quantmodel — Improve Link GGC Enhanced Mid-Month corr vs AS (2nd pass)

Objective

Improve Link GGC Enhanced Mid-Month corr vs AS (2nd pass).

Notes: 2nd pass: OECD MEI edition gap-fill shipped; corr stays 0.907. Residual shared GGC-family vintage noise (err corr~0.84 vs monthly). day12=0.919 rejected (AS labels 15th). Findings: link-ggc-enhanced-mid-month-corr-t108.md

Approach

Reuses existing report research/findings/link-ggc-enhanced-mid-month-corr-t108.md — tables/metrics preserved verbatim via source_report pointer. Body below summarizes grounded evidence from that report and adjacent artifacts.

Files / code / data changed

Associated commits (inferred from log grep t108):

bccf83da research: t108 GGC mid-month corr 2nd pass (OECD DI gap-fill) (#93)
bccf83da research: t108 GGC mid-month corr 2nd pass (OECD DI gap-fill) (#93)
Source report: research/findings/link-ggc-enhanced-mid-month-corr-t108.md
Task-history entry: research/findings/task-history.json#t108

Results

Key table from source report (preserved):

variant	corr	CAGRΔ	verdict
baseline mid12 asof	0.907	+0.82%	—
ME-ffill mom (t42)	0.892	+1.32%	reject (t66 already)
inclusive DI (`>=`)	0.913	+1.18%	reject (Enhanced blog is `>`)
decision / edition lag-1	≤0.884	—	reject
prefer M-2 / first-print / modern-17	≤0.880	—	reject
lookback 6/9/15 mid periods	≤0.891	—	reject
same-day 252d / calendar-12m	~0.904–0.906	—	no gain
hybrid mid-on / ME-off mom	0.908	+0.76%	flat

Conclusions and metrics are in research/findings/link-ggc-enhanced-mid-month-corr-t108.md; see detail page for preserved tables/metrics.

Conclusions

2nd pass: OECD MEI edition gap-fill shipped; corr stays 0.907. Residual shared GGC-family vintage noise (err corr~0.84 vs monthly). day12=0.919 rejected (AS labels 15th). Findings: link-ggc-enhanced-mid-month-corr-t108.md

Problems / follow-ups

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