Newsletter · 2026-09-11 · Ashutosh Agarwal

Treasury Buyback Earns a C-Minus as the 10-Year Yield Nears Five Percent - The Dollar Brief - Week of September 11, 2026

The Dollar Brief for the week of September 11, 2026. Podcast synthesis on the day the house showed its hand: the Treasury's first supersized reverse buyback accepted only 5.19 billion of a possible 6 billion dollars and drew a C-minus from Rick Santelli, the 10-year yield climbed to 4.95 percent anyway, the FT's Unhedged branded the I am the house now line egonomics, State Street put a 70 percent chance on the dollar-yen top being in, and hot PPI plus 105 dollar oil set up CPI as the decider for next week's Fed.

The Dollar Brief

Week of September 11, 2026: Treasury Buyback Earns a C-Minus as the 10-Year Yield Nears Five Percent

Yesterday we said the "I am the house now" era would get its first real test at 1:40pm Eastern, when the Treasury actually ran its supersized bond buyback. The results are in, and the grade is not flattering. CNBC's Rick Santelli, who has watched government auctions for decades, gave it "somewhere around a C, maybe even a C minus." The Treasury wanted to buy back up to $6 billion of older bonds; it managed only $5.19 billion, because the people who own those bonds were, in Santelli's word, "gun-shy" about selling. And the number that was supposed to calm the bond market did the opposite: the 10-year Treasury yield climbed to 4.95%, its highest since 2023, and kept climbing toward the round number that spooks everyone: 5%. Meanwhile a hot wholesale-inflation report landed the same morning, oil is stuck above $100, and the consumer-price report that may decide next week's Fed meeting drops today. The house made its bet. So far, the table is winning.

(Quick plain-English glossary. The "dollar" is the US dollar; the "DXY" measures it against a basket of other big currencies. A "Treasury" is a loan to the US government; its interest rate is its "yield," and yields go up when bond prices go down. A "buyback" is the Treasury buying its own older bonds back, a "reverse auction," where the government is the buyer and investors decide how much to offer for sale. "Offer-to-cover" measures how eagerly investors offered to sell: high means lots of willing sellers, low means few. "PPI" is the producer price index (wholesale prices, what companies pay); "CPI" is the consumer price index (what you pay at the till). The "FOMC" is the Federal Reserve's rate-setting committee; it meets September 16–17. When the Fed "hikes," it raises its short-term interest rate. The "carry trade" is borrowing cheaply in a low-rate currency (for years the Japanese yen) to buy higher-yielding things elsewhere; "PPP," purchasing-power parity, is a way of estimating a currency's fair value.)

TL;DR

The buyback happened, and drew a C-minus. The Treasury accepted $5.19 billion of a possible $6 billion in its reverse auction of 10- and 20-year notes. Investors only offered about $10.5 billion for sale, an "offer-to-cover" of just $2.02, versus a normal 2.5–3.0. Rick Santelli's grade: "a C, maybe even a C minus" (CNBC's Power Lunch, Sep 10).
Yields went up anyway. The 10-year hit 4.95% (highest since 2023), the 30-year 5.36%. "$6 billion sounds like a ton of money" but it is tiny against "over $30 trillion in Treasuries out there, over $8 trillion… trade every single day" (Brew Markets, Sep 10).
The FT laughed out loud. On the Financial Times' Unhedged, Robert Armstrong said traders find "this whole I am the house now thing legitimately hilarious because it puts a massive target on your back. It smacks of pride coming before a fall," and flagged the strangeness of a Treasury Secretary giving "forward guidance" on the Bank of Japan while the Fed has stopped giving it on itself (Unhedged, Sep 10).
The most credible yen call of the week. State Street's Michael Metcalfe put a 70% probability that the peak in dollar-yen is already in: "there's been 13 [turning points] in the past 50 years, and 9 of them have occurred around intervention", with a longer-term target near 120 (Street Signals, Sep 10).
The intervention was tiny. Analyst Matt Klein noted the famous Bessent note said buy "$5 to $10 billion worth of yen," but the actual reported intervention was only $500 million, "very small amounts relative to the size of the currency markets" (Top Traders Unplugged, Sep 9).
Wholesale inflation ran hot. August PPI rose 0.4% on the month and 5.4% from a year ago, "slightly hotter than expectations," led by energy (Brew Markets, Sep 10). Fed-hike odds for next week jumped to about 70%, up from a coin-flip (Best Stocks Now, Sep 10).
Today's CPI is the decider. JPMorgan's Mike Ferroli: he forecasts core CPI of 0.21% and thinks 0.20% keeps the Fed on hold, 0.29% triggers a hike: "how important is that second decimal point? Pretty important" (Bloomberg Surveillance, Sep 9).
A hedge-fund "Titanic" warning. Andromeda's Alberto Gallo said Bessent is chasing an "impossible trinity" (strong dollar, no austerity, and controlled yields), and "something has to give… initially it will be the currency. So there will be more debasement. But eventually, it might be the credit market too" (The Credit Edge, Sep 10).
Oil is the political time bomb. Brent pushed past $105–107, diesel is at records (up ~24% in August alone), and Trump's own message is that prices stay high until after November, a gift to Iran's leverage and a headache for the midterms (Unhedged, Sep 10; Wall Street Unplugged, Sep 10).
Europe hiked, and isn't done. The ECB delivered its second 25bp hike, and JPMorgan now pencils in a third in December, with the door open to a fourth (Global Data Pod, Sep 4).

