quantmodel — Audit Novell SPY-COMP 2002 unemployment residual

Objective

Audit Novell SPY-COMP 2002 unemployment residual.

Notes: Novell-only UNRATE >12m avg +0.10pp improves SPY-COMP corr 0.954→0.970 and DB 0.960→0.977; yearly SPY corr 0.946→0.971. Rejected require-rising despite 2002 fit because full-window return corr regressed.

Approach

Reuses existing report research/findings/novell-spy-comp-ue-t230.md — tables/metrics preserved verbatim via source_report pointer. Body below summarizes grounded evidence from that report and adjacent artifacts.

Files / code / data changed

No commit SHA confidently associated with t230 in sampled log [reconstructed — uncertain].
Source report: research/findings/novell-spy-comp-ue-t230.md
Task-history entry: research/findings/task-history.json#t230

Results

Key table from source report (preserved):

variant	SPY-COMP corr	Dynamic Bond corr	yearly SPY corr	yearly SPY MAE	2002 SPY
baseline (t179)	0.9537	0.9602	0.9459	3.63%	8.7%
UNRATE margin +0.10pp	0.9701	0.9767	0.9705	2.61%	25.8%
require UNRATE rising	0.9467	0.9531	0.9864	2.17%	59.1%
AS observed	—	—	—	—	58.1%

Conclusions and metrics are in research/findings/novell-spy-comp-ue-t230.md; see detail page for preserved tables/metrics.

Conclusions

Novell-only UNRATE >12m avg +0.10pp improves SPY-COMP corr 0.954→0.970 and DB 0.960→0.977; yearly SPY corr 0.946→0.971. Rejected require-rising despite 2002 fit because full-window return corr regressed.

Problems / follow-ups

See source report or PR discussion for follow-ups. No unsupported follow-ups fabricated.