Newsletter · · Ashutosh Agarwal

A Big Dumb Constant Buyer Steps In for the Dollar - The Dollar Brief - Week of September 23, 2026

The Dollar Brief for the week of September 23, 2026 (podcasts recorded roughly September 16 to 23, 2026): a synthesis of the week's macro and FX podcasts on Treasury Secretary Bessent's push to make stablecoins a 'big, dumb, constant buyer' of Treasury bills, the near-consensus that his bond buybacks are failing as the 30-year hit 5.33%, the fight over whether Warsh's hike was a credibility win or a policy mistake, Barry Eichengreen's warning that the real threat to the dollar comes from Washington, and how the midterms are already in the bond math.

The Dollar Brief

Week of September 23, 2026: A Big Dumb Constant Buyer Steps In for the Dollar

Here's the sentence that stuck with me from this week's podcasts. The Treasury Secretary, Scott Bessent, has been saying for years that America "needs stablecoins because we need a big, dumb, constant buyer of treasury bills." A stablecoin, for the uninitiated, is a digital token meant to be worth exactly one dollar, and to keep that promise its issuer parks the real money in short-term US government IOUs. So every dollar of stablecoins is, in effect, a small loan to the US government. Bessent wants that pile to get very large, and this week the podcasts finally started arguing about whether it will actually work.

That question sits on top of a bigger one that ran through the week: the US government is now fighting the bond market on two fronts at once. On one front it is trying to hold down its own long-term borrowing costs by buying back its own bonds, and a remarkable chorus of investors this week said, in a dozen different metaphors, that it is failing. On the other, it is trying to manufacture a brand-new class of lenders out of thin air, via stablecoins. Meanwhile the dollar itself just had another firm week and sits back above 100 on the DXY, still riding new Fed chair Kevin Warsh's decision to defy the President and raise rates. But a louder camp now says that hike was a mistake, and the most credible voice of the week argued the real long-run danger to the dollar isn't China at all. It's Washington. Let's get into it.

(Quick plain-English glossary, skip it if you live in this stuff. The "dollar" is the US dollar; the "DXY" measures it against a basket of other rich-world currencies. A "Treasury" is a loan to the US government; its interest rate is its "yield," and yields rise when bond prices fall. "T-bills" are the short-term version (under a year); the "10-year" and "30-year" (the "long end") set mortgage and corporate borrowing costs. A "real" yield is the yield after subtracting inflation. "Term premium" is the extra yield lenders demand for the risk of lending far into the future. A "breakeven" is the inflation rate bond prices imply. A "buyback" is the Treasury purchasing its own outstanding bonds to support their price. "Operation Twist" is an old tactic of buying long bonds and issuing short ones to push long-term rates down. The "basis trade" is a heavily-borrowed hedge-fund strategy that has quietly become a huge buyer of Treasuries. A "stablecoin" is a digital dollar token backed by real dollar assets. The "carry trade" is borrowing cheaply in a low-rate currency (for years, the yen) to buy higher-yielding things elsewhere. "De-dollarization" is the world slowly using and saving in the dollar less. "Fed independence" means the central bank sets rates free of political orders.)

