quantmodel — Idle recurring: hone least-accurate replications vs AS returns + yearly allocs

Objective

Idle recurring: hone least-accurate replications vs AS returns + yearly allocs.

Notes: Vitral MAM protection_b sweep: 11.5 corr 0.9536 vs default 10.5 corr 0.9522, but worse CAGR delta/MAE and conflicts with published formula; retained 10.5. Findings: vitral-mam-hone-t166.md

Approach

Reuses existing report research/findings/vitral-mam-hone-t166.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 t166 in sampled log [reconstructed — uncertain].
Source report: research/findings/vitral-mam-hone-t166.md
Task-history entry: research/findings/task-history.json#t166

Results

Key table from source report (preserved):

`protection_b`	monthly corr	CAGR delta	monthly MAE
8.5	0.9461	+0.73%	0.3695%
9.5	0.9496	+1.12%	0.3531%
10.5 (current)	0.9522	+1.41%	0.3583%
11.5	0.9536	+1.58%	0.3870%
12.5	0.9511	+1.67%	0.4286%

Conclusions and metrics are in research/findings/vitral-mam-hone-t166.md; see detail page for preserved tables/metrics.

Conclusions

Vitral MAM protection_b sweep: 11.5 corr 0.9536 vs default 10.5 corr 0.9522, but worse CAGR delta/MAE and conflicts with published formula; retained 10.5. Findings: vitral-mam-hone-t166.md

Problems / follow-ups

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