Higher yields are finally a real problem for crowded growth trades
Rising Treasury yields have stopped being background noise for the market’s most crowded growth names. With the 10-year at 4.631%, the 30-year above 5%, and AI leaders still priced for near-perfect execution, duration risk is starting to bite where positioning is most crowded.
The market has spent months acting as if higher yields were somebody else’s problem. That was easier to defend when AI leaders kept delivering numbers so strong that valuation discipline barely mattered. It looks harder now. When the 10-year Treasury is back at 4.631% and the 30-year has climbed to 5.159%, the discount rate is no longer an abstract macro talking point; it is a direct test of whether investors still want to pay peak multiples for earnings that arrive far into the future.
The key point is not that growth is broken or that AI demand is fake. It is that the market’s favorite growth trades are no longer getting a free pass on duration.
is the clearest example. The company just posted extraordinary fundamentals, with revenue growth of 65.5% and net margin of 63.0%, yet even that was not enough to fully re-ignite the tape after results. That is what priced-for-perfection looks like: the business can still be excellent while the stock becomes more vulnerable because the valuation already assumes near-flawless execution.
That vulnerability matters more in a bond market that is repricing higher-for-longer. We are not talking about a mild move in yields anymore. A 30-year Treasury above 5% changes the conversation for the longest-duration equities because investors can suddenly demand more immediate proof of cash generation instead of paying up for distant optionality. The market can tolerate rich multiples when money is cheap and inflation is cooling. It gets much less forgiving when inflation-sensitive data keep rate pressure alive and every future dollar is worth a little less.
The comparative setup across the crowded growth complex makes that risk hard to ignore:
That spread tells the story better than any macro slogan. MSFT is not cheap, but it is a different kind of growth asset than AMD or TSLA. Microsoft is still expensive relative to the broad market, yet a roughly 25x earnings multiple and 39.3% net margin look far easier to defend in a higher-yield world than AMD at nearly 147x earnings or Tesla at 230.70x while revenue is shrinking and net margin sits at 4.0%. If rates stay elevated, the market is likely to keep separating durable cash machines from stories that still rely on a lot of future promise.
Yes, bulls will say the AI leaders are earning their premiums. They are not wrong to point to Nvidia’s 65.5% revenue growth, Microsoft’s still-powerful cloud engine, or AMD’s 164.4% EPS growth. But that argument misses the market regime shift. When yields were easier to dismiss, strong growth could paper over almost any valuation concern. When long bonds are making fresh highs and even Nvidia’s blockbuster quarter fails to produce unqualified enthusiasm, the burden of proof rises. Great companies can keep growing and still see their multiples compress.
That is also why the speculative end of the growth trade looks especially exposed. TSLA is the cleanest warning sign. The company is still valued like a transformational AI-and-robotics platform, but the current operating profile looks much less forgiving: negative revenue growth, negative EPS growth, and thin profitability. In a lower-rate world, investors can indulge that optionality. In a higher-rate world, they start asking what is being funded today, what the return on that spending will be, and how long they are willing to wait. The same logic applies to baskets like ARKK, which remain shorthand for long-duration equity exposure and have already struggled to regain leadership.
Meanwhile, the contrast with a name like JPM is useful precisely because it shows this is not a blanket bearish call on stocks. Banks do not carry the same duration profile as crowded AI trades, and a higher-yield backdrop can be less hostile to them than to premium-multiple tech. JPM at 14.43x earnings and 2.83x sales is not being asked to clear the same perfection bar as Nvidia, AMD, or Tesla. That does not make financials exciting. It does make the relative valuation gap across the market look harder to justify if the bond market keeps tightening financial conditions on its own.
The market signal this week was subtle but important. AI is no longer automatically overpowering rates. That is the shift. For the last year, investors could treat every yield spike as temporary noise because AI enthusiasm overwhelmed everything else. Now the tape is telling us that even the best narrative in the market has to compete with a real cost of capital again. Once that happens, crowded growth stops being a one-way trade and starts behaving like what it always was: a duration bet with exceptional fundamentals attached.
We do not think this has to end in a broad crash. The more plausible outcome is a slower, more frustrating unwind in which the biggest growth winners keep posting solid numbers but struggle to expand their multiples because the bond market will not let them. That is enough to change leadership, especially when some of the most crowded names are still valued as if the macro backdrop barely matters.
What would change our mind? A clear cooling in inflation-sensitive data and a retreat in long yields would help restore the old playbook. Short of that, the market has to prove it can keep paying peak prices for long-duration growth while the 10-year sits in the mid-4% range and the 30-year stays above 5%. Right now, that looks like a much tougher ask than the AI bulls want to admit.
Our take, not advice. This is opinion commentary — informational only, not personalized investment recommendations. Markets carry risk. Do your own research and consider your own situation before any trade.
▌The Daily Briefing · Free
A new stock idea, every evening.
One stock worth watching each weekday, plus the analysis behind it. Free, in your inbox.