TL;DR

  • Bessent wants stablecoins to become the dollar's permanent new lender. As a CNBC panel recounted his line, the US "needs stablecoins because we need a big, dumb, constant buyer of treasury bills," and he has vowed to be "the world's leading bond salesman," with the share of US debt issued as short-term bills already up to about 24% (Squawk on the Street, Sep 21).
  • But the new demand may be an illusion, warns Jim Bianco. If Americans just move cash from money-market funds into stablecoins, "the net impact is zero," the money already sat in T-bills. The real prize is 2 billion people abroad with unstable currencies actually buying dollars for the first time (Bitcoin Magazine, Sep 21).
  • Even the bullish math is modest. On a Bitcoin-focused show, Saifedean Ammous walked through the academic estimates: a six-fold jump in stablecoins (to ~$1.5 trillion) might shave only about 20 basis points off US interest rates; the most bullish study caps it near 40bp (The Bitcoin Standard, Sep 22). (Openly pro-Bitcoin venue, weigh accordingly.)
  • The Treasury's bond buybacks are being mocked as futile. It doubled them to $4 billion a round while selling roughly $550 billion of new bonds this quarter, "a garden hose on a wildfire," and by week's end the 30-year hit 5.33%, its highest since 2007 (Money Rehab, Sep 21).
  • Robin Wigglesworth: "even if he 10x's it, it's like peeing on a forest fire." The FT writer argues the real point of the buybacks is to calm the hedge-fund "basis trade", which authorities fear "would be the next big shoe to drop" (Top Traders Unplugged, Sep 16).
  • Jeffrey Gundlach thinks the buybacks are cover for a coming Operation Twist. With the deficit "headed towards 50 trillion," he sees only two real exits: inflate the debt away, or restructure it by stretching maturities and cutting coupons (The Julia La Roche Show, Sep 16).
  • A growing camp calls Warsh's hike "a policy mistake." David Rosenberg said the Fed's statement reads "like we're in some sort of economic boom. You would not even know that the housing market exists" (RiskReversal Pod, Sep 17). Wellington's Mike Medeiros took the other side: "cutting rates in this environment would be clinically insane" (The Wall Street Skinny, Sep 18).
  • The most authoritative voice of the week: the threat to the dollar is home-grown. Berkeley's Barry Eichengreen (a leading historian of reserve currencies) argued "if there is a threat to the continued global role of the dollar, it's coming from Washington, D.C.," and noted foreign investors have gone from holding a majority of US Treasuries a decade ago to a minority today (The Disciplined Investor, Sep 20).
  • The midterms are quietly a floor under long-term rates. Trump's proposed $5,000 "dividend" check means that if a GOP win looks likely, investors will "demand additional yield" for the deficit risk, capping how far long rates can fall (Macro Horizons, Sep 18).

What's New

The dollar's new buyer of last resort: a "big, dumb" stablecoin

Start with the idea that got the most fresh airtime this week, because it reframes the whole dollar debate. For years the worry has been that the world is drifting away from the dollar. Bessent's counter-move is to invent a new source of demand for it, and the vehicle is the stablecoin.

The logic, as a CNBC panel laid it out, is blunt. Bessent "over the years has said we need stablecoins because we need a big, dumb, constant buyer of treasury bills," one panelist recounted. "You totally need it if you want to be able to finance the deficit in this way for a long period of time. He says, I'm going to be the world's leading bond salesman." The mechanism is simple: to keep a token worth a dollar, the issuer holds real dollars in T-bills, so a bigger stablecoin market is automatically a bigger buyer of US debt. The panel noted the Treasury is already leaning hard on short-term borrowing ("the bill share is hitting 24 percent," well above the historic norm), which is exactly the kind of debt stablecoins soak up (Squawk on the Street, Sep 21).

How big could it get? On a crypto-industry show, a stablecoin-platform executive put the current market "right around 300 billion" and made the striking point that "a little under two percent of that is yield bearing," most stablecoins just sit idle as trading collateral. Bessent, he said, is "pounding the table to get stablecoins bigger and more proliferated… because it's a great place to stuff his bonds in and his bills. And export the world dominance of the U.S. dollar to every nook and cranny of the world" (CRYPTO 101, Sep 21). (Crypto-platform venue, the guest sells stablecoin infrastructure, so weigh the enthusiasm accordingly.)

