BASKET SIMULATOR · STRUCTURAL LAB PROXY
DO NOT FORCE 20. DO NOT RANK BY BTC.
No weighting method is selected because it beat BTC. Returns from size proxies would be circular and hindsight. Compare nEffective, concentration, turnover, and capacity — then freeze only after a vendor point-in-time free-float panel exists. Live DDEI is not published.
RESEARCH · SIMULATION · NOT INVESTMENT PERFORMANCE · NOT INVESTABLE · £0 REAL CAPITAL · PAPER
Headline size is a research ordinal/Zipf proxy with an era BTC-dominance overlay. It is not a vendor market-cap print and is not a forecast. · BEAT_BTC=FORBIDDEN · FROZEN=FALSE · LIVE=FALSE
LATEST SNAPSHOT 2025-12
Capped at 20 after qualification. Qualification ≠ weighting.
QUALIFIED names used: 20
| Method | N | N effective | Largest | Top 5 |
|---|---|---|---|---|
| MCAP | 20 | 2.65 | BTC 59.3% | 85.7% |
| CAPPED_MCAP | 20 | 14.16 | BTC 10.0% | 49.7% |
| EQUAL | 20 | 20.00 | BTC 5.0% | 25.0% |
| SQRT | 20 | 16.16 | BTC 10.0% | 44.0% |
| RISK | 20 | 19.59 | XLM 6.7% | 29.5% |
| LIQUIDITY | 20 | 13.92 | BTC 10.0% | 50.0% |
| HYBRID | 20 | 15.87 | BTC 10.0% | 45.1% |
MEAN TURNOVER (L1/2 ACROSS SNAPSHOTS)
- MCAP: 16.0% · mean N_eff 3.21
- CAPPED_MCAP: 24.5% · mean N_eff 13.64
- EQUAL: 25.5% · mean N_eff 18.67
- SQRT: 23.7% · mean N_eff 15.36
- RISK: 31.6% · mean N_eff 17.62
- LIQUIDITY: 22.8% · mean N_eff 13.42
- HYBRID: 23.7% · mean N_eff 15.11
WHY EACH METHOD EXISTS
- MCAP — Headline size proxy, uncapped. Concentration typically extreme. Not vendor market cap.
- CAPPED_MCAP — Headline size proxy with a 10% single-name cap (or 1/N if N is small). Research default cap, not frozen.
- EQUAL — 1/N among QUALIFIED names only. Ignores size. Capacity on small names is a known failure.
- SQRT — Square-root of headline size, then 10% cap. Diversifies vs raw mcap without going equal.
- RISK — Inverse liquidity-band proxy (not realised vol). Not a covariance model. RESEARCH PROXY.
- LIQUIDITY — Liquidity-band multiplier as size. Not vendor ADV. RESEARCH PROXY.
- HYBRID — 0.5 SQRT + 0.5 LIQUIDITY, renormalised, then 10% cap. Not optimised on returns.