US-Labor-Market
Top drivers
⌁ mcp.call("adw-254") vADW-254-live-1.0 Is the US labor market weakening (unemployment level + recent trend)?
US-Labor-Market
Top drivers
⌁ mcp.call("adw-254") vADW-254-live-1.0 A macroeconomic scenario agent polls ADW-254 monthly and, when stress_score crosses above 55 on a rising trend (indicating both unemployment level has moved above the 3.5% baseline floor AND the 3-month trend is deteriorating), automatically shifts the downstream credit-model's macro-scenario weight from base to adverse — adjusting expected-loss provisioning inputs without manual model-governance intervention, provided methodology_version has not changed. The backtest shows the score has fallen from 50.0 to 22.5 (a 55% decline) over its observation window, placing the current reading well below the neutral midpoint; the agent's threshold of 55 is thus anchored to a historically meaningful stress boundary rather than an arbitrary number.
A bank CFO uses the stress_score, unemployment_rate, and change_3mo fields from ADW-254 in the monthly ALCO deck to support the allowance for credit-loss calibration discussion, replacing a manual pull of BLS headline unemployment that captures level only. The blended level-plus-momentum methodology means the CFO can show regulators a documented, reproducible composite that leads net charge-off deterioration by one to two quarters — versus headline unemployment, which typically coincides with or lags credit-quality deterioration — providing stronger supervisory justification for early reserve builds.
0.5*level-stress(vs 3.5% floor) + 0.5*3mo-trend-stress
Version ADW-254-live-1.0 · validated to beat a naive baseline · benchmark: Headline unemployment alone; this blends level + momentum