Here's the catch, and it came from a genuine market strategist. Jim Bianco of Bianco Research pointed out that if the growth is just Americans shuffling money around, it changes nothing: "if you sell your money market fund and you put your money into a stablecoin, the net impact is zero," because your money-market fund already owned T-bills. "They already own stuff that's backed by T-bills; they're just going to swap into something else that's backed." The only way stablecoins create real new demand for US debt is if the buyers are new, and Bianco thinks they are, abroad: "there's 2 billion people" in Latin America, Africa, the Middle East and southern Asia with "unstable banking systems" and "currencies that are constantly devalued" who "need this product desperately." If Bessent's $3-4 trillion vision is driven by those people, "that could be big demand for the US treasury." If it's just domestic reshuffling, it's a wash (Bitcoin Magazine, Sep 21).

And even the optimistic case is smaller than the hype suggests. On a Bitcoin podcast, Saifedean Ammous did the unglamorous work of reading the academic papers. Economists Azzimonti and Quadrini, he said, model a six-fold rise in stablecoins (from about $250 billion to $1.5 trillion) and find it would "lower U.S. interest rates by only about 20 basis points." Their most bullish scenario tops out near 40bp, a figure Fed Governor Stephen Miran has since stretched into an economy-wide claim. His verdict: helpful at the margin, but "20 to 40 basis points" is the realistic range, not a rescue (The Bitcoin Standard, Sep 22). (Anti-fiat venue; the host's broader argument is that this is more bullish for Bitcoin than for the dollar.)

The buyback that isn't working

While one hand of the Treasury tries to conjure new lenders, the other is trying to hold down long-term borrowing costs directly by buying its own bonds, and this week the podcasts were close to unanimous that it isn't working.

The clearest plain-English autopsy came from Money Rehab. The Treasury "doubled the amount of its own long-term bonds that it buys back from $2 billion to $4 billion per round starting September 9th," the host explained. "And for exactly one day, it worked. Yields dropped. Then investors did the math." The problem is scale: the Treasury is buying back roughly $14 billion while selling "roughly $550 billion of new bonds this quarter alone. That is a garden hose on a wildfire." By the end of the week the 30-year Treasury "hit 5.33%… the highest it's been since 2007." Her framing of why is worth keeping: the yield curve is steepening, but "that's not a recession alarm. It's an inflation and a debt alarm. Investors aren't worried that the economy is about to stall. They're worried about getting paid back in dollars that buy less than they do today." The bottom line: "The Treasury can buy back bonds for a day. It can't buy back the market's confidence" (Money Rehab, Sep 21).

Robin Wigglesworth (the Financial Times writer who wrote a book on the rise of these leveraged trades) was blunter. "Whenever this pod goes out," he said, "he will not succeed. The buybacks (increasing those buybacks), I mean, even if he 10x's it, it's like peeing on a forest fire." His reasoning is the iron law of government finance: "every dollar the U.S. Treasury spends is a dollar they have to borrow. They're just switching one dollar for another dollar of debt." But he offered the most interesting theory of what Bessent is really up to. The buybacks, he suspects, are meant to "cap the volatility" in the bond market to protect the hedge-fund "basis trade", a huge, heavily-borrowed strategy that has become one of the market's biggest Treasury buyers. Authorities "were quite rightly worried that the basis trade would be the next big shoe to drop," and there are "already signs of weakness in some of the weaker players in that trade" (Top Traders Unplugged, Sep 16).

Jeffrey Gundlach, the bond investor, thinks the stated reason for the buybacks doesn't even add up, which tells him something bigger is coming. The Treasury claims it's smoothing out illiquid older bonds, but "the bid-ask spread on an off-the-run treasury is probably about three-eighths of a point worse at most… it's like a couple of basis points. What difference does it make?" His read: "I think it might be just… a full-on Operation Twist is what they really want to do, to control long-term rates." And he framed the endgame starkly. With a deficit "headed towards 50 trillion" and potentially "10 or 12% of GDP in the next recession," he sees only two ways out: "inflate the deficit away by printing money," or "restructure the treasury debt by extending maturities and or reducing coupons." Cutting entitlements, he added, "isn't going to do it because you've already got all that debt." That's why he keeps moving to shorter maturities and higher-quality bonds, "further and further away from the areas that are going to suffer the most" (The Julia La Roche Show, Sep 16). Former Fidelity manager George Noble, on Steve Eisman's show, put the same view more colorfully, describing Bessent's whole effort to "repress interest rates" as a losing battle in which the Treasury Secretary has been steadily losing credibility as rates rise anyway (The Real Eisman Playbook, Sep 21).

