Steve Cohen's move looks more like selective rotation than a tech exit
Cohen's filing looks more like a selective rotation than a retreat from growth. Point72 reported $77.95 billion in 13F portfolio value at the end of March, and NvidiaNVDA-- remained the largest reported holding even after the manager trimmed Amazon by 6% and Nvidia by 24%. At the same time, he increased his Boston ScientificBSX-- position by 50% from the start of the year.
What the 13F can and cannot tell you
A 13F is a quarterly report of holdings. It provides a snapshot, not a real-time view of trading, so it is useful for spotting interest rather than proving what Cohen owns today.
Boston Scientific stands out because the business is still producing proof
The more important question is not whether Cohen owns healthcare, but where operating results can still outrun bad sentiment. Boston Scientific fits that setup better than a portfolio headline alone. After a stronger first quarter first-quarter financial results, the stock has room to be reassessed if execution continues.
Why the business case still looks plausible
A bigger Boston Scientific position in a broad Point72 portfolio looks less like a rescue bet and more like a disciplined buy into temporary weakness. The company's revenue mix remains heavily weighted toward cardiac care and medical-surgical products, so the case depends on procedural demand and device adoption holding up rather than on a single breakthrough announcement.
AI is still central to Point72, even after the sales
Point72's top holdings still include Nvidia, AmazonAMZN--, and Arista, with the top five names accounting for just 7.68% of the portfolio. Separately, Cohen added roughly $2.8 billion in fresh positions across five AI giants, including $870 million in Taiwan Semiconductor, $510 million in Nvidia, $492 million in Amazon, $487 million in Equinix, and $439 million in Broadcom. Even with an 18% weighting in AI infrastructure, this still looks like a portfolio that remains exposed to the AI trade rather than one abandoning it.
Why the overlooked-name case can still work
The opportunity is not that AI is wrong. It is that attention can become overcrowded. When investors concentrate on a few visible winners, slower-compounding businesses can trade as if they are broken when they are merely overlooked. Boston Scientific is a cleaner example of that mismatch, with first-quarter net sales of $5.203 billion and adjusted EPS of $0.80.
What would confirm or weaken the setup
The practical move is to watch for confirmation, not headlines.
Quick Backtesting Tool
Signals that support the case
- Growth is holding, not just sounding defensive. After Boston Scientific beat its prior 10.5% to 12.0% revenue growth guide and matched the top end of the $0.78 to $0.80 adjusted EPS range, the next check is whether that pace sustains.
- Core segments are doing their job. Because Boston Scientific is heavily weighted toward cardiac care and medical-surgical products, investors should look for evidence that device traction is translating into steady procedural demand.
- Margins stay intact. A clean quarter can be emotional relief; durable margins would suggest better underlying operating leverage.
Signals that weaken the case
- Procedure recovery slows.
- Device momentum fades after the post-result bounce.
- Compliance costs or other structural expenses start to cap operating leverage.
- Debt-related drag becomes more than a background issue.
That last point matters because the Boston Scientific signal still rests on a 13F filing, and a 13F is only a quarterly report of holdings. It can flag interest. It cannot prove what Cohen owns today. So the better approach is to watch the next operating updates closely.
