The Bond Market Isn't Testing Warsh — It's Testing the Treasury

Things started getting a bit ridiculous in the long bond again. The 30-year Treasury yield has pushed above 5.3% — you have to go back to 2007 to find a level like it — and the 10-year keeps climbing, pressing toward 4.7% this morning even though the Fed's short-term rate hasn't moved since December. The headlines are calling it a bond squeeze and asking whether the new Fed chair, Kevin Warsh, is losing control of the bond market. Warsh delivers his first Jackson Hole keynote today, and the argument he's stepping into is the precise opposite of the scare: the claim, made by Peter Tchir of Academy Securities, that Warsh doesn't run much risk of losing control of bonds. Say what you want, but that is the more interesting claim, because it's a statement about mechanics rather than mood — and the mechanics decide what you should actually be watching with your own money.

The bond market isn't betting on the Fed's interest rate. This is where most people misfire. The Fed owns the front end — the overnight rate, which has sat at 3.50%-3.75% since December after a cumulative 1.75 percentage points of cuts that began in September 2024. Here's the thing: longer-term yields are higher today than they were the day those cuts started. Rate cuts are supposed to pull long yields down. When the long end rises anyway, the market has stopped betting on the Fed's rate path and started repricing the compensation for owning thirty years of US government credit. That's the term premium, and decomposition work puts it back near the levels Warsh first saw as a governor in 2006. What's driving it: a national debt heading toward $40 trillion that keeps growing, a debt-financed war with Iran, and — in former Fed chair Ben Bernanke's retelling — a shift from a "global savings glut" to a "global savings squeeze," in which governments, aging populations, and a wall of AI capital spending all compete for the same dollars. An overnight policy rate barely touches any of that.

The person actually running the long end is the Treasury Secretary. And this is where the "who's in control" question stops being a metaphor. On August 19, the day after the 30-year hit its highest level in 19 years, Treasury Secretary Scott Bessent doubled the Treasury's buyback operations in long-dated debt from $2 billion to at least $4 billion per operation. The 30-year dropped — about nine basis points in a day — and then, get this, gave it all back within two days. Bessent says he's ready to go bigger. CNBC reported the Treasury could fund purchases from its roughly $950 billion cash account at the Fed, which two senior Treasury officials called "considerable firepower." He's taken to calling the operation a "Treasury Twist" — the fiscal cousin of the Fed's 2011 Operation Twist.

Read the plumbing, not the name. The Treasury General Account is money parked inside the Fed's balance sheet. A bigger TGA drains reserves from the banking system; a smaller one injects them. When Bessent spends that cash buying bonds, he does two things at once: he puts a buyer under the long end, and he pushes reserves back into the banking system — a liquidity injection that happens with the Fed nowhere in sight. That is real firepower, and it's probably a big part of why the equity market can sit a step from record highs while the 30-year trades where it trades. The plumbing is being held up by the Treasury.

Here's the problem with the whole construction: the buyback program Bessent scaled up was created in May 2024 to lubricate trading in older, less liquid bonds — a market-making aid, not a price-management tool. Deutsche Bank does the arithmetic that exposes the gap: roughly $100 billion of 30-year buybacks would shorten the average maturity of the national debt by about a month. A month. This is a support operation, not a solution. The debt isn't shrinking, and no one is pretending that the deficit is the problem being solved. Tchir, for one, expects the interest-expense line to keep drifting higher no matter how many bonds get bought back.

Which brings us back to Tchir, and to why he isn't worried. He calls the buyback expansion "kind of mediocre," and his prescription is where his real confidence lives: the Fed needs to take rate hikes off the table, talk about cutting, and — if it actually wants to flatten this curve — run its own Operation Twist, buying the long end and selling the short. On this read, Warsh doesn't run much risk of losing control of bonds because the controls are all still sitting unused. The front-end lever is un-pulled. The balance-sheet tool is un-deployed. Add a Treasury Secretary willing to stand behind the market, and the picture isn't a Fed losing its grip. It's a transfer of control — from the committee's overnight rate to the Secretary's cash pile.

Now grant the bull case its due: a real loss of central-bank control has signatures, and we're not seeing them yet. Inflation expectations haven't broken loose. The Fed's short rate is pinned. The S&P 500 is up about 13% this year, and — worth noting — the equal-weight index is up more, so this isn't the usual mega-cap mirage (today, in fact, is the exception, with the cap-weighted index higher while the equal-weight one is lower). The options book shows heavy put open interest relative to calls at implied volatility near 11.5% — a market hedged against a fall, not screaming about one. And when the long-end selloff knocked stocks lower for three straight sessions earlier this month, equities clawed back into this week once the Treasury answered. By those measures, Tchir has a point. No loss of control. Yet.

The fragility is specific. The Treasury market's foundational assumption since 1982 is that the issuer is regular and predictable — that the government sells debt on a schedule and doesn't manage its own market. Bessent has now surprised that market twice in a month, and the compensation investors demand for surprise is a higher term premium. That's the 48-hour story in one line: an intervention can smooth the move, but it cannot change the sign of the repricing, because the buyer and the source of the supply are the same institution. You can't buy back your way out of a deficit that finances the purchases. And the cracks have names. A large foreign holder of US debt, Japan, is under pressure to defend the yen — the same incentive that has preceded foreign selling before — and the dollar's slide is itself an inflation import the Fed would rather not have to answer for.

Which is why the real question today isn't whether Warsh has lost control. It's what he's willing to do with it. The whole case that control holds rests on one assumption: that Warsh won't validate the market's repricing. Three of his own colleagues voted for a quarter-point hike in July, and traders have at times priced better-than-even odds of a hike at the September meeting. If Warsh leaves that on the table, the front end reprices higher, the front end starts leading the long end, and the Fed and the bond market begin chasing each other — the loop that broke risk assets in October 2023. That's the actual loss-of-control scenario, and it arrives through your stocks, not through a yield chart: the day bonds stop being ignored. So watch the things that would tell you the transfer of control is holding or failing: what Warsh says today; whether Bessent's first expanded buyback on September 9 does more than pop yields for an afternoon; the September 15-16 meeting; and the day the 10-year stops rising alone and starts bringing the whole book down with it.

Here's the part worth internalizing. The bond market isn't warning Warsh that he's losing control. It's warning him that control now has a price — a $950 billion account, a September pivot, an Operation Twist away — and that the seller of all this supply is, for the moment, also the buyer. The moment that stops being true, the headlines stop being wrong.