Powell Out, Warsh In: A Fed Inheritance Built on 4% Inflation and 5% Bonds

Compress the past three days. Tuesday, May 13th: Senate confirmation, 54–45, Fetterman the only Democrat crossing the aisle. CNBC and Bloomberg writeups described it the same way - the narrowest margin in modern Fed history, a clean partisan split, and a vote shadowed by Democratic concern that Warsh would bend rapidly to Trump's demand for cuts. Wednesday, May 14th: Stephen Miran resigned from the Board of Governors, effective on the swearing-in of his successor. Miran had only joined the board in September 2025, filling an unexpired term ending January 31st, 2026 - he was always a placeholder, and now he is gone. Thursday, May 15th: Powell's term as Chair formally ended. And on the same day, in opening remarks for a webinar on central banking, Cleveland Fed President Beth Hammack stated, plainly and pointedly, that "an independent and accountable central bank is essential for monetary policy."

Hammack did not name anyone. She did not need to. Hammack's remarks read as a marker laid down by the institution itself - independence first, accountability second, in that order. Her audience understood. Markets, too...the 10-year Treasury yield finished May 15th at 4.59%, the 30-year at 5.12% per the Treasury's daily curve.

Then there is the part that gets less attention. Powell decided to stay on as Governor through the end of his term in 2028. This is rare. Standard practice has been for an outgoing Chair to resign the Board seat at the same time. By holding the seat, Powell does three things simultaneously: denies President Trump a fresh Board appointment for two and a half years, preserves a counterweight to Warsh inside FOMC deliberations, and provides what Richard Fisher, former Dallas Fed president, has called "political cover" for the new Chair. Fisher gave Powell an A-minus this week for handling political pressures, downgrading only because of the "transitory" inflation misstep of 2022. Fisher's historical parallel was striking. He reminded CNBC viewers that President Truman once called Fed Chair William McChesney Martin a "traitor" for refusing to cut rates, and that Lyndon Johnson physically confronted Martin at his Texas ranch to force cuts that Martin reportedly regretted until his death. Fed has been here before. Institution survived. But it never enjoyed it.

Now reset the picture from Warsh's seat. He walks in on Day 1 with one ally departing (Miran resigning), one predecessor staying as counterweight (Powell), and a committee that has just dissented against itself in the most fragmented vote since 1992. That April FOMC vote is the single most important piece of context for what comes next.

Here is the data Warsh inherits. Not forecasts. Just delivered prints.

Read across that table and you see the entire argument the committee is going to have. Headline is hot. Core is sticky but not breakout. Energy and gasoline are doing nearly all the headline damage...exactly the supply-side shock central banks have long argued they should look through. But bleed into services is where this gets dangerous. April services rose 0.5% month-on-month, the largest monthly print since March 2024. Services inflation is notorious for being the slowest to come down and the most predictive of where headline lands in twelve months.

Bank of America's research note that has been making the rounds this week - I keep coming back to it - lays out the path arithmetic with brutal simplicity. BofA's Michael Hartnett ran the monthly-MoM scenarios out to November midterm elections. If monthly CPI averages 0.1% from here, annual CPI lands around 3.0% by November. If the recent 0.4% MoM pace persists, annual CPI hits 5.2%. Four percent matters historically because once headline CPI crosses that line, the S&P 500 has, on average, declined four percent over the following three months and seven percent over six. Risk assets become structurally vulnerable above that line. We are not there yet. We are knocking on the door.

And then there is the part nobody is talking about loudly enough - producer price index. April PPI rose roughly 6% year-over-year, the biggest annual rise since December 2022. PPI services were up 5.5% - broadest pass-through I have seen in this cycle. ING's Padhraic Garvey, whom I generally read with respect, was blunter than usual in his May 15th note: "Headline US producer price inflation has hit 6%, and headline US consumer price inflation is approaching 4%. By the May readings, they'll be higher still. US economy will become a base 4% inflation economy." When a sober rates analyst tells you "base 4% inflation economy" in print, you take notice.

The optimistic counter is that energy is transitory. It usually is. The 2022 Russia–Ukraine shock added roughly four months of rising energy inflation before it rolled over. The same trajectory now would carry the energy CPI peak into late summer 2026. But that trajectory assumed Hormuz reopens. As of this morning, it has not. Iran's new supreme leader Mojtaba Khamenei has publicly committed to keeping the strait as leverage. So the duration assumption everyone is leaning on is itself fragile. 17.9% energy YoY can spread into services and rents far longer than markets are currently pricing.

