Skip to content

Feat/strategy development#9

Open
enriquedlh97 wants to merge 3 commits into
benedictbrady:mainfrom
enriquedlh97:feat/strategy-development
Open

Feat/strategy development#9
enriquedlh97 wants to merge 3 commits into
benedictbrady:mainfrom
enriquedlh97:feat/strategy-development

Conversation

@enriquedlh97
Copy link
Copy Markdown

No description provided.

  - Add engineering standards (CLAUDE.md) and methodology doc (APPROACH.md)
  - Add Makefile with check, lint, test, validate, run, sweep targets
  - Add ruff linting and pytest config to pyproject.toml
  - Add static fee sweep script and baseline strategy template
  - Fix .gitignore: track strategies/, ignore .claude/ and .solc-select/
  - Commit uv.lock for reproducible builds

  Phase 1 results: static fee sweep (10-100 bps) shows 80 bps as
  optimal static fee (edge 380) vs normalizer baseline of 344 at 30 bps.
  Systematic search across 18 candidates in 4 strategy families using
  paired-seed evaluation (same seeds, per-seed deltas) found that a
  dual-EWMA volatility ratio (lambda=0.90/0.98) with ±3 bps regime
  adjustment beats static 80 bps by +27.92 edge points (t=127.91,
  100% win rate, 1000 sims).

  Key findings:
  - Static asymmetry: zero effect
  - Absolute EWMA vs nominal variance: badly miscalibrated (delta -45)
  - Directional adjustment: harmful (delta -1 to -5)
  - Dual-EWMA ratio: self-calibrating, +28 delta across all variantsv
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant