← Results index

Improve Keller FAA corr vs AS

2026-07-23T19:08:13Z Strategy implementation done FAA corr 0
replication t116

Source report: research/findings/faa-corr-t116.md

Objective

Improve Keller FAA corr vs AS.

Notes: FAA corr 0.946→0.973 via ci=corr(asset,EW others); DBC→PDBC stitch also shipped. Live hist-alloc MAE~0. Findings: faa-corr-t116.md

Approach

Reuses existing report research/findings/faa-corr-t116.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 t116):

Results

Key table from source report (preserved):

variantmonthly corrrecent hist-alloc mean MAEnotes
t87 pairwise-matrix mean ci0.946~0.10baseline
DBC→PDBC stitch only0.947~0.10IPO gap not the residual
w_v=w_c=0.5 (tie hack)0.952~0.10rejected — AS docs 49.99%
me_excl vol window0.863worserejected
daily vol/corr0.912worserejected
ci = corr(asset, EW others)0.973~0shipped

Conclusions and metrics are in research/findings/faa-corr-t116.md; see detail page for preserved tables/metrics.

Conclusions

FAA corr 0.946→0.973 via ci=corr(asset,EW others); DBC→PDBC stitch also shipped. Live hist-alloc MAE~0. Findings: faa-corr-t116.md

Problems / follow-ups

Links

Source artifact: research/results/t116.md