PRIM3 Brief #7: Treasury Issuance H2 2026 and the Crypto Carry Trade
The U.S. Treasury's refunding announcement for H2 2026 implies meaningful skew toward short-duration issuance, bills and 2-year notes, to fund the deficit through the year-end. The headline number isn't the news. The composition is. And the composition has real, measurable downstream effects on stablecoin economics, basis spreads, and the structured-DeFi carry trades that the institutional cohort has been quietly building positions in for the past 18 months.
Most crypto-side coverage isn't connecting the macro to the on-chain mechanics. PRIM3 has been doing exactly that work for our LP communications and we think the framing is worth making public.
The Treasury Mechanics, Compressed
The Treasury issues debt across the curve to fund operations. The composition decision — how much short-duration (bills, 2-year notes) versus medium-duration (5- and 7-year notes) versus long-duration (10-, 20-, 30-year bonds), is determined by the Treasury Borrowing Advisory Committee recommendations and Treasury staff calls on optimal financing cost.
In H2 2026, the composition skew is meaningful. Treasury is expected to issue roughly $1.4–1.6T in net new debt in H2, of which a disproportionate share (~70%) is short-duration. That skew has the effect of holding short-duration yields slightly elevated relative to what they'd otherwise be given the Fed's cutting trajectory.
For the crypto-adjacent reader, that may sound technical. The implication is concrete: the SOFR curve in H2 2026 is going to be flatter than it would otherwise be, with the short end holding above 3.5% even as the Fed cuts further. The flatness of the short-end yield curve is the single most important macro variable for stablecoin economics in 2026.
What This Does To Stablecoin Issuers
The major USD-pegged stablecoin issuers, Tether, Circle, and the new generation of yield-bearing stablecoin operators (Ethena, Sky's USDS, the second-tier MMF-collateralised entrants) — all earn the spread between (a) the yield on their reserve assets (predominantly short-duration Treasury bills and tokenized MMFs) and (b) the yield they pay holders (mostly zero for USDC/USDT, partial for yield-bearing variants).
The H2 2026 short-duration yield environment is favourable for issuer economics in two ways. First, the absolute level of the short-duration yield (3.5%+) is a sustainable revenue line. Second, the flatness of the short end means the issuers can hold a relatively short-WAM reserve portfolio without sacrificing yield against a longer-duration alternative, which preserves their flexibility on redemption capacity.
The implication for the issuer cohort: 2026 net interest income is on track to be the highest in stablecoin industry history. Circle's full-year reported NII for 2026 is on track to exceed $2.5B by our estimate, against $1.7B in 2025. Tether's is, conservatively, going to clear $8B. These are not businesses in any kind of stress. They are businesses producing record cashflow in a yield environment specifically tuned to their economics.
What This Does To Yield-Bearing Stablecoins
The yield-bearing stablecoin segment — USDtb from Securitize-Ethena, Sky USDS, and the second-generation MMF-passthrough entrants, has more complex economics because they pay yield to holders. The H2 2026 environment is still favourable, but the competitive dynamics are different.
In a 4% short-duration environment, a yield-bearing stablecoin can credibly pay 3.5–3.8% to holders and capture 20–50bps for the issuer. That's an attractive product for both sides. The cohort that's compounding fastest in 2026, by AUM and by integration depth, is exactly this segment. We expect aggregate AUM in yield-bearing stablecoins to cross $30B by year-end 2026, up from roughly $9B at the start of the year.
For founders building in this segment: the next 18 months are a structural opportunity, but the moat is execution and distribution rather than economic design. The unit economics will get commoditised fast as more issuers enter; the winners are the ones that build distribution into institutional treasury management, embedded fintech, and high-volume on-chain protocols where the yield-bearing stablecoin can replace idle stablecoin balances by default.
The Crypto Carry Trade
The most interesting downstream effect is on the structured carry trades the institutional cohort has been quietly assembling.
The core trade: hold a tokenized MMF or yield-bearing stablecoin earning 4%-ish. Use it as collateral for a perp short position that pays funding when basis is positive (which it has been, for sustained periods, in 2026). Net annualised carry: 8–15% with delta-neutral exposure to the underlying spot price.
This trade isn't new, basis trades have existed since 2017. What's new in 2026 is (a) the institutional-grade collateral layer (tokenized MMFs with credible settlement infrastructure) and (b) the execution surface on intent-based DEX venues that allow the trade to be assembled and unwound at institutional size without market impact.
The institutional desks running this trade in 2026 are the same multi-strategy hedge funds we wrote about in PRIM3 Brief #1. The volume they're running is much larger than the public-facing data suggests. The downstream implications for the protocols providing the rails, MMF issuers, perp DEXes, on-chain settlement venues, are meaningful.
What This Means For Builders
A few specific implications:
For stablecoin and tokenized-MMF founders: the 2026 macro tailwind is real and durable. The constraint is institutional distribution. The teams that win this category are the ones building credible institutional sales channels, not the ones with the most elegant on-chain implementation. Hire your first institutional sales head in Q2 2026 if you haven't already.
For perp DEX founders: the basis-trade volume is a meaningful share of your flow in 2026. Optimise for institutional execution quality — tight spreads, deep books, professional API access, credible custody integrations. The retail-perp era is winding down. The institutional-perp era is starting.
For DeFi infrastructure founders: the carry-trade activity composes naturally with structured product platforms, on-chain prime brokerage, and credible cross-margining infrastructure. We see three sub-sectors as underbuilt for the volume that's coming.
For protocol treasurers: if you're not running a yield-bearing stablecoin or tokenized MMF position for your treasury, you're leaving free yield on the table. The operational bar for credible institutional positions has dropped meaningfully. Talk to your finance team.
The Risk Frame
The H2 2026 carry trade is favourable but not riskless. The main risks: (a) a basis collapse driven by a CEX dislocation could compress trade economics quickly; (b) regulatory changes to yield-bearing stablecoin treatment could re-rate the underlying collateral; (c) a counterparty failure at the perp-execution layer could create idiosyncratic loss exposure that the structured trade doesn't price.
Sophisticated desks hedge each of these. Less-sophisticated desks running similar trades have produced the largest loss events in DeFi history. The trade is real and durable; the implementation discipline matters more than the trade idea.
Where We're Allocating
PRIM3 is overweight on yield-bearing stablecoin infrastructure, tokenized MMF distribution rails, and institutional-grade perp execution venues. We're closely watching the structured-products layer that composes the basis trade for institutional users. We're cautious on consumer-facing yield-bearing stablecoin distribution plays — the consumer wedge is harder to defend than the institutional wedge in this segment.
If you're building credibly in any of these sub-sectors, we'd be interested. Pitch us via prim3.vc.