What's new

The auction ran, and Santelli graded it a C-minus

This was the moment the whole "I am the house" story had been building toward, and it's worth walking through exactly what happened, because the detail is the story. On CNBC's Power Lunch (Sep 10), Rick Santelli reported the results live, and he was careful to explain the mechanics first. A buyback is a reverse auction: the government is the buyer, and investors decide how much of their old bonds to offer up for sale. So instead of the usual "bid-to-cover," you look at "offer-to-cover": how much investors threw out there versus how much got bought.

The numbers: "$10.489 billion were offered for sale. They accepted $5.19 billion." Divide the two and you get an offer-to-cover of $2.02. Santelli's read: "It's weak. Most reverse auctions, you're looking at $2.5 to $3 is normal… the sellers thrown out about $10.5 billion worth. It should have been closer to $15 or $20 billion." And the Treasury didn't even use its full $6 billion budget, which "probably means that the Treasury thought some of that stuff was a bit too expensive to buy." His verdict: "If I had to grade this auction, it would be somewhere around a C, maybe even a C minus."

There's a genuinely two-sided way to read that, and Santelli gave both. The glass-half-full version: reluctant sellers can be a good sign: "The Treasury wanted to buy it, and the sellers were reluctant… A hot IPO, you would say it's five or six times oversubscribed. This was like, ah, two times." In other words, holders of these bonds didn't want to give them up at current prices, which tells you they don't think yields are done falling. The glass-half-empty version is simpler: the operation meant to reassure the market barely moved, and Santelli flagged that the next one, on September 24, will be the real tell for whether sellers show up. Sharing the desk, private-equity veteran Orlando Bravo of Thoma Bravo shrugged at the fiscal hand-wringing ("if you have tools at your disposal to manage the balance sheet, go for it. I think it's astute"), but even he conceded the bigger force is out of anyone's hands: "$100 oil because there's conflicts in Russia, Ukraine, and Iran. I have no idea what an interest rate move is going to matter to that."

The bond market's answer: no thank you

The point of the exercise was to take pressure off long-term rates. It did the opposite. On Brew Markets (Sep 10), the hosts put it plainly: "more demand should push bond prices up… That should push yields lower. But instead, the opposite happened." The 10-year hit 4.95%, the highest since 2023; the 30-year pushed to about 5.36%. Their explanation is the one that matters: "$6 billion sounds like a ton of money. But it is pretty tiny compared to the overall bond market… over $30 trillion in Treasuries out there, over $8 trillion in Treasuries trade every single day." The real story, they argued, "is really becoming a story about just how much control the government actually has over long-term borrowing costs", and the answer the market gave was "no thank you." As one host put it, the posturing "has been very much like, I am the captain now… But the debt market is a market just like any other market… The market is going to decide that."

Why does the 10-year matter so much? Because, as Brew's hosts noted, it's "the anchor rate for the economy": most mortgages, though written for 30 years, get refinanced or sold within five to ten, and most auto loans run about six years, so the 10-year is the better proxy for real-world American borrowing than the 30-year everyone quotes.