The piece I find most important, and most under-covered in the financial press, is what's happening at the long end of the curve. Here is the yield ladder as of close of business May 15th:

The curve has bear-steepened. Front-end yields have drifted up modestly as rate-cut bets get pushed further into the future. Back end has moved much more. 30-year cleared five percent at last week's auction - priced just above 5.04%, per the Treasury Department - the first time a fresh 30-year supply has done so since August 2007. Intraday peak hit 5.20%. That is the bond market saying something specific: it does not buy Fed's ability to control the long end and it demands compensation for what it sees as persistent inflation and unsustainable fiscal trajectory.

Two structural pressures are at work behind these levels.

Fiscal arithmetic. U.S. federal debt is roughly $39 trillion, around 120% of GDP per FRED's series. Annual interest cost has roughly tripled in five years - from approximately $500 billion in 2020 to over $1.2 trillion now. Every basis point on a $40 trillion stock matters. Treasury sold $25 billion of 30-year bonds last week at 5.04% and is sitting on roughly $2 trillion in annual fiscal deficits with no plausible near-term path to balance. Issuance volume is structurally elevated, demand is structurally constrained.

Marginal buyers. Japan, the largest foreign holder of U.S. debt, has begun selling Treasuries to defend the yen against 160 level against the dollar. Bank of Japan has been marginal foreign defender of that exchange-rate threshold; their interest aligns with appreciating yen, which means selling dollar assets. Meanwhile, U.S. hyperscalers - Alphabet, Amazon, Meta, Oracle, Microsoft - are now issuing record volumes of corporate bonds to fund AI capex. Oracle's annual free cash flow has turned to roughly negative $24.7 billion. Spread between hyperscaler bonds and Treasuries (roughly 60 basis points wide for the highest-grade names) is now wide enough that some institutional investors are choosing the corporate paper. Treasury is competing for capital with Alphabet, and Alphabet is paying up.

Layer on the energy-driven inflation premium and the long end has every reason to remain elevated. ING's Garvey targets 4.5% on the 10-year by mid-year with overshoot risk to 4.75% in the near term. I think that's the right neighborhood. The asymmetry, though, is brutal: if Hormuz remains closed and CPI prints another 0.4% MoM in May (which I think more likely than not), 10-year could test 4.85% before any cooling. And that has cascading effects across mortgages, corporate debt rollovers, commercial real estate, and equity multiples.

One number to anchor on: U.S. credit card 90+ day delinquency just hit 12.7%. That is within striking distance of the Global Financial Crisis peak. Mortgage delinquencies remain below 1%, genuinely reassuring and is why I do not believe we are facing a 2008 redux. But credit card, auto and student loan distress is at or near 2008 levels, and the cushion is thinning. A 10-year at 4.85% does not help.

What does Warsh actually want to do? Strip the political noise away and Warsh's doctrine is clearer than press has been giving him credit for. Three pillars - all stated in his Senate Banking Committee testimony, all reinforced in his prior writing and CNBC appearances over the past year.

Read those three pillars in sequence and a coherent strategy emerges - a strategy that has a name in economic history. It is called financial repression. It is what Fed and Treasury jointly engineered after World War II to bring federal debt down from 122% of GDP in 1946 to roughly 30% by 1974 - a thirty-year program of holding nominal interest rates below the rate of inflation, devaluing dollar gradually, and growing real GDP faster than nominal debt. Savers paid the price. Hard-asset owners did well.

Warsh has not used the phrase publicly. But the doctrine - lower short rates, smaller balance sheet pressing on longer rates, a structurally lower inflation target through AI productivity - is structurally consistent with a modern repression program. A seasoned macro analyst at Friedrich Global Research who explicitly named it that way in a piece I read this week, may be reading the signal correctly.

There is also the unstated fourth pillar: communicate less, dot-plot dies, fewer speeches. Warsh has called the FOMC's Summary of Economic Projections a problem and described the dot plot as the wrong instrument. He prefers letting outcomes speak. Outgoing Chair Powell himself said earlier this year that he was "never the world's biggest fan of the dot plot" but failed to find committee consensus on a replacement. Warsh will try again. If he succeeds, expect higher front-end rate volatility. Markets will have less to anchor on and a wider term premium on the long end.

One more honest beat: Warsh genuinely admires Paul Volcker. He has said so. Volcker killed inflation of the late 1970s, at the cost of a double-dip recession and unemployment north of 10%, and then set up the longest post-war expansion. Warsh's quiet bet may be that you can run a Volcker-style discipline on the long end (via QT) while running a Greenspan-style accommodation on the short end (via cuts). I find this analytically elegant. I also find it operationally treacherous.

Here is what every breathless headline about "Warsh's Fed" misses: he is one vote of twelve. FOMC operates by majority, and the Chair's vote counts the same as any other. Powerful Chairs - Greenspan, Bernanke at his peak - have been able to move committees through persuasion and institutional weight. But there are cases - 1979 Volcker most famously - where a strong Chair faced majority resistance and had to compromise. Warsh now sits in a committee where his own predecessor will hold a Board seat and almost certainly vote against the dovish elements of his doctrine.

