Warsh Said Inflation Risks Eased. Markets Heard a Rate Cut. They Are Not the Same Thing.

On July 1, on a stage in Sintra, Portugal, Fed Chair Kevin Warsh said two things in the same breath. He said prices are still “too high.” And he said that “expectations of inflation over the first four weeks of this period have come down, inflation risks have come down.”^[1,2] The two-year Treasury yield, the bond that most directly tracks where traders think the policy rate is going, reversed lower while he was still talking, easing back toward 4.14% after touching near 4.20% overnight.^[3] Equities held near a record. The headlines the next morning were about relief.

Here is what actually moved. Before Warsh spoke, CME FedWatch had a roughly 37% chance of a rate hike at the July 29 meeting. After he spoke, the hike odds dropped to about 30.5%.^[3] That is it. That is the whole rally. Nobody repriced a cut, because there was no cut to reprice. Futures still put the odds of at least one hike by September at 77%, and above 90% by October.^[3]

So the market did not front-run a cut. It front-ran the temporary absence of a hike, and a lot of people are reading that as a pivot. If you own long-duration growth stocks or you are sitting on a down-payment waiting for mortgage rates to break, that distinction is the entire ballgame.

What the June dot plot actually said

On June 17, Warsh's first meeting as chair, the FOMC voted 12-0 to hold the federal funds target range (the overnight rate banks charge each other, and the anchor for everything else) at 3.50% to 3.75%.^[4] The statement was rewritten to be dramatically shorter and, crucially, to strip out the language that implied the next move would be down. Warsh described it as “a bit shorter, a bit simpler,” a statement that “just gives you the facts as best we can judge it.”^[5]

Then came the Summary of Economic Projections, the quarterly document that includes the dot plot (an anonymized chart where each official marks their forecast for the policy rate). The median for end-2026 moved to 3.8%, up from 3.4% in March.^[5]Read that carefully. The current midpoint of the range is about 3.625%. A median of 3.8% does not mean “hold.” It means the committee, on median, thinks the next move is up. Of eighteen officials, nine projected a higher rate by year-end, eight projected no change, and exactly one projected lower.^[5]

The risk assessment is even more lopsided. Seventeen of eighteen officials flagged inflation risks as tilted to the upside. Zero saw downside risk. The Fed raised its 2026 PCE inflation forecast to 3.6%, from 2.7% in March.^[5] That is not a committee looking for an excuse to ease. That is a committee that just told you it thinks inflation is going to be worse than it said three months ago.

The transmission chain nobody watches

Most retail investors track the fed funds rate the way you track a weather forecast. Fed cuts, stocks go up, mortgages get cheaper. That is a cartoon. The policy rate is an overnight rate. Almost nothing you own is priced off an overnight rate. What actually happens is a chain, and every link in it can break.

Start with the 10-year Treasury. Its yield is roughly the market's expected average short rate over the next decade, plus a term premium (extra compensation for the risk that the forecast is wrong). Warsh saying inflation risks have eased should compress both pieces. But it did not compress much. The 10-year closed July 2 around 4.49%.^[6] With the policy rate midpoint near 3.625%, the market is telling you it expects short rates to average higher than today over ten years. That is a curve that has already ruled out a cutting cycle.

Now the discount rate. Every valuation model you have ever seen prices a company as the present value of its future cash flows, discounted back at some rate that starts from the risk-free yield. A growth stock is one where most of the cash flows land far in the future. Push the discount rate up a percentage point and near-term cash flows barely flinch, while cash flows in year eight get crushed. That asymmetry is why the AI infrastructure names, the software multiples, and anything trading at 40 times forward earnings live and die on the 10-year, not on the Fed statement.

Why mortgage rates don't track the Fed

This is the part that costs people real money. The Fed does not set your mortgage rate. It never has. A 30-year fixed mortgage is priced off the 10-year Treasury plus a spread, and that spread is a living thing that moves for reasons that have nothing to do with monetary policy.^[7]

Why a spread at all? Three reasons. First, prepayment risk: you can refinance whenever you want, so the lender is short an option and has to get paid for it. Second, credit and servicing costs, including the guarantee fees Fannie and Freddie charge. Third, and most importantly right now, who is buying the mortgage-backed securities. The Fed used to be the biggest buyer. It is not anymore. Banks have been cautious. That leaves money managers, who demand a wider spread.