Saxo's John Hardy added a detail that deepens the disappointment. On Saxo Market Call (Sep 10), he noted the Treasury announced $6 billion for this operation but signaled subsequent months would drop back to $4 billion, so "the market was underwhelmed, and Treasury yields shot higher." Tellingly, Hardy pointed out, "the US dollar didn't really do anything, nor did the Japanese yen." The drama was entirely in the bond market. His bar for real trouble: "we would need to be blasting above 5%" on the 10-year for this to become "destabilizing."

The FT's verdict: "egonomics"

The most quotable takedown of the week came from the Financial Times' Unhedged (Sep 10), where columnists Katie Martin and Robert Armstrong devoted an episode to what they cheerfully called "egonomics: what happens when powerful people are spoiling for a fight." They played the Bessent quote ("I have asymmetric information. I am the house now. You can bet against me if you want"), and Armstrong, to his credit, tipped his hat twice. First: the yen is up "about 6% since the end of July… appears to be working. Maybe he's just lucky, but let's give him credit." Second, for sheer nerve: "you say something like that and the facts go against you, you look like a humongous idiot… You've got to respect the guy's courage."

Then came the puncturing. "Macro people… find this whole I am the house now thing legitimately hilarious because it puts a massive target on your back. Like, it smacks of pride coming before a fall." And Martin nailed the deeper weirdness: "the Federal Reserve in the U.S. has given up on forward guidance. It's not going to tell us what it's going to do. But now you're telling us what the Bank of Japan is going to do… this is extremely weird." On the buyback, they reached for Crocodile Dundee: the Treasury said it would buy "$6 billion worth of bonds. Normally, it would buy more like two. And the market was like, you call that a buyback?" ("That's not a knife. You call that a knife?") Their bottom line on the bond move: buybacks "that aren't really moving the dial… they're just not big enough. Inflation that is already elevated. And now also you've got the oil price kicking higher again. And the market no likey."

The serious yen call: State Street says the top is probably in

Cut past the mockery, though, and the week produced a genuinely careful, data-grounded case that the yen's long slide really has turned, and it came from an operator, not a pundit. On State Street's Street Signals (Sep 10), Michael Metcalfe, State Street's Global Head of Macro Strategy, laid out why the yen is "an underdog, it's a puzzle, and now it's a battle." The puzzle: the yen recently hit "its weakest level against the dollar in nearly 40 years," yet "10-year yields actually in Japan are at a 30-year high", and higher yields usually mean a stronger currency. On State Street's PPP model, the yen has been "more than 40% cheap."

His headline call is the one to write down: "I'm going to say there's a 70% chance we've seen [the top in dollar-yen], and that's just based on looking at past turning points in dollar-yen, and there's been 13 of them in the past 50 years, and 9 of them have occurred around intervention." What makes this intervention different, he argued, is that it's being "followed up by the BOJ really tightening policy, getting rates back to neutral": intervention plus rate hikes, not intervention alone. His fair-value target: "120 would be the obvious" longer-term level. And a nice trader's nuance: if this is really about value coming back as a factor, the dollar may not be the best thing to sell against the yen: "you'd prefer to buy the yen against the Swiss," which is expensive where the yen is cheap. The catch he flagged honestly: investors are still "quite underweight or short the yen," which leaves room for more strength, but the risk is "the BOJ doesn't follow through" on the tightening the market has now priced.

Metcalfe also drew the line back to the dollar's own plumbing that ties this whole story together: the messaging from Bessent "and the actions that they've done on Treasury buybacks. The two would seem to be connected. You know, who's going to buy long-dated issuance in the US?" A stronger yen keeps Japan, a huge Treasury holder, from having to dump US bonds to defend its currency. That is the quiet logic underneath the noise.

