Sizing lock · Phase A3.7 final · Review & decide

Lock the crypto alpha sizing.

For each task, tap YES to approve my recommendation, NO to skip, or MODIFY and tell me a different number. Decisions auto-save. When done, hit Submit — I'll rewrite YAMLs, run pre-commit, commit clean diffs. Containers stay DOWN.

Subject BTC + ETH crypto alpha portfolios Decisions 26 tasks · ~3 min Containers stay DOWN Recommended BTC ≤60% / ETH ≤35% overlap

Yes

Modify

Pending

0 of 26 reviewed

Appendix A — Recommended operating point summary

Account	Ceiling	Total notional	CAGR	Max DD	Calmar
BTC	≤60% overlap	747.5%	+269%	-25.8%	10.4
ETH	≤35% overlap	304%	+28%	-22.3%	1.3
Combined 50/50	—	—	—	-25.7%	—

All numbers gross of paper-to-live Sharpe degradation (already applied 0.75× to f1). Combined max DD ≈ BTC's because BTC dominates dollar weights after compounding.

Appendix B — Why these numbers

Operator-set ceilings 9%/25%/40% dropped per your earlier instruction
Per-strategy random search across 12,000 samples per account, with 25% diversification cap
Equity sim: cost model (BTC 12% / ETH 10% APR funding, 0.06% fees, 0.05% slippage) + MAE-aware intra-trade DD
f1_follow_short returns × 0.75 to reflect bootstrap-CI lower-bound Sharpe
Operating point = Calmar peak per account (best CAGR per unit of max DD under realistic assumptions)
Diversification cap forced sma4_btc to its appropriate "workhorse" weight, not the c4-concentrated artefact of unconstrained search

Lock the crypto alpha sizing.

Appendix A — Recommended operating point summary

Appendix B — Why these numbers

Submit final decisions?