The numbers: the spread between the 30-year mortgage and the 10-year Treasury sat around 201 basis points in early January 2026, and roughly 189 basis points by late April.^[8,9] The post-2009 average is closer to 170 basis points.^[7] So we are still carrying somewhere around 20 to 30 basis points of excess spread relative to normal. That is a quarter-point of mortgage rate that has nothing to do with the Fed and everything to do with who shows up to buy the paper.

Which is why the actual mortgage rate has been so stubbornly boring. The Freddie Mac survey put the 30-year fixed at 6.43% for the week of July 2, down from 6.49% the week before, against 6.72% a year ago.^[10] A year of drama, a new Fed chair, a war that started and ended, and the 30-year fixed moved about 30 basis points. That is the whole story of rate transmission in one line.

What a lower path would actually do to housing

Let's do the landlord math, because this is where the abstract becomes a check you write. NAR put the median existing-home price at $434,300 in May, with the housing affordability index at 105.6, up from 97.5 a year ago.^[11] An index above 100 means a median-income family has slightly more than enough income to qualify for a median-priced home on standard terms. So affordability has improved, but barely, and mostly because incomes grew, not because rates fell.

Take a $434,300 home, 20% down, so a $347,440 loan. At 6.43%, principal and interest is roughly $2,180 a month. Knock a full point off, down to 5.43%, and it is roughly $1,955. That is about $225 a month, or 10%. Meaningful, not transformative. And that assumes the full point of Fed easing passes through cleanly to the 30-year, which the last three years say it will not.

For a landlord the arithmetic is worse than most people admit. Lower rates cut your debt service, which lifts cash-on-cash return. Good. But lower rates also compress cap rates, which pushes acquisition prices up, which eats the benefit on the way in. If your yield falls as fast as your borrowing cost, you have gained nothing except a bigger balance sheet and more duration risk. The landlords who actually got rich in this cycle were the ones who bought when the debt was expensive and the seller was motivated, not the ones who waited for the rate they wanted.

My call: the market is paying for a cut that is not on the menu

Here is my position, and I'll be specific enough to be wrong. There will be no cut in the second half of 2026. The base case is a long hold with a live hike risk, and the market is not priced for that. Goldman has pushed its cut forecast out to 2027, and the “no cuts in 2026” camp has gotten crowded for a good reason.^[12,13]

The three legs of the argument:

1. The improvement is energy, and energy is a gift, not a trend. Warsh himself tied the improvement to energy prices coming down after the US-Iran memorandum of understanding.^[1] Energy is the most mean-reverting, most geopolitically fragile component in the basket. Building a rate-path thesis on it is building on a fault line. Core services, the part the Fed actually cares about, is the slow-moving thing, and it did not get fixed by a peace deal.
2. Warsh deliberately removed the tool that lets the market front-run him.Deleting forward guidance is not a communication tweak. It is a statement of philosophy. Warsh has spent a decade arguing the Fed talks too much and that guidance makes the committee a hostage to its own past sentences. A chair who refuses to pre-commit is a chair who will not deliver the cut you priced just because you priced it. Warsh saying the committee will “have a good family fight” about July is not a man laying groundwork.^[2]
3. Seventeen of eighteen officials see upside inflation risk.^[5] A cut requires that committee to change its mind collectively, quickly, and against a PCE forecast it just raised by nearly a full point. That does not happen in a quarter absent something breaking.

Which brings me to the honest caveat. The thing that produces cuts in H2 2026 is not good inflation news. It is bad labor news. A cutting cycle that starts because unemployment jumps is not the cycle equity holders think they are buying. If you are long growth stocks on a “rates are coming down” thesis, understand that the world where you are right about rates is probably a world where you are wrong about earnings. That is the trade nobody wants to say out loud.

What would change my mind

• Two consecutive core CPI prints at or below 0.2% month-over-month driven by services, not energy. Core services is the tell. Everything else is noise wearing a suit.
• The unemployment rate rising four tenths of a point or more from its recent low. That is the classic trigger, and it would mean the cuts arrive for the ugly reason.
• A shift in the dot plot at the September SEP where the median for end-2026 falls back below the current range midpoint. Nine officials would have to move. Watch that number, not the headline.

Absent those, do the boring thing. Underwrite the deal at today's rate, not the rate you want. Do not extend duration in your equity book on the assumption that the discount rate is about to fall. And when you see a headline saying the Fed chair sounded dovish, go find out whether what moved was the odds of a cut or just the odds of a hike. Nine times out of ten in this cycle, it has been the second one.

Warsh told you inflation is still too high. He told you he will not tip his hand. He told you the committee is going to fight about it. The market heard a rate cut. That gap between what was said and what was priced is where the money gets made, and lost.