"A policy mistake", the case against Warsh's hike, and the case for

The dollar is firm largely because Warsh raised rates when the President wanted cuts, a story we've followed all week. But this week a serious camp argued he got it wrong.

The loudest was David Rosenberg, the economist, whose episode was literally titled "The Fed Just Made a Policy Mistake." His complaint isn't the quarter-point itself; it's the story the Fed told to justify it. The statement, he said, was "basically pretty ridiculous… It's like we're in some sort of economic boom. You would not even know that the housing market exists when you look at that press statement. Everything is resilient, robust, strong." He noted the market's cool response ("the bond market really didn't respond that constructively," with the two-year jumping and the ten-year losing its rally) and predicted the doves who "threw in the towel" won't move before the election, so the next real hike lands "in December" (RiskReversal Pod, Sep 17).

The opposite corner was well argued too. Mike Medeiros, macro strategist at Wellington, said the decision "had nothing to do with President Trump's wishes" and that "cutting rates in this environment would be clinically insane." His case: "financial conditions are too loose… the economy is growing above trend, the unemployment rate is falling, not rising, and inflation has been above target for five and a half years and is still rising." He called it "a recalibration" and warned this is "the start of, at a minimum, a short tightening cycle," because "the Fed rarely does one hike and then stops" (The Wall Street Skinny, Sep 18).

The most useful swing vote came from Joseph Wang, a former Fed trader, who explained why the hike mattered for anyone lending to the US long-term. The great fear, he said, had been "that President Trump would appoint someone, and then Trump would end up running monetary policy," and Warsh had earlier "talked a big talk, then went on stage, did nothing," even floating the idea of revising the inflation index, which for "a bond investor" was "reason to panic." This week changed his mind: "today, we see that he is indeed someone who is serious about getting inflation under control." Wang also flagged two plumbing shifts worth watching: Warsh may scrap the Fed's "dot plot" forecasts entirely (former Bank of England Governor Mervyn King, a dot-plot skeptic, is on his communications task force), and the Fed could quietly tilt its bond holdings toward T-bills to "match perfectly what the Treasury is doing": "Treasury issues more bills, Fed buys more bills" (Monetary Matters, Sep 16).

The real threat isn't Beijing. It's Washington.

If you read one thing this week, make it Barry Eichengreen. A Berkeley professor and one of the world's leading historians of reserve currencies, he cuts through the noise on de-dollarization with unusual authority, and his conclusion should unsettle the dollar bulls even as the currency sits near its highs.

He refuses the false choice. "The de-dollarization story can be overstated and the dollar can be at risk going forward, both at the same time," he said. Yes, officials worldwide "are actively trying to create alternatives, develop alternatives, hold their reserves in other currencies," he pointed to fast-payment systems like Brazil's PIX and India's UPI being linked across borders so countries can swap currencies "without having to go into the foreign exchange market and first buy U.S. dollars," and to banks tokenizing deposits on blockchains to bypass the US banking system. But "developing these alternatives takes time. And until they are more fully developed, the dollar will continue to be the central currency" (The Disciplined Investor, Sep 20).

