quantmodel — Improve Varadi GI corr vs AS

Objective

Improve Varadi GI corr vs AS.

Notes: Varadi GI confirm=6: corr 0.909→0.953 CAGRΔ +1.62%. Geek note length undisclosed; 3d was hist-alloc heuristic. Findings: varadi-gi-corr.md

Approach

Reuses existing report research/findings/varadi-gi-corr.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 t107):

973d3511 research: t107 Varadi GI confirm=6 (corr 0.953) (#92)
973d3511 research: t107 Varadi GI confirm=6 (corr 0.953) (#92)
Source report: research/findings/varadi-gi-corr.md
Task-history entry: research/findings/task-history.json#t107

Results

Key table from source report (preserved):

variant	corr	CAGRΔ	notes
confirm=1 (original)	0.864	+5.09%	high turnover
confirm=3 (t95)	0.909	+2.11%	hist-alloc day-of match 6/6
confirm=4	0.941	+1.98%
confirm=6	0.953	+1.62%	peak corr
confirm=8	0.945	+1.21%
confirm=10	0.917	−0.44%
TR→price index	0.875	+2.48%	worse
BestFolio EW+XLY+SMA	≤0.890	—	rejected

Conclusions and metrics are in research/findings/varadi-gi-corr.md; see detail page for preserved tables/metrics.

Conclusions

Varadi GI confirm=6: corr 0.909→0.953 CAGRΔ +1.62%. Geek note length undisclosed; 3d was hist-alloc heuristic. Findings: varadi-gi-corr.md

Problems / follow-ups

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