Warsh was sent to cut rates. The dot plot went up.

Justice Clarence Thomas swore Kevin Warsh in as chair of the Federal Reserve on May 22, 2026, four weeks after the Senate confirmed him 54 to 45, the most divided confirmation vote in the Fed's history. He was the president's pick because the market read him as the rate cutter. Twenty-six days later, running his first meeting of the Federal Open Market Committee, the Committee held the federal funds rate at 3-1/2 to 3-3/4 percent for a fourth consecutive time, 12 to 0, and then raised its own median forecast for where that rate lands at year end. Not lowered. Raised, from 3.4 percent in the March projections to 3.8 percent in June.

That is not the outcome anyone underwriting Warsh's nomination in January had priced.

Why did the rate-cut chair's committee just raise its own forecast

Because the data stopped cooperating, and the Committee chose to believe the data over the mandate story that got Warsh the job. The FOMC statement is direct about why: inflation "remains elevated relative to the Committee's 2 percent goal, in part reflecting supply shocks that have driven price increases in certain sectors, including energy," while activity is "expanding at a solid pace despite elevated uncertainty that owes, in part, to the conflict in the Middle East." Consumer prices were up 4.2 percent year over year in May, more than double the target the Fed has promised since 2012. The Summary of Economic Projections released the same day puts core PCE inflation at 3.3 percent for the year. Real GDP growth is projected at 2.2 percent, unemployment at 4.3 percent. Nothing in that table says recession. Everything in it says the committee has more work to do than it expected in March.

Warsh reportedly declined to submit his own dot at his first meeting, though he encouraged the rest of the Committee to submit theirs.

That is a small, telling detail. A new chair who wanted to stamp his own rate path on the institution on day one would have put a number on the page. Warsh let the Committee speak first, and the Committee spoke hawkish: traders are now pricing the next move as a hike, not a cut, with October as the earliest live date. The longer-run dot still sits at 3.1 percent, so nobody on the Committee thinks 3.8 percent is the resting rate. They think it is the rate that gets inflation back under control first.

The curve already told you the cost of capital changed

Here is the part development teams should sit with longer than the politics of who chairs the Fed. The overnight rate and the long rate have decoupled from where they sat for most of the tightening cycle. SOFR is running around 3.66 to 3.68 percent in early July. The 10-year Treasury closed at 4.49 percent on July 2. That is an 83-basis-point gap between the financing index and the long bond, and it has widened since the 10-year troughed near 4.24 percent in mid-2025. A construction loan prices off SOFR. A take-out or a stabilized asset's value prices off something closer to the 10-year plus a risk premium. When that gap widens, the bridge between "we can build this" and "we can sell this" gets more expensive to cross, independent of anything a county planning board does.

Run the arithmetic once, plainly. A construction loan priced at a reasonable current spread, SOFR plus 300 basis points, floats at roughly 6.66 percent all-in today. A stabilized multifamily asset traded in the first quarter of 2026 at an average cap rate of 5.75 percent, per RealPage's read of the RCA data. Compare those two numbers and the debt costs more than the asset yields on a current basis. That is not a crisis. It has been the normal condition of this cycle since 2022. But it means every month a project sits in entitlement rather than under a lease is a month spent paying more for money than the finished asset would throw off, and that gap is a function of the rate path this post is about, not of anything site-specific.

I want to be specific about what changed and what did not, because "higher for longer" has become a phrase people use to avoid saying anything falsifiable.

What changed on June 17: the ceiling moved up. What did not change: the committee's own longer-run view, still 3.1 percent, and the market's structural expectation that rates eventually fall from here. What also did not change, and this is the point most rate coverage skips, is where the inflation the Fed is fighting is actually coming from. Shelter costs rose 3.4 percent year over year in May, down hard from an 8.2 percent peak in March 2023, and they had actually bottomed near 3.0 percent in November 2025, the lowest print since August 2021, before ticking back up, a move partly distorted by a rental-sample data gap during the October 2025 government shutdown. Shelter is running below the headline 4.2 percent CPI print, not above it. The category the Fed is worried about is energy and the tariff-adjacent supply shocks named directly in the statement. Housing is not the thing scaring this Committee. It is decelerating while the rest of the basket is not.

That distinction matters for how a development team should read this Fed. A hawkish committee fighting energy and supply-shock inflation is not the same animal as a hawkish committee fighting a housing-cost spiral, and it does not necessarily resolve the same way. Energy shocks can pass through a system in quarters. Structural housing-cost inflation does not.

The same rate path lands differently by asset type

A flat 3.8 percent policy rate does not hit a garden-style multifamily deal in Ohio the same way it hits a 1.2-gigawatt data-center campus. Multifamily debt sizing runs on debt yield and a stabilized cap rate that moves with the 10-year, so a rate that stays higher than the market wanted mostly shows up as fewer basis points of proceeds and a bigger equity check, the kind of arithmetic we have run before on specific deals. Data-center debt increasingly prices off a signed power contract and a hyperscaler credit, not a market cap rate at all, so the rate path matters less to whether the deal gets funded and more to the discount rate an operator applies to a 15 or 20-year contracted cash flow. The Fed's June forecast makes that discount rate higher for longer across every property type. It just does not make it the same number twice.

What this actually does to a go/no-go

Nothing in a single meeting should flip a project from yes to no. What it should do is move the discount rate you run your sensitivity case at. If your base case still assumes the rate cuts the market expected in March, that case is now stale, and it was stale the moment the dot plot printed 3.8 instead of 3.4. Rebuild the downside case around a fed funds rate that could sit at 3.75 to 4.00 percent into 2027 rather than sliding toward 3 percent on schedule, and rebuild the construction-loan line at whatever spread your relationship lender is actually quoting today, not the spread they quoted in 2024.

The Mortgage Bankers Association counts $875 billion of commercial and multifamily mortgage balances, 17 percent of the $5.0 trillion outstanding, coming due in 2026. Every one of those loans refinances into whatever this Committee decides between now and the maturity date. A sponsor who assumed a friendlier 2026 than the one that just showed up on June 17 is not looking at a modeling error. They are looking at a real cash call.

My own call, stated so it can be checked later: the Committee does not actually hike in October. I think the June dot plot is a hedge, a way to look serious about the 4.2 percent headline print without committing to the political fight a hike under a five-week-old chair would invite, and I think shelter's continued deceleration gives the doves enough room to hold again at the September meeting. If Warsh's Committee does raise the funds rate before year end, I was wrong, and every pro forma built on the market's cut-happy priors from January was wrong twice over.

Either way, stop underwriting to the Fed you wish you had. Underwrite to the one that showed up on June 17, and rerun the file every time the dot plot moves, because it just proved it will.

This analysis is a source-cited research summary drawn from public records, not legal advice. It can contain errors and should be verified independently before any investment decision.