Now layer on the April FOMC vote and the picture sharpens further. On April 29th, committee held federal funds rate at 3.50%–3.75% in an 8–4 split decision - the most dissents on a single decision since October 1992. Dissents went in both directions: Miran wanted a 25 basis-point cut. Three regional presidents - Hammack (Cleveland), Kashkari (Minneapolis), and Logan (Dallas) - voted to hold but opposed retaining the easing bias in the statement. Translation: even before Warsh arrives, the committee is split between those who want to cut and those who want to remove all signaling toward cuts.

Add Warsh's vote when he arrives, subtract Miran's once he formally departs, and you have a committee with roughly seven members biased toward holding rates against the energy shock and four-to-five biased toward cutting - a soft majority against fast cuts. Powell staying on as Governor through 2028 effectively gives the cautious bloc a senior anchor and denies President Trump a fresh appointment slot until then. Warsh doctrine does not get vetoed - it gets diluted. A first cut probably happens in 2026, but the path to 2.75% takes longer than Warsh would like.

One other piece of math worth respecting: Warsh's first scheduled FOMC meeting as Chair is June 16th–17th, just a month from now. He will not have time to renegotiate inflation metrics, write a new framework, or rebuild forward guidance before that meeting. Committee will function on Powell's framework for at least one more cycle. That is, in itself, a reassurance to bond markets - whether they are pricing it that way yet is another question.

I find scenario thinking more useful than point forecasts here because the variance of outcomes is unusually wide. Three distinct paths sit in front of us, each with very different implications for rates, dollar, and equities.

My center case is the middle scenario - stagflation lite. That is not a comfortable place to land, but the data supports it. Energy shock is real and will take months to filter out. Services inflation is just turning. Powell's presence prevents the dovish tail. Iran's stated leverage on Hormuz prevents the hawkish tail's clean resolution. We sit in the middle, frustrated, with a Fed that wants to cut but mostly can't, and a bond market that prices for trouble.

What would change my view? Three things. First, a clean reopening of Hormuz with negotiated terms - I'd shift weight toward dovish. Second, May or June CPI MoM at 0.5% or higher - I'd shift weight toward hawkish tail. Third, two more Governor resignations giving Trump appointment slots - I'd shift weight toward dovish, but with a much wider variance band because Fed independence becomes a live concern. None of these are my modal view this morning. All of them are plausible enough to monitor weekly.

Let me be practical. The asset implications differ meaningfully by scenario, but a few positioning calls hold across at least two of the three roads.

A few positioning conclusions I am willing to put my name on. Short-dated Treasuries remain the best risk-adjusted instrument in the base case. A 1–2 year T-bill at ~3.9% with negative real yield of about 10 basis points is unappetizing, but it preserves optionality. Dividend growth equities outperform in two of three scenarios. Quality companies with 3%+ initial yields and consistent dividend raises north of inflation are the post-WWII repression playbook, and they work if Warsh's doctrine succeeds or if the committee diluted it lands at stagflation lite.

Long-dated bonds are a strict bet on the dovish path. I would not bet the farm on it. Friedrich Global's view that long bonds underperform across the regime change makes sense to me. Regulated utilities - water, electricity in growing-demand regions - remain a structural hold regardless of scenario, because AI capex is driving electricity demand higher and trims neither side of the political ledger has any incentive to change.

One area I am watching but not yet buying: gold. Precious metals tend to outperform during financial repression - this is a multi-decade pattern, well-documented. If Warsh's program is what I think it is, gold has a structural tailwind. But real rates remain positive at the front end, dollar is strong, and consensus repression trade is already crowded. I prefer waiting for either a dollar reversal or a confirmed cut path before adding.

One last thought, written plainly. Post-2008 era of zero rates, predictable Fed cadence, and consensus-driven communication is over. It has been over for some time, but the institutional acknowledgment of its end happens this weekend - in the swearing-in of a Chair who has openly criticized the framework that defined it. Warsh's tenure will not be smooth. It will not be quiet. It will not follow the playbook investors have been trained to read since 2010. What it will be, if we are honest about it, is more like the Fed of the 1950s through the 1970s - a Fed that has to negotiate with politics, with fiscal arithmetic, with global capital flows, with energy shocks it did not cause. That is harder. That is messier. And it changes how every portfolio should be built.

Inflation is not transitory this time either. Anyone who tells you Warsh has the answer already is selling something. Honest position: build for resilience, prefer quality income, keep duration moderate, and watch bonds more closely than stocks. Bonds are where Warsh's verdict gets delivered first.

Research Desk, Bellwether Research, May 16th, 2026