How big was the intervention, really? About $500 million

Here's a fact that deflates a lot of the drama. On Top Traders Unplugged (Sep 9), analyst and writer Matt Klein (formerly of FT Alphaville) pointed out that the celebrated Bessent note (the one written in big letters and photographed) said to buy "$5 to $10 billion worth of yen." But according to the actual reported data, "they bought only 500 million. Either way… these are very small amounts relative to the size of the currency markets." Klein's read on the underlying motive is the most coherent one going: Bessent's real fear is Treasury yields. Normally Japan would defend the yen by selling its US dollar assets (i.e., selling Treasuries), "which… leads you back to the same point, which is higher borrowing costs for Uncle Sam." So Bessent preferred almost anything that spared the Treasury market, including nudging Japan to use a Federal Reserve facility to raise dollars against its bonds rather than selling them outright. In other words: the yen theatrics are, at heart, a bond-market defense.

Klein's broader frame is worth keeping too. He's skeptical the whole trade-and-currency campaign is even aimed at the right targets: "most of the energy the U.S. has expended in terms of tariffs and trade conflict have been with countries that are not contributing to global imbalances," while "China has comparatively been let off the hook." The latest fight with Canada he called simply "absurd."

Inflation ran hot, and today's CPI is the whole ballgame

The buyback wasn't even the morning's most important number. August PPI (wholesale, or "factory-gate," inflation) came in 0.4% on the month and 5.4% year-over-year, "slightly hotter than expectations," with "a lot of that acceleration… led by energy" (Brew Markets, Sep 10). On Best Stocks Now (Sep 10), Barry Kite flagged that the prior month's figure had been revised up from zero to 0.1%, with core PPI running "in the 4.7-ish range year-over-year, which is high," and made the mechanical point: "when you've got producer price index going up, eventually… these companies are pretty good at passing costs on to us." Fed-hike odds for next week, he said, are now "70%… so it was 50-50 for a little while." He also clocked the global picture: the ECB hiked 25bp that morning, and Japan carries "a 97% chance" of a quarter-point hike in the days ahead.

The angriest version came from Wall Street Unplugged (Sep 10), where Frank Curzio hammered the 5.4% figure against the Fed's 2% target ("the Fed's credibility right now is an effing joke") and zeroed in on the real culprit: diesel, "up 24.1% in August alone," and up roughly 80% year-over-year. His co-host Daniel Creech relayed a fresh Nestlé warning that higher Middle East energy costs are "forcing us to reformulate products," with the company vowing to be "making sure consumers come along if we have to increase prices." Curzio's fear is a pincer: either companies eat the cost and cut earnings estimates, or they pass it on and risk sales, pointing to consumer-discretionary names down 20–40% in a month as the tell. (He borrowed strategist Darius Dale's line for the mood: investors are "just frogs in boiling water on a debasement trade.")

But the number that actually decides next week lands today. On Bloomberg Surveillance (Sep 9), JPMorgan's Mike Ferroli was precise about how much rides on a single decimal: he forecasts core CPI of 0.21% month-over-month, and "if it's 0.20, I think we're in good shape [Fed on hold]. I think if it's 0.29, we'll probably change our call and look for a hike." He stressed the recent acceleration is concentrated in "the core goods bucket, which we think is both tariff pass-through and potentially energy price pass-through" (supply-side, not demand-driven) and noted wages are actually cooling, with average hourly earnings down to 3.1% year-over-year. The committee itself is on a knife-edge: Ferroli sees "a case we could see no hikes… [and] a case we could see four hikes," floated the idea that a September move might be a one-off "nip and tuck… to bolster credibility" rather than a real hiking cycle, and only half-joked about the rising influence of Governor Waller ("some people are calling him Chair Waller").

The hedge-fund view: an "impossible trinity" and a Titanic

For the bleakest and most structured take, The Credit Edge (Sep 10) hosted Alberto Gallo, chief investment officer at hedge fund Andromeda Capital and a veteran macro-credit strategist. He reached for the Titanic: after the ship hit the iceberg, "the passengers were not worried immediately. They actually wanted to stay… it's just after two hours after the impact, they started to worry." His argument is that Bessent is trying to achieve an "impossible trinity" (keep the dollar strong, avoid austerity, and keep yields under control), and "at some point… you have to let go of one of the three sides of this triangle." His base case for which side gives: "initially it will be the currency. So there will be more debasement. But eventually, it might be the credit market too."

