3 Stocks That Prove the AI Trade Isn't Over, It Moved

Key Points

• As AI infrastructure spending peaks near $725 billion in 2026, investors are shifting focus toward companies profitably applying AI in real businesses.
• Lemonade, AppLovin, and Hinge Health each use AI to run profitable operations in insurance, advertising, and healthcare, despite recent stock volatility.
• TradeSmith's Keith Kaplan recommends sizing positions by risk level, favoring large-cap names first while treating smaller AI application stocks as higher-risk opportunities.

Everyone has already made money on the AI buildout. The next money gets made by whoever's actually using it.

For the past two years, the easiest trades in the market were the ones tied directly to AI infrastructure: the chips, the data centers, the power contracts. That money has largely been made. The stocks that ran hardest on that story have already been rewarded, and in some cases already punished, as investors start asking a harder question: who's actually turning all this spending into profit? That's the tension running through this year's AI story, and it's why the next round of winners may not look anything like the last one.

The Buildout Is Over, the Payoff Begins

Alphabet Inc. NASDAQ: GOOGL, Amazon.com, Inc. NASDAQ: AMZN, Microsoft Corporation NASDAQ: MSFT, and Meta Platforms, Inc. NASDAQ: META are on track to spend a combined $725 billion on AI infrastructure in 2026, up 77% from last year's $410 billion. That's a staggering number, and it's only a cost until someone builds a real business on top of it.

History keeps repeating this pattern. The railroads laid track across America in the 1800s, and it was Montgomery Ward and Sears, not the railroad men, who got rich shipping catalog goods to every town the tracks reached. Amazon did the same thing with the internet: it built a bookstore on top of someone else's infrastructure and became the biggest retailer in the country.

The application layer is where that's happening now, according to Keith Kaplan, CEO of TradeSmith. These are companies running real businesses on AI, in insurance, advertising, and healthcare, and most of them don't even screen as AI stocks. That's exactly why they're still cheap relative to the infrastructure names.

Lemonade Turns a Sell-Off Into an Entry Point

Lemonade, Inc. NYSE: LMND uses AI to underwrite insurance in seconds instead of weeks. Its second-quarter revenue rose 79% to $294.4 million, and in-force premium climbed 32.5% to $1.43 billion, the 11th straight quarter of accelerating growth. Gross profit has grown roughly tenfold over the past several years while headcount has barely moved, a sign of just how much of the workload AI has absorbed. The stock still fell roughly 20% on the report.

Why the drop? Wall Street wanted a guidance raise, not a repeat of solid execution. Adjusted EBITDA losses narrowed to $19 million from $41 million a year earlier, and the company still expects to reach positive adjusted EBITDA by Q4 2026. A newly announced CFO transition added some uncertainty, but nothing about the underwriting engine itself changed. When a cash machine gets discounted on expectations rather than execution, that's typically an entry point, not a red flag.

AppLovin's Margins Make Chipmakers Look Inefficient

AppLovin runs a different kind of machine. AppLovin Corporation NASDAQ: APP built Axon AI, an AI engine that decides which ad hits your phone, billions of times a day, with no factory and no inventory behind it. Its most recent quarterly revenue rose 59% year over year to $1.84 billion, with net income around $1.2 billion, and the company carries a market cap near $131 billion. Institutional ownership sits at roughly 41%, which leaves real room to grow if larger funds decide those margins are too good to keep ignoring.

The catch is an ongoing SEC investigation into AppLovin and how Axon AI collects the data that makes it work. That's a real risk worth sizing modestly for. But the numbers are hard to ignore: this is one of the most profitable software businesses in public markets, and its next earnings report lands Aug. 5 after market close, which could clarify how much the probe is actually weighing on growth.

Hinge Health Rebuilds Physical Therapy Around AI

Hinge Health, Inc. NYSE: HNGE is the smallest and newest name here, a nearly $6 billion company that put a physical therapist's judgment into a phone camera. It watches you exercise, corrects your form, and targets one of the largest categories of employer health spending in the country.

Revenue grew 53% year over year in the latest quarter to $213 million, and Hinge Health again raised its full-year guidance, now calling for 2026 revenue of $856 million to $860 million. The Aug. 4 report gave investors a stronger read on the company’s post-IPO momentum, with free cash flow of $100 million and a new agreement to acquire Cylinder Health adding another layer to the growth story.

Because Hinge Health only went public in 2025, many institutional investors may still be building their view of the stock, but the latest quarter gives them more to work with than a simple early-stage growth narrative.

How to Size the Risk

Not every name in this group carries the same risk, and Kaplan's approach is to size positions accordingly. The safest starting point is the large-cap names where the business model is already proven and profitable. Mid-cap names carry bigger swings and sometimes hinge on a single industry trend or regulatory decision. The smallest, newest names carry the most speculative upside and the most volatility.

Lemonade, AppLovin, and Hinge Health are three names from a broader list Kaplan and his team at TradeSmith have grouped by this same risk framework, sorting names by whether the AI is making decisions, buying ads, or rebuilding an entire industry from scratch. The practical takeaway is to treat a list like this as a menu, not a checklist: pick the handful of names that make sense for your own risk tolerance rather than trying to own all of them equally, and keep position sizes small enough to ride out the swings without panic-selling into weakness.

Where the Risk and the Upside Actually Sit

The upside across all three names is the same: they didn't build the infrastructure, they bought into it, and that means faster payback with a fraction of the capital commitment the hyperscalers are making. The risk is just as consistent: every one of these stocks can swing hard on a single earnings print, a regulatory headline, or a guidance number that doesn't match inflated expectations.

None of that changes the underlying setup. The infrastructure money has already been moved. Stay focused on who's actually running a business on top of it, because that's what moves these stocks next.

Continue following MarketBeat

Add MarketBeat as your preferred source on Google to see our latest stories in your feed.

Add As Preferred Source