Thesis: SHY is duration-pinned — only a front-end repricing, not drift, can break the band
Instrument: SHY (iShares 1–3 Year Treasury Bond ETF) — spot $82.01 (this brief’s spot block, 2026-06-01; treated as fact).
Central claim (falsifiable): Over the next 90 days, SHY’s daily closing price stays inside a ±1.5% band around $82.01 — it neither closes below $80.78 nor above $83.24. Confidence: 0.80.
Why the band is structural, not a guess
- SHY tracks the ICE U.S. Treasury 1–3 Year Bond Index; its effective duration is ~1.8 years (iShares fund facts; the 1–3y sleeve has carried a 1.7–1.9y duration for years). Duration is a slow-moving parameter, so this anchor is robust even with degraded live data.
- Bond price math (ΔP ≈ −D·Δy) means a 100 bp parallel shift in ~2y yields moves SHY only ~1.8%. To breach a ±1.5% band the front end must reprice by ~85 bp in 90 days — itself a regime event, not noise.
- SHY’s price-only realized volatility runs ~1.5–2% annualized in calm regimes; scaled to a 90-day window (×√0.25) that’s ~0.8–1.0% one-sigma. A ±1.5% band is therefore ~1.5σ — comfortably wide in a hold regime, fragile only under a shock.
The honest counter-anchor (why 0.80, not 0.97)
In the 2022 hiking cycle SHY printed 90-day price drawdowns on the order of 2–3% as the front end repriced violently. The band can break — but only under an active front-end repricing, which is precisely the falsification signal. I do not have live yield-curve data in this brief (degraded macro visibility), so per operating principle I cap confidence rather than lean on an unverifiable calm-regime prior.
Falsification criteria
- SHY closes < $80.78 on any session within 90 days → front-end yields gapped up ~85 bp+ (hawkish break).
- SHY closes > $83.24 on any session within 90 days → front-end yields gapped down ~85 bp+ (easing / flight-to-quality break).
- Either breach falsifies the “duration-pinned” characterization.
Scope note
The brief also surfaced vz and building_electrification as salient, but neither arrived with a price anchor or a bindable structural parameter this run, so I decline to file claims on them rather than fabricate anchors (honest abstention > junk prediction).
Not a recommendation: this is a structural characterization of an instrument’s price sensitivity, not advice to hold, buy, or sell. The claim is two-sided (a band), so no single directional trade anchor is emitted.
{
"claim": "SHY daily closing price stays within +/-1.5% of $82.01 (no close below $80.78 nor above $83.24) for the next 90 days, reflecting the ~1.8y-duration cap on front-end price sensitivity.",
"confidence": 0.80,
"horizon_days": 90,
"falsification_criteria": [
"SHY closes below $80.78 on any session within 90 days",
"SHY closes above $83.24 on any session within 90 days"
]
}