Cyber / Captive Crypto Fund — Backtest Model

Arrington Capital · stylized four-bucket scoping model
Parameters

Bucket weights (targets)

Raw sum 100 → normalized to 100% for the portfolio.

How to read these weights

The four buckets are target value shares of the fund that sum to 100% — on a $100M fund the base case 10 / 20 / 20 / 50 is $10M / $20M / $20M / $50M. With quarterly bucket rebalance on, the fund resets to these targets each quarter; off, the dollar shares drift with each bucket's return. The collar protects only its own slice; cold and trend hold naked crypto and cash earns the basis-trading rate. It is a carve-out bucket, NOT an overlay across the book.

Live parameters

Structure

Sleeve composition

The 80/120 collar caps each roll period to ±20%: it blocks a fast crash but not a slow grind (~46% standalone drawdown vs ~76% for raw cold). The 20% target is a portfolio number, not the collar bucket alone.

Equity curve vs BTC buy-and-hold

FundBTC b&h

Per-bucket standalone returns

Each bucket's standalone return stream over the selected window (the inputs to the blend), plus its contribution = bucket weight × bucket total return. Cold ≈ the BTC-heavy sleeve buy-and-hold. Contributions are a first-order attribution and do not sum exactly to the fund (compounding + rebalancing cross-terms).

Methodology & disclosures

Trend-following trade audit

Trade-by-trade log of the trend bucket for the currently selected filter and exposure — every rotation between long crypto and cash, with the BTC-proxy trigger context and the Trend-vs-Cold edge each trade earned. Updates live with the trend filter and the Exp. high / Exp. low / cash-rate sliders.