The numbers behind it: he sees "10-year rates hitting 5% pretty soon… and 30 years could hit 6%," and warned that at current levels federal net interest costs grind from about 3% of GDP "to over 4%… over the next few years." His deeper worry is Japan, "the anchor of low yields for the last decades", because if the BOJ is forced to hike faster and Japanese savers repatriate money into their own bonds, "that's going to put even more pressure on treasuries." Gallo's frame for the whole developed world is "pro-cyclical fiscal policy" (the US "never had a 7% deficit with unemployment… below 4%"), which keeps inflation sticky and leaves bond investors in "financial repression": nominally juicy yields (a 4.8% 10-year) that barely beat real inflation. His base case is central banks that "hike, but they'll hike less than needed": one from the Fed, one from the BOJ.

Oil: the political time bomb under the whole story

Everything above runs through the price of crude, which just won't quit. Brent traded around $105 on Unhedged (Sep 10), "getting close to the peaks of the early days of the war… that 110 level", and Brew Markets clocked a 4% overnight spike past $107, the first time since May. Armstrong's rule of thumb: below $150 "it's not an energy crisis. It may be a political crisis", and with national gas at $4.27 a gallon (up a dollar in a year) and diesel near $6 (from $3.70 a year ago), "people notice that sort of thing." The politics are hard to miss: as Martin put it, "Donald Trump seems to have just sort of given up. His message to the electorate now is, oh well, prices are going to stay high into the midterms." Saxo's John Hardy noted the same resignation: Trump insists the conflict "will end immediately after the election," while "top advisors" warn Iran could dig in "through to the 2029 election."

Europe hiked, and it isn't finished

The one part of the week that went to script was the ECB. It delivered its second 25bp hike, and on JPMorgan's Global Data Pod (Sep 4), economist Raphael Brun-Aguerre said he'd just "adjusted the call this week" to add a December hike ("a third hike feels more and more likely"), with the door not closed on a fourth into 2027. The driver is the same one hitting the US: energy. He sees euro-area headline inflation "running basically three and a half until the very early part of next year," core "stuck at two and a half until next spring," and the Middle East war making a durable de-escalation hard to bank on. The mechanical read-through for the dollar is unchanged from yesterday: Europe tightening while the Fed hesitates is, all else equal, a dollar headwind, even if this week the dollar barely moved while all the action was in bonds.

The debate

Was the buyback a flop or a clever bluff that's quietly working? A real split, and Santelli voiced both sides himself. The bear case is simple and was everywhere: $6 billion is a rounding error against a $30-trillion-plus market, the operation didn't even fill, and yields rose regardless, proof the Treasury is "billions against trillions." The steel-man is subtler, and it's not naive: reluctant sellers (offer-to-cover of just 2.0) can mean holders think yields still have room to fall, so they won't part with these bonds cheaply, and Orlando Bravo's "if you have the tools, use them… it's astute" is a fair defense of active debt management. The tie-breaker is the September 24 buyback: if sellers show up in size, the C-minus was first-day jitters; if they don't, the "hubris" chorus gets its proof.

Has the yen genuinely turned, or is this another false dawn? A serious, evidence-based disagreement. Metcalfe's 70%-the-top-is-in call is the most rigorous bullish-yen case in weeks: 40% undervaluation, 30-year-high Japanese yields, and the historical pattern that big dollar-yen turns cluster around intervention. But he flagged the honest risk himself: the whole thesis rests on the BOJ actually delivering the hikes now priced in, and "the BOJ has had a history of being cautious." Matt Klein is the skeptic's anchor: the real intervention was a trivial $500 million, and the yen's failure to appreciate despite rising Japanese yields suggests the market has already repriced growth and inflation, muting the move. Both agree the yen is cheap; they disagree on whether policy will finally let it revert.

Does something break, and if so, what? Flagged as unresolved, and the scariest question on the table. Gallo's "impossible trinity" says one of three things has to give, and he thinks it's the currency first (more debasement), then possibly credit. The optimists on the same tape (Bravo, Ferroli's cooling-wages read, even the "0.20% and we're fine" scenario) say this is a supply-side, oil-driven bump that fades, wages aren't accelerating, and a single "insurance" hike settles it. The gold-and-doom crowd (see below) says the break is already underway in slow motion. What everyone agrees on: the 10-year at 5% is the line where "bad vibes" turn into something the stock market can no longer ignore.

The trades in play

These are speakers' own stated positions and views, not advice.

Buy the yen, but maybe against the Swiss franc, not the dollar. State Street's Michael Metcalfe sees a 70% chance the dollar-yen top is in, targeting ~120 over time, and argues that if value is the driver, "you'd prefer to buy the yen against the Swiss" (Street Signals, Sep 10).
Own real assets over paper. Alberto Gallo argues bond investors face "financial repression" (nominal yields that barely beat true inflation), so "you need returns that are linked to real assets, not only paper assets" (The Credit Edge, Sep 10).
Fade the "buyback saves the day" trade. Saxo's John Hardy frames the risk as one-directional until proven otherwise: yields keep leaking higher unless "we see a thaw in both the treasury market… and in the energy action", with 5% on the 10-year the level that turns a wobble into a rout (Saxo Market Call, Sep 10).
Gold as the end-of-empire hedge. Francis Hunt of The Market Sniper (a gold bull, so treat the framing as an interested opinion) points to "South Korea, for the first time in 13 years, buying gold, dumping treasuries. Same for Norway, the largest wealth fund." His reserve-shuffling data points are worth noting even if you discount the "Western collapse" thesis wrapped around them (Palisades Gold Radio, Sep 10).

Read-throughs

Watch September 24, not just today's yield print. The single cleanest test of whether the buyback strategy is working isn't the level of the 10-year (that can rise for a dozen reasons) but whether the next reverse auction draws willing sellers and a healthier offer-to-cover than this week's weak 2.0. A repeat C-minus turns "first-day jitters" into "the market is calling the bluff."
Today's CPI is a decimal-point event. Ferroli's framework is the one to hold in your head: roughly 0.20% core keeps the Fed on hold next week; 0.29% flips it to a hike. With PPI already hot and oil surging, the risk is skewed toward the upside surprise, which would send the 10-year at (or through) 5% and hand the "market beats the house" camp its evidence three trading days before the FOMC.
The yen debate is really a US-rates debate. Metcalfe's own caveat (that even a hawkish BOJ can be overwhelmed if the Fed out-hikes it) means the September 17 Fed decision may matter more for dollar-yen than the September 18 BOJ meeting. If you watch one currency for the health of the dollar system, keep it the yen; but keep your eyes on Washington.
Oil has quietly become the master variable. It feeds the inflation prints, which feed the Fed, which feeds yields, which feed the buyback's futility, and it's wired straight into the November midterms. Every week the Strait of Hormuz stays contested, the harder next week's Fed call gets and the worse the political math for the incumbents. This is the domino that moves all the others.
"Strong dollar" and "eroding trust" still coexist. The dollar and yen barely budged this week even as the bond market convulsed. Metcalfe and the plumbing logic say the dollar is well-supported, while Gallo and the gold-repatriation crowd say the foundation is quietly cracking. Both can be true: a currency can be the strongest in the room and still be slowly losing the automatic trust that made it the reserve. That's why the dollar grinds sideways-to-firm while the debasement headlines pile up.

What changed this week

The buyback stopped being a plan and became a scorecard. It ran, it under-filled at $5.19 billion of $6 billion, drew a weak 2.0 offer-to-cover, and earned a "C-minus" from the market's most-watched bond reporter, with the next test set for September 24.
Yields broke to new highs and are testing 5%. The 10-year hit 4.95% (highest since 2023) and the 30-year 5.36%, after the operation designed to pull them down, the clearest sign yet that the Treasury's tools are small against the market's size.
The mockery went mainstream. The FT's Unhedged turning "I am the house" into "egonomics" (and needling the oddity of US forward guidance for the Bank of Japan) marks the moment the slogan flipped from bravado to liability in serious financial media.
A rigorous bull-yen case arrived. State Street's 70%-the-top-is-in call, grounded in 50 years of turning points, is the most credible argument yet that the yen's multi-decade slide has actually reversed, and it reframed the buyback as part of the same dollar-defense machine.
Inflation turned back up, right before the Fed. Hot PPI (5.4% YoY), record diesel, and $105+ oil pushed hike odds to ~70% and set up today's CPI as a decimal-point coin-flip for the September 16–17 meeting.