The T-Bill Skew: Why the Treasury Is Betting Short — and Why That Bet Is Being Tested Right Now

For the past several quarters, the U.S. Treasury has run a quiet but consequential experiment in how a government finances a multi-trillion-dollar deficit: lean hard into short-term T-bills, hold coupon (note and bond) issuance roughly flat, and hope the gap gets filled later once conditions look friendlier for locking in long-term rates.

Market participants call it the "T-bill skew." The logic sounds clean in a slide deck. In practice, it's a bet — on the Fed's path, on oil, and on the world's appetite to keep rolling over America's debt every few weeks without asking for more in return. As of this week, that bet is genuinely in play, not just theoretically at risk.

The strategy, in plain terms

Treasury debt comes in two broad flavors: bills (maturing in a year or less) and coupons — notes and bonds (2 to 30 years). When the Treasury issues more bills relative to coupons, it's choosing to finance the government with debt that has to be refinanced constantly, in exchange for (usually) a lower interest rate today.

The scale of the tilt is real. Four-week bill auctions have averaged roughly $94 billion each — more than double the $47 billion average seen throughout 2016 — making the 4-week bill the single largest security the Treasury offers. Coupon auction sizes have grown too, but nowhere near as fast. Treasury has signaled it intends to hold nominal coupon and Floating Rate Note (FRN) auction sizes roughly flat for the next several quarters, leaning on bills to absorb whatever financing gap shows up.

The next Quarterly Refunding Announcement — where Treasury and its advisory committee (TBAC) formally set the borrowing mix for the coming quarter — lands August 5, 2026. That's the next real checkpoint for whether this tilt continues.

Why "skew short" made sense when the curve was inverted — and why it makes even more sense now that it isn't

For most of 2022–2024, the yield curve was inverted: short rates sat above long rates, so issuing bills was actually the more expensive choice in the near term. Treasury tolerated that cost anyway, partly to avoid locking in high 10- and 30-year rates for decades.

That constraint eased once the curve normalized. But "normalized" doesn't mean "steep." As of early August 2026 the

3-Month T-Bill is at 3.78% and the

10-Year Note is at 4.70% (near an 18-month high) a spread of 90 basis points.

That's a real, meaningful discount for going short — close to a full percentage point saved on every dollar financed at the front end rather than the long end. The curve has normalized into a genuinely upward slope, not just a marginal one.

The AI capex angle: staying out of the long end on purpose

One under-discussed rationale for the skew is what's happening in corporate credit markets. Hyperscalers, semiconductor manufacturers, and utilities are issuing enormous volumes of long-duration corporate debt to fund data centers, chip fabs, and power infrastructure for the AI buildout. That's real demand for long-duration capital, competing for the same pool of investors who'd otherwise absorb 10- and 30-year Treasuries.

If Treasury flooded the long end with new supply at the same moment corporate AI issuance is peaking, the combined supply would likely push long yields — and by extension corporate borrowing costs and mortgage rates — higher than either side wants. Staying bill-heavy is, in part, a deliberate choice to leave room at the long end for private capital rather than competing with it directly.

This is a genuinely underappreciated crowding-out dynamic, but it's worth being honest about its limits: it's a reasonable inference from issuance patterns and market commentary, not something Treasury has stated outright as policy. Treat it as the most persuasive unstated rationale rather than official doctrine.

The vulnerability: this isn't a hypothetical anymore

Here's where the original bull case for the bill skew runs into trouble, and where the story has moved fastest.

The whole strategy assumes short rates stay low or keep falling, so that rolling over bills every few weeks stays cheap. That assumption is being actively tested right now:

Rate Hikes Priced In: Markets are currently pricing roughly a 68% chance of a 25-basis-point Fed rate hike in September 2026 — the opposite of the cutting cycle the bill-skew strategy was built to exploit.

Fed Dissents: Three Fed officials dissented at the last meeting, warning that waiting too long to act could eventually force more aggressive tightening.

Macro Headwinds: The 10-year yield has been sitting near an 18-month high, while oil and Treasury markets have both been trading heavily on Middle East tension headlines, directly tying geopolitical risk to bond market volatility.

Put together: the "sticky inflation from a geopolitical oil shock forces the Fed to stay restrictive" scenario isn't a tail risk anymore. It's arguably the market's base case for the next few months. The ~90bp discount bills currently offer over the 10-year is still healthy — but it's a discount built on the expectation of near-term Fed cuts. If a September hike materializes instead, that expectation gets repriced, short rates ratchet up directly, and the bill-heavy financing structure means Treasury feels that ratchet almost immediately across trillions in debt — rather than being insulated from it the way a coupon-heavy structure would be.

That's the real story here: Treasury built its financing plan around a rate-cut path that the market is now pricing against, and the stress test is arriving in real time.

The plumbing that's quietly making this survivable: FIMA and the FX backstop

One reason a rollover-heavy strategy hasn't already caused visible strain: the Fed's Foreign and International Monetary Authorities (FIMA) Repo Facility lets foreign central banks pledge their Treasury holdings as collateral for overnight dollars, rather than selling those Treasuries outright when they need dollar liquidity to defend their currencies. That keeps large blocks of long-term Treasuries from hitting the secondary market during FX stress episodes, protecting long-end yields from an additional source of selling pressure.

It's a genuine safety valve, but it's worth being precise about what it does and doesn't do: it prevents forced foreign selling from adding to long-yield pressure. It does nothing to offset the risk sitting squarely on the short end — the risk that Treasury itself must pay whatever the market demands, week after week, if short rates stay elevated or rise further.

Where this leaves things

The bill skew delivered real, calculable savings when the curve was steep and cuts looked likely. Both of those conditions are now shakier than when the strategy was first adopted:

*The spread risk is front-and-center. While a ~90bp discount still offers near-term relief, any upward push in short-term rates quickly eats into those savings.

*The rate-cut assumption is being challenged in real time, with hike odds — not cut odds — currently the more actively discussed scenario in rate markets.

*Geopolitical energy risk is live, not theoretical, and is already showing up in daily Treasury price action.

What to watch in the August 5 QRA

As Treasury releases its borrowing statement, three key indicators will reveal whether the strategy holds or shifts:

*Bill Share vs. TBAC Recommendation: Whether bill issuance remains above the historically recommended 15%–20% range of total debt.

*Weighted Average Maturity (WAM): Any sign that Treasury is preparing to lengthen maturity profiles to lock in long rates before further volatility.

*Coupon Guidance: Whether the "flat coupon auction size" guidance holds firm for another quarter.

The strategy is still collecting a real discount today on every dollar financed short rather than long. But its entire value proposition rests on the Fed continuing to ease. Because Treasury has structurally maximized its bill share, it has correspondingly minimized its insulation from being wrong. The framing worth carrying forward isn't "a clever hedge against locking in high long-term rates." It's that the U.S. government's financing costs are now unusually exposed to two variables almost entirely outside Treasury's control: the Fed's next move, and energy market stability in the Middle East.

Generated by Gemini Flash and Claude Sonnet 5

Note: Gemini, Sonnet and everyone including humans can make mistakes !