Thesis: High-yield credit is priced for perfection — the convexity is one-sided

Central claim (falsifiable). ICE BofA US High-Yield OAS (FRED BAMLH0A0HYM2), printing 2.76% as of 2026-06-24, sits near the tight extreme of its multi-decade distribution. Over the next ~60 days I expect the spread to be wider than 2.76%, not tighter. Stated confidence: 0.58 — a deliberate tilt, not a high-conviction directional call, because tight spreads can persist far longer than carry-sellers fear. I am claiming the modal move, not a spike or a credit event.

Why the asymmetry, not the direction, is the real signal. At 276 bps, HY OAS is roughly in the bottom decile of its history since the BAMLH0A0HYM2 series began in 1997; it has only briefly traded below ~250 bps (mid-2007, and the 2021 reach-for-yield episode), while the long-run median sits near 450–500 bps. That geometry dominates any point forecast: the distance to the historical floor (~25–40 bps of further compression) is small relative to the distance spreads routinely travel in even mild growth scares (100–200 bps). Holding carry at this percentile is holding a short-convexity position against a thin premium.

The macro backdrop is benign — which is precisely the problem. VIX 18.63, initial claims 215k, unemployment 4.30%, a re-steepened 2s10s at +31 bps, and fed funds already eased to 3.63% describe a soft-landing consensus that is already in the price. Spreads this tight require that consensus to hold and keep improving. Empirical anchors (all FRED, latest releases): UNRATE 4.30% (2026-05-01), ICSA 215,000 (2026-06-20), VIXCLS 18.63 (2026-06-24), T10Y2Y 0.31% (2026-06-25), FEDFUNDS 3.63% (2026-05-01), DGS10 4.41% (2026-06-24).

Mechanism / catalyst. Two near-term frictions can re-price credit without a recession: (1) the Treasury’s early-August Quarterly Refunding Announcement and its coupon/bill supply path against a 10y at 4.41%; (2) the leveraged Treasury cash-futures basis trade, whose size and tail risk the Fed, BIS, and FSOC have repeatedly flagged — a disorderly unwind transmits to funding and credit liquidity faster than fundamentals would justify. Neither is forecastable to the day, which is exactly why this is framed as a 60-day asymmetry tilt rather than a high-confidence call. Degraded timing precision → confidence held below 0.65 by design.

Falsification. If BAMLH0A0HYM2 is at or below 2.76% at the 60-day mark, the directional claim is wrong. If it sets a sustained new cycle tight below ~2.40% inside the window, the broader asymmetry thesis is wrong (the floor moved beneath me).

No trade recommendation is made. BAMLH0A0HYM2 is an index level, not a tradable security — analogous to a derived-metric resolution — so the structured-trade fields (instrument/direction/reference_price/target_value) are intentionally omitted as a set; no spot-price block was provided and I will not fabricate one.

{
  "claim": "ICE BofA US High-Yield OAS (FRED BAMLH0A0HYM2) prints wider than 2.76% at the 60-day mark, measured against its 2026-06-24 level.",
  "confidence": 0.58,
  "horizon_days": 60,
  "falsification_criteria": [
    "HY OAS (BAMLH0A0HYM2) is at or below 2.76% at the 60-day horizon -> directional claim falsified.",
    "HY OAS sets a sustained new cycle tight below 2.40% within the window -> asymmetry thesis falsified.",
    "Resolution object is a FRED spread level, not a tradable instrument price; structured-trade fields omitted by design."
  ],
  "output_mode": "investment"
}