Why PlainIndex doesn't score crypto ETPs
Spot bitcoin and ether ETPs are outside the catalog. The scoring model was built for diversified funds, and a one-asset trust leaves almost nothing to measure except the fee.
Not an allocation page
Nothing here is an argument for or against holding crypto. Same as gold, same as small-cap value: the site scores how a fund implements an exposure, not whether you want that exposure. People still ask about IBIT and FBTC, so this page is the reason they are not in the catalog.
What a spot-crypto ETP is, structurally
IBIT, FBTC, GBTC, and ETHA are grantor trusts, not 1940-Act investment companies. Each trust holds exactly one asset — bitcoin or ether in custodial cold storage — and issues shares against it. There is no portfolio manager, no index of securities, no rebalancing, and no income: the underlying asset pays no dividends or interest, so the trusts make no distributions. The sponsor fee accrues by selling small amounts of the underlying over time, which is why the asset-per-share figure declines slowly even when the price doesn't move.
Structurally these are siblings of the physical gold trusts (GLD, IAU) rather than of index funds — single non-productive asset, grantor-trust wrapper, fee paid in kind. The gold trusts make it into the catalog because gold has a long-standing slot in published lazy-portfolio allocations (Permanent Portfolio, Golden Butterfly) that the site documents. Crypto ETPs don't have that anchor, and as the next section shows, the scoring framework has even less to say about them.
Why the scoring axes collapse
The PlainIndex composite weighs four sub-scores: cost, tax efficiency, liquidity, and concentration. It was built for diversified index funds, and a single-asset trust breaks three of the four:
- Concentration is definitionally 100%. The sub-score measures top-10 holdings as a share of net assets to surface diversification differences within a category. A trust holding one asset has no diversification to measure — every fund would pin to the floor, telling you nothing.
- Tax efficiency has no distribution profile to score. The trusts distribute nothing, so the distribution-rate component is vacuously perfect, and the asset-class base in our table doesn't have a researched entry for grantor-trust crypto (the IRS treats the trusts as pass-throughs; gains are property gains, and the in-kind fee accrual creates small reportable sales each year that brokers handle inconsistently). Scoring this with a made-up base would be exactly the kind of false precision the methodology page exists to prevent.
- Holdings analysis — the layer that powers the overlap tool, the similar-funds links, and the compare pages — has nothing to work with. There is no N-PORT filing and no holdings table; "overlap" between two bitcoin trusts is 100% by construction.
- Cost and liquidity survive — but with only those two axes live, the composite stops being a multi-dimensional score and degenerates into a fee comparison with an AUM tiebreaker. We can publish that comparison honestly as a table (below) without dressing it up as a 0–100 score it isn't.
The one fact the methodology can speak to: fees
Because every spot-bitcoin trust holds the same asset, the funds are near-perfect substitutes for one another, and the sponsor fee is the main structural differentiator. The spread is unusually wide:
| Ticker | Trust | Underlying | Sponsor fee | AUM (June 2026) |
|---|---|---|---|---|
| IBIT | iShares Bitcoin Trust (BlackRock) | Bitcoin | 0.25% | ≈ $49B |
| FBTC | Fidelity Wise Origin Bitcoin Fund (Fidelity) | Bitcoin | 0.25% | ≈ $16B |
| GBTC | Grayscale Bitcoin Trust (Grayscale) | Bitcoin | 1.50% | ≈ $16B |
| ETHA | iShares Ethereum Trust (BlackRock) | Ether | 0.25% | ≈ $5B |
AUM figures are rounded and move with the underlying asset's price — treat them as order-of-magnitude indicators as of June 2026, not live data. Fees are the prospectus sponsor fee.
The notable row is GBTC: at 1.50%, it charges six times the fee of IBIT and FBTC for the identical underlying asset — a legacy of its pre-2024 closed-end-fund structure, where conversion costs and embedded capital gains have kept long-time holders in place despite the spread. Holders weighing that fee against their own embedded gains are making a tax decision specific to their situation, which is outside what this site models.
If the data gets better
A score would need something that actually differs across these trusts: custody, creation/redemption, premium and discount. We do not have a public source for that that is clean enough to score. Until we do, the fee table is the analysis. The methodology page is what the composite does measure, and the guides cover the diversified categories the model was built for.