quantmodel — Hard reverse-engineer Black Box Defense First vs AS

Objective

Hard reverse-engineer Black Box Defense First vs AS.

Notes: Black Box DF RE+replicated 2015+: corr 0.886 CAGRΔ -0.19%. Spec/port backtest/black_box_df.py. Findings: as-black-box-defense-first-re.md. t96 RE wave complete.

Approach

Reuses existing report research/findings/as-black-box-defense-first-re.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 t106):

5ba9ac18 research: t106 Black Box Defense First reverse-engineered (corr 0.886) (#91)
5ba9ac18 research: t106 Black Box Defense First reverse-engineered (corr 0.886) (#91)
Source report: research/findings/as-black-box-defense-first-re.md
Task-history entry: research/findings/task-history.json#t106

Results

Key table from source report (preserved):

AS date	notes
2026-01-30	GLD/PDBC/SPY weights ~exact
2025-07-31	GLD/PDBC/SPY weights ~exact
2025-08-29	GLD/SPY directionally right
2026-06-30	AS TLT/UUP/CASH vs ours TLT/UUP/SPY — SPY vs cash gate residual

Conclusions and metrics are in research/findings/as-black-box-defense-first-re.md; see detail page for preserved tables/metrics.

Conclusions

Black Box DF RE+replicated 2015+: corr 0.886 CAGRΔ -0.19%. Spec/port backtest/black_box_df.py. Findings: as-black-box-defense-first-re.md. t96 RE wave complete.

Problems / follow-ups

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