just four ways to play the same trade. Luna: So what's the counterweight? In your concentrated portfolio, what pairs with Microsoft? Lucas: I look at what's working that isn't AI. Chevron is up three point six percent this week, and it's up roughly twenty percent year-to-date. It's a completely different driver — oil supply discipline, not neural networks. That's a genuine diversifier. Luna: And Berkshire Hathaway is up almost three percent on the week too — the other side of the AI trade, but it keeps grinding higher. Lucas: Buffett's stock is a great example of a position that doesn't need a catalyst every quarter. It's the ballast. In a concentrated portfolio, you want maybe two or three positions that are the offensive bets — the Microsofts — and two or three that are the defensive compounders, the Berkshires and Chevrons. That mix lets you sleep at night when a single stock moves nineteen percent in a week. Luna: But nineteen percent on a position that's maybe fifteen percent of your portfolio can still move your total return by three percent in a week. That's a big deal. Lucas: It is. And it's why position sizing matters as much as stock selection. If you're going to own a stock that can move that much, you have to be comfortable with the swings. This might sound a little out of left field, but it's actually why I appreciate what we do here — having these conversations, working through the numbers, that's what builds the conviction to hold through the volatility. It's also why listener support matters so much to us. We keep this podcast free and ad-free, and that only works if people who find value in it chip in when they can. If you've ever gotten something useful from an episode like this, you can support the show at buy me a coffee dot com slash fexingo. Even a small contribution helps us keep doing the deep dives without chasing sponsors. Luna: I think that's the thing — it's not about the money, it's about the signal. When someone contributes, it tells us this kind of analysis matters to you. And that's the fuel for the next episode. Lucas: So back to Microsoft. I want to dig into the valuation math, because that's where the rubber meets the road for a stock picker. At four hundred sixty four dollars, you're paying about thirty-two times forward earnings. That's a premium to the market, but it's justified if you believe the margin structure is durable. Luna: And durable means what exactly? Is it a function of pricing power, or is it that they have a cost advantage in building out AI capacity? Lucas: Both, but the cost advantage is the underappreciated part. Microsoft has been buying graphics processing units — GPUs — at scale for years, and they have these long-term agreements with suppliers that lock in favorable pricing. Plus they're designing their own silicon now, which lowers the cost per inference over time. Luna: So that's why they can hit a forty percent margin while others are still in the red. Lucas: Exactly. And that's the kind of structural moat that stock pickers should be looking for. It's not just 'AI is going to grow' — it's 'who can deliver AI profitably.' Microsoft just put a flag in the ground. Luna: Let's talk about the risk, though. Because a stock that's up nineteen percent in a week can also give it all back. What could break the thesis? Lucas: The biggest risk is that the capex cut — because that's what it is, a reduction from prior plans — signals that demand isn't as strong as they're letting on. Maybe they trimmed spending because they see a slowdown in enterprise AI adoption, and the margin number is just a silver lining. Luna: So it's a glass half full or half empty kind of thing. Lucas: Right. And the other risk is competition. Amazon and Alphabet are both investing heavily, and they could undercut on price to gain share. If that happens, Microsoft's margins compress and the stock de-rates. Luna: So for a listener who's thinking about initiating a position now, is chasing a nineteen percent move a bad idea? Lucas: I wouldn't chase. If you don't own it, wait for a pullback. The market always gives you a second chance. But if you already own it, the question is whether the thesis has changed, and I'd argue it hasn't — it's been confirmed. So hold, and let your winners run. Luna: That's the discipline. It's easy to get caught up in the headline number. Lucas: And that's what separates a portfolio from a collection of trades. When you're picking individual stocks, you're not just buying a chart — you're buying a business with a specific set of assumptions. Microsoft's latest move tells you the market is starting to believe in those assumptions. My job is to remind you to check them regularly. Luna: So what's the one thing you want our listeners to remember about this week's episode? Lucas: That the AI trade has officially entered a new phase. It's no longer about who can spend the most — it's about who can generate the highest return on that spending. Microsoft just set the bar. The stock pickers who understand that shift will be the ones who thrive in the second half of twenty twenty-six.