The sting is where he locates the danger. Dominant currencies, he argued, "are often damaged from within rather than displaced by a rival," and that is exactly the conclusion of his new book, Money Beyond Borders: "if there is a threat to the continued global role of the dollar, it's coming from Washington, D.C. It's the United States of America shooting itself in the monetary and financial foot by issuing large amounts of public debt that the markets come to view as unsustainable, through threats to the independence of the Federal Reserve, which are deeply demoralizing to international investors." The receipt he cited is the one that ties this whole issue together: "We've moved from a situation 10 years ago where foreign investors held the majority of U.S. Treasury securities. Now they hold a minority. We can't count on them to the same extent to hold our assets and finance our deficits" (The Disciplined Investor, Sep 20).

That is the quiet thread under everything else this week. Foreigners are stepping back. So the Treasury is scrambling for replacement buyers (hedge funds via the basis trade, stablecoins via the GENIUS Act, its own buybacks) and betting a visibly independent Fed keeps enough trust alive to hold the whole thing together. Eichengreen's point is that the same government propping the dollar up with Warsh's independence is the one undermining it with its debt and its politics.

The midterms are already in the bond market

One more fresh, concrete point on why long-term rates won't fall easily. An analyst on Macro Horizons zeroed in on Trump's proposed $5,000 "dividend" payout (a giant tariff-funded check to households) as a specific election risk the bond market is starting to price. "If the pendulum of sentiment starts swinging in favor of a GOP victory, investors may begin to demand additional yield to compensate for the potential deficit and inflation risks," the analyst said. Even if the checks never arrive, "the risk of a dividend should limit the extent to which long-end rates can decline." He also anchored just how relentless the move up has been: the 10-year closed "above 5% for the first time since July 2007" and is "en route for their seventh consecutive month of monthly increases, matching the longest streak since 2011 and before that since 1978," and crucially, "this has largely been a real yield story," with inflation expectations up less than 10 basis points against a nearly 100-basis-point rise in real yields since the summer's conflict began (Macro Horizons, Sep 18). The base case that show is working with: the House flips to the Democrats, the GOP holds the Senate, a split Congress, and a president who plans to govern mostly around it.

The Debate

Where the podcasts genuinely disagreed this week.

Do stablecoins actually create new demand for the dollar, or just reshuffle old demand? This week's freshest argument. Bessent's team and stablecoin backers see a "big, dumb, constant buyer" that finances the deficit and exports the dollar to billions of un-banked people (Squawk on the Street, Sep 21; CRYPTO 101, Sep 21). Jim Bianco's rebuttal is that domestic adoption is a wash ("the net impact is zero"), so the entire benefit rides on genuinely new foreign buyers (Bitcoin Magazine, Sep 21), and the academic ceiling even in the bullish case is only 20-40bp of rate relief (The Bitcoin Standard, Sep 22). Referee: whether the stablecoin float grows from foreign wallets or just drains US money-market funds.

Was the hike a credibility win or a policy mistake? Wellington's Medeiros and former-Fed Joseph Wang say it was necessary and restores trust: loose conditions, above-trend growth, inflation above target for years, and a Fed that just proved it won't be bullied (The Wall Street Skinny, Sep 18; Monetary Matters, Sep 16). Rosenberg says the Fed is tightening into a slowdown it refuses to acknowledge, describing a boom that doesn't exist (RiskReversal Pod, Sep 17). Referee: the next two jobs and inflation prints, and whether December's hike actually lands.

Can the Treasury control the long end at all? The near-consensus this week is no: buybacks are "a garden hose on a wildfire" (Money Rehab, Sep 21), "like peeing on a forest fire" (Top Traders Unplugged, Sep 16), and only a full Operation Twist or debt restructuring would move the needle (The Julia La Roche Show, Sep 16). The unspoken other side is the Treasury's own bet that steady buybacks plus a stablecoin buyer plus a hawkish Fed can, together, keep a lid on things. So far the 30-year at 5.33% is siding with the skeptics.

The Trades in Play

These are speakers' own stated positions and views, not advice. Treat openly-interested sources accordingly.

