Skip to content

KB/signal

10Y-3M Treasury Spread

Last verified

Auto-generated. This article is rebuilt from app/signals/config/signal_definitions.json by scripts/build_signals_kb.py. Edit the registry entry and re-run the script β€” do not edit this file directly.

10Y-3M Treasury Spread

What it is

10Y-3M Treasury Spread β€” registry key t10y3m_spread.

NY-Fed near-term recession spread (10y minus 3-month Treasury; the academically-preferred curve-inversion measure vs. the existing 10y-2y); ingested + persisted, not yet scored. No band classifier in this PR.

Data-pyramid tier

T1 β€” Signal. This metric sits at the signal layer of the platform’s six-tier data pyramid (T0 raw inputs β†’ T5 actionable read). The tier reflects how far the value is from a raw measurement β€” not how strongly it is validated. Abstraction and validation are separate axes: a higher tier is not a claim of stronger evidence.

Horizon & validation

Horizon: 40-60d β€” the window this metric’s measured skill speaks to. A horizon stamp is an evidence claim, not a trading signal.

Evidence: Macro drawdown-lens leg: forward-maxdd correlation grows with horizon and peaks at 40-60d (breakeven_10y +0.28@20d -> +0.42@60d; ofr_fsi_funding ~+0.278, highest tested). FAILS clean walk-forward (single 2022 bear, in-sample; RT-3 falsifier open). Weak - ships labeled, never as a standalone forecast.

Source

Source module: credit
Data source: back_fillable

Fetched directly from app/sources/credit.py; see source code for the upstream API call and any provider-specific handling.

How it’s computed

10y-minus-3-month Treasury spread from FRED series T10Y3M, in percent. The NY-Fed-preferred near-term recession curve measure. Persisted raw with no band classifier. See app/sources/credit.py.

Where it surfaces

Health-score / alignment role

Data carrier β€” no implication, no health-score contribution.

Persisted for downstream consumers (sparklines, base-rate matcher, calibration substrate) but does not classify into BULLISH / NEUTRAL / BEARISH and does not contribute to the 0-100 health score.

Release cadence

See also