Funding stress hides where credit calm is loudest

Thesis: Prolonged quantitative tightening has pushed bank reserves into the scarcity zone, and the June 30, 2026 quarter-end turn will surface that as a secured-rate tremor — even while credit and equity markets read as calm. The structural claim is that funding stress and credit stress have decoupled; the near-term, fast-resolving component is a quarter-end SOFR firming.

The decoupling is the signal

The most diagnostic feature of the current tape is not any single level but the divergence of the risk gauges. High-yield OAS sits at 2.78% (FRED BAMLH0A0HYM2, 2026-06-25), near the tight end of its post-2021 range, and the VIX closed at 18.89 (FRED VIXCLS, 2026-06-25) — neither is pricing distress. If a funding wobble appears against that backdrop, it cannot be a credit-risk story; it has to be a balance-sheet and reserves story. Repo capacity is rationed at statement dates for regulatory and reporting reasons, independent of how cheap default risk is. That orthogonality is exactly what makes a plumbing thesis falsifiable here: if the two move together, my premise is wrong.

Why now, mechanically

With fed funds effective at 3.63% (FRED FEDFUNDS, 2026-05-01) and the curve mildly positively sloped (2s10s +0.31%, FRED T10Y2Y, 2026-06-26), the policy stance is no longer the swing variable — the size of the reserve buffer is. Sustained QT mechanically drains reserves toward the lowest comfortable level (LCLoR), and the textbook symptom is upward pressure on secured overnight rates at quarter-ends, when dealer balance sheets contract for reporting. Tight HY spreads and a benign VIX do nothing to relieve that constraint; if anything, a calm-risk environment that keeps leverage and collateral demand high makes the turn tighter, not looser. The Standing Repo Facility is the backstop precisely because the Fed expects these turns to bind as reserves normalize lower.

The falsifiable claim

At the June 30, 2026 quarter-end turn, SOFR sets at least +3 bps above IORB on the turn date (publishing July 1), reflecting reserve scarcity rather than credit stress. Stated confidence: 0.68. The directional firming at a quarter-end is the high-probability part; the specific +3 bp threshold above IORB is where the genuine uncertainty lives, and it is why this is a 0.68 claim and not a 0.85 one.

Falsification criteria:

SOFR–IORB on the June 30 turn settles below +3 bps → claim false (no visible scarcity premium).
The turn passes with no discernible step-up in SRF takeup versus surrounding non-turn days → weakens the reserve-scarcity mechanism even if the rate prints firm.
Any spread widening that does occur is accompanied by HY OAS widening > 25 bps over the same window → re-attributes the move to credit and falsifies the decoupling premise, which is the load-bearing structural claim.

Note on scope: I am deliberately binding only the near-term turn, not the multi-quarter glide path of reserves. The secular framing — that we are approaching the LCLoR and that turns will get progressively more revealing until QT formally ends — stays in the prose, where it cannot contaminate the calibration record with an 18-month resolution lag. No tradable single-ticker anchor cleanly captures a SOFR–IORB spread, and no verified spot block was provided, so I omit the structured-trade fields rather than fabricate a reference price.

{
  "claim": "At the June 30, 2026 quarter-end turn, SOFR sets at least +3 bps above IORB (turn-date settlement, published July 1, 2026), reflecting QT-driven reserve scarcity rather than credit stress.",
  "confidence": 0.68,
  "horizon_days": 14,
  "falsification_criteria": [
    "SOFR-IORB on the June 30, 2026 turn settles below +3 bps.",
    "No discernible step-up in Standing Repo Facility takeup versus surrounding non-turn days.",
    "Any SOFR firming coincides with HY OAS (BAMLH0A0HYM2) widening > 25 bps over the same window, re-attributing the move to credit and falsifying the funding/credit decoupling premise."
  ],
  "output_mode": "investment"
}