Joseph Carlson — 6 Stock Update

Salesforce jumped about 22% after earnings. Carlson's point is that the good news is mostly real but smaller than the headlines: part of the revenue growth came from buying Informatica, and most of the earnings "beat" came from the rising value of Salesforce's stake in Anthropic, not from selling more software.

What impressed him was the buyback. Salesforce borrowed about $25 billion and used it to buy back roughly 15% of its own shares in a year. Borrowing costs it about $1 billion a year in interest, which lowers total profit — but because the profit is now split among far fewer shares, each remaining share earns more (about $13.16 instead of $12.50 in his example). And while the interest bill stays fixed, the business keeps growing, so the trade gets better over time. The downside is a weaker balance sheet and less room for acquisitions. He calls it smart financial engineering — though he still prefers Meta, which he sold Salesforce to buy.

Netflix today grows by making or licensing its own shows. Carlson's bigger idea is that it becomes a "hub": you could sign up for Peacock, Paramount+ or Fox inside the Netflix app and watch their shows there, with Netflix keeping a slice of each subscription. Smaller streamers would happily pay, because those are customers they might never have reached — the same way a hotel is glad to fill a room that would otherwise sit empty.

That would let Netflix earn money from other companies' content without spending billions on it, similar to how Visa earns more whenever people spend more. The risk is ruining Netflix's clean interface with a clutter of paid add-ons. On engagement worries, he argues a live event watched by millions is worth more than its small share of viewing hours suggests. At about 23 times earnings he calls it a great company at a not-expensive price.

Reddit is one of the few places online where real people discuss things in depth, which matters more as the internet fills with bots. It makes money by letting advertisers target very specific communities, and by licensing its conversations to AI companies. Revenue is growing about 66% a year, the number of advertisers about 70%, and it has $3 billion of cash.

The worry is that AI tools now answer questions directly, so fewer people click through to Reddit from Google — and Reddit stopped reporting a user metric that would show whether that is happening. After a 35% fall this year, Carlson thinks the price (around 25 times next year's earnings) is reasonable for a business he expects to be around a long time.

AppLovin runs an advertising system that matches advertisers (first mobile games, now online shops) with ad space inside apps, using its AI model to predict which ad will pay off. The stock fell about 60% from its peak even though revenue is still growing about 47% a year and profits even faster, which is why it has become popular on Qualtrim.

Carlson was looking at it for the first time. Optimists say a recent slowdown was just the timing of a model upgrade; skeptics say growth depends on unpredictable breakthroughs and that the push beyond games is unproven. He likes the setup — a beaten-down, fast-growing, cash-rich company at a low teens earnings multiple — but hasn't done the work to take a position.

SpaceX makes most of its money from Starlink satellite internet; its rockets mainly exist to launch and replace those satellites cheaply, and it is adding AI data-center hosting and other long-shot businesses. Starlink is profitable and growing fast, but the company is pouring most of its spending into AI computers and the Starship rocket.

At about $1.84 trillion with only a couple of public quarters, Carlson says there is no reliable way to value it — you are buying a belief in Elon Musk's ability to make money from many bets. He has no problem with a small position like that, as long as it is balanced by companies whose growth can actually be modelled.