US-Labor
Top drivers
⌁ mcp.call("adw-348") vADW-348-live-1.0 Is the US labor market actively deteriorating (claims/unemployment up, quits/hires down)?
US-Labor
Top drivers
⌁ mcp.call("adw-348") vADW-348-live-1.0 A macro-strategy portfolio agent monitors ADW-348 weekly and, when the cooling signal rises above 60 on a rising trend (700-observation backtest since 1968: range 0–100, mean 43.7; current 49.3 at 64th percentile and rising), it automatically reduces equity beta in the model portfolio and increases allocation to investment-grade bonds, logging the IOM's top_drivers—which decompose the signal across unemployment rate momentum (UNRATE), initial claims (ICSA), quits rate (JTSQUR), and hires rate (JTSHIR)—to explain which labor-market deterioration pathway is driving the shift. Unlike ADW-311, which measures labor-market level, ADW-348's momentum framing means the agent detects directional turning points even when the absolute tightness level remains high.
An HR analytics lead at a large employer uses ADW-348 to time voluntary-separation management decisions: a rising cooling signal warns that quits are falling and unemployment claims are rising, meaning workers are less likely to leave voluntarily—so offering voluntary early-retirement packages now, while cooling is still below 60, will cost less than waiting until the labor market deteriorates further and workers have fewer outside options. This is a concrete, quantified improvement over relying on quarterly BLS press releases, which combine these four series into narrative summaries rather than a single directional composite.
per-series momentum vs baseline -> [0,1] sub-signals -> weighted *100
Version ADW-348-live-1.0 · validated to beat a naive baseline · benchmark: Bloomberg LCI; GS CAI labor; Indeed Hiring Lab