  • Move shorter and higher-quality; brace for a shock. Jeffrey Gundlach keeps shifting to shorter maturities and better credit, "further and further away" from the AI-borrowing epicenter, and "would be surprised if there wasn't some sort of a shock… maybe an inflation shock, maybe a supply shock" within a year (The Julia La Roche Show, Sep 16). (Bond fund manager, DoubleLine.)
  • The dollar is overvalued; short the euro. On Forward Guidance, macro investor Vincent argued the dollar is overvalued and made the case for euro weakness, pairing short-euro positions against the yen and Brazilian real. His thesis: Europe, like Japan before it, will ultimately "debase" its currency because it can't fix its finances through productivity, Japan let the yen fall ~50-65% over 15 years and "it did work" (Forward Guidance, Sep 18). (Research-firm macro strategist.)
  • Position defensively; stocks could fall 10% by year-end. Wellington's Medeiros expects a Fed that isn't done and sees downside for risk assets into year-end (The Wall Street Skinny, Sep 18). (Asset-manager macro strategist.)
  • Own the long bond (carefully) as a bet that Warsh wins. Joseph Wang feels "much better" about the long bond after the hike, framing it as a wager "that the Fed is going to get inflation under control," now more plausible with a credible chair (Monetary Matters, Sep 16). (Former Fed trader, research publisher.)
  • Fade the buyback rally. The shared trade behind Rosenberg, Wigglesworth and Gundlach: don't trust any dip in long-term yields driven by Treasury buybacks, because the supply is too big for the tool (RiskReversal Pod, Sep 17; Top Traders Unplugged, Sep 16).

Read-Throughs

  • The Treasury, not the Fed, is now the dollar's main character, and it's fighting a two-front war. Front one: hold down long-term rates by buying back bonds (the podcasts say it's failing). Front two: manufacture a permanent new lender out of stablecoins (unproven, and possibly just a reshuffle of existing money). Watch the 30-year: at 5.33% it's telling you front one is losing.
  • Stablecoins are a real lever, but a small and conditional one. The honest bull case is 20-40bp of rate relief, and only if the growth comes from genuinely new foreign holders rather than Americans swapping out of money-market funds. Treat "$3-4 trillion saves the deficit" as a sales pitch, not a forecast.
  • The long-end move is a real-yield story, not an inflation story. Breakevens are flat; real yields are doing all the work, sitting near an 18-19-year high. That's a message about trust and supply, not prices, which is precisely why the Fed's rate decisions can't fix it.
  • Washington is the swing factor for the dollar, in both directions. Warsh's independence is a genuine prop under the currency right now. But Eichengreen's warning is that the same government is eroding the dollar's foundation through unsustainable debt and political pressure on the Fed. The independence win and the debt threat are two sides of one coin.
  • The election is now a rates story. A visible tilt toward a GOP win (and Trump's $5,000-check promise) puts a floor under long-term yields regardless of what the Fed does. Between now and November 3, political polling is a bond-market input.

What Changed This Week

  • Stablecoins graduated from a crypto sideshow to a serious dollar-funding debate. The question shifted from "are they legal?" to "do they actually create new demand for US debt, and how much?", with a credible skeptic (Bianco) and sober academic math (20-40bp) puncturing the biggest claims.
  • The consensus that Treasury buybacks are failing hardened into near-unanimity, captured in a run of vivid images (garden hose on a wildfire, peeing on a forest fire) as the 30-year pushed to a post-2007 high of 5.33%.
  • A louder "policy mistake" camp emerged to challenge the Warsh-credibility narrative, led by Rosenberg, even as Wellington and a former Fed trader defended the hike as necessary and trust-restoring.
  • The de-dollarization debate got its most authoritative voice yet, with Eichengreen relocating the real threat from Beijing to Washington and putting a number on it: foreigners have gone from a majority to a minority of Treasury holders in a decade.
  • The midterms entered the bond math explicitly, via Trump's $5,000-check proposal as a concrete reason long-term rates struggle to fall.