The market is pricing about a 95% chance of a 25 basis point hike today, and one view holds the Fed has been pushed into a corner with no choice but to hike. I think the Fed could still hold. The market has forced the corner, but if anyone has the nerve to say "we are not doing what the market demands," it would be Kevin Warsh, and he has cover to pause and wait.
The core inflation argument
Strip out food and energy and inflation runs at 2.4% year-over-year, one of the lowest readings since 2021. The trouble is an energy spike. Oil is out of control, and that is where the inflation pressure comes from. A Fed rate hike will not produce a new barrel of oil, will not open a new Strait of Hormuz, and will not settle the Middle Eastern crisis, which is the source of the price pressure. So the Fed has a leg to stand on to wait and see. If oil settles, inflation should work its way out of the system.
The counterpoint is real: food and gas are the two costs Americans feel most, and the two things I want to strip out are exactly those. Warsh said in his first press conference that Americans are feeling the strain of higher costs and that it has gone on too long. Can he now call that a short-term energy blip and strip it out? Yes, he can - the Fed has done stranger things in the last decade. Should he? In our view, we are probably on the verge of a hike within hours.
The political trap
This has become far more political than it needs to be. Warsh has two choices: side with the administration, which clearly does not want a hike and lists reasons at every press conference why rates should not be cut, or side with the bond market, which demands the Fed pay attention to inflation and to the national debt, which is out of the Fed's control. That is the tightrope. Warsh is the next trapeze artist to walk it.
Credibility and street credit
If Warsh does what the market wants, he earns credibility - street credit. The big question is whether the Fed has credibility in the bond market that it will attack inflation. The 30-year and 10-year yields have both moved higher. A 10-year over 5% is historically bad for stocks. So is a rate hike, which usually puts a wet blanket on the market.
What is unusual now: a month ago I would have said a hike ends the bull run. Today I think the market has largely priced the hike in, because a 10-year note over 5% by itself threatens the bull run - it pulls money out of stocks into a guaranteed yield instead of reaching for 6% or 7% in equities at higher risk. The good news: a hike may not kill the bull run.
The AI narrative and existential risk
The other threat to the bull market is a shift in the AI story - from heavy enthusiasm toward talk of a "SaaS apocalypse," and now, from comments this weekend by Dario Amodei and other CEOs, even a human-race risk: a stated 10% chance of annihilation, a 90% chance of survival. That is worth watching.
Does it put a wet blanket on the market? Yes. Does it put the fire out? No. Earnings keep growing, and earnings growth is what keeps taking the market higher. AI adoption is broadening out. It may cause money to rotate out of tech and the AI trade into the broader market, but that rotation already happens seasonally - the market typically cools in September, and we positioned clients for that. We are not abandoning the tech trade. Jensen Huang of Nvidia says some data centers earn their money back after the first year. As long as AI spending continues, AI revenues climb, and earnings grow, the bull market stays. We are bullish on the AI trade into 2028.
The picks
- Apple (AAPL): owns the consumer of AI, has a new CEO and a lot of tailwind, a long-term hold.
- Marvell (MRVL): sold off on earnings and still not valued right by the market; exploding data center revenue, a potential front runner, since AI demands data.
- Palantir (PLTR): a little on sale now with a lot of upside.
- A packaged options ETF built around QQQ - a put-call spread, described as the September version - called "the AI ride with a seat belt." It offers as much as 18% upside on the Qs and protects the first 15% of losses on a September-to-September basis. A good place to put money if you are not sure the AI thesis holds but want to take part if it does.
