Onchain Index Tokens Test a New Structure for Crypto Exposure

Passive ETFs, replicating the performance of an underlying index mechanically by holding the same assets at the same weights, have become the default vehicle for institutional and retail participants alike, displacing active management as the preferred way to own the market versus cherry-picking single stocks. High-conviction, directional exposure demands genuine edge and skin in the game, and history shows most asset managers cannot sustain it in the long run.

Over the decade through 2025, just 21% of active funds survived and beat their average indexed peer, according to Morningstar data spanning nearly 9,248 funds and $26tn in assets. In US large-cap equity – the most institutionally significant and most analyzed market in the world – the same headline drops to 10%. The failure rate makes the cost disadvantage of active management not a headwind but a near-structural impossibility, a verdict that a decade of capital flows has since confirmed.

(Source: Investment Company Institute, FactBook 2026)

Passive US funds crossed active ones in total assets for the first time in 2024, and the gap has kept widening since, with passive vehicles holding $19.03tn against $17.57tn for actively managed funds and ETFs. Yet the latter headline is largely an assets’ appreciation residual, with active AUM expanding on the back of a decade-long equity bull market roughly tripling over the period. Strip out price appreciation and the honest ledger is flows, and from flows emerge a different story.

(Source: Morningstar)

Over the 10 years ending in 2025, US active funds have hemorrhaged $2.41tn in cumulative net flow according to Morningstar, while passive vehicles absorbed $6.43tn in fresh capital – an $8.84tn flow differential that represents the largest sustained reallocation in the history of modern asset management. Active funds managed to post net inflows in a single year out of ten – 2021’s post-COVID exception – coinciding with the broadest equity rally in a generation, when rising tides briefly made stock-picking look viable again.

That dynamic is now migrating onchain. Index products – the vehicle that displaced active management in traditional finance – are beginning to find their footing in Ethereum, bringing the same rules-based, basket-exposure logic into a composable, permissionless settlement layer. Onchain indexing refers to the creation of tokens or structured products whose value is mechanically linked to a basket of digital assets, with ownership, rebalancing and fee distribution handled entirely through smart contracts rather than through intermediaries or active managers. Although onchain indexing is still in its infancy, the topic has started to enter broader protocol design discussions. Ethereum co-founder Vitalik Buterin recently explored the construction of index-tracking assets using options-based mechanisms, pointing to growing interest in native onchain approaches to passive exposure.

If passive indexing won traditional finance by making broad market exposure cheap, transparent and structurally superior to discretionary selection, the question for onchain markets is whether the same logic holds, and whether the infrastructure now exists to deliver it at the same fidelity. Unlike traditional index funds or ETFs, which charge ongoing management fees and require separate custody arrangements, onchain index products can remove both the management layer and custody risk within a single smart contract structure.

Return of diversification 

Over the past six months, the equally weighted basket outperforms the median single asset pick across four of five sectors – by five percentage points in Layer-1, 12 in Layer-2, 9 in lending, and a decisive 42 in derivatives. The spread between the top and worst performers within any single sector ranges from 70 to 188 percentage points, meaning the cost of picking wrong is not a rounding error but the difference between +96% and -91% in the same basket.

(Source: Trading View, indices compiled by Sandmark)

The basket absorbs that dispersion and converts it into a positive return. That is the mechanical case for indexing in crypto: not that every constituent wins, but that no single bad pick is fatal. In a market where information asymmetry is high, volatility extreme and narrative cycles rotate faster than due diligence allows, removing single-asset selection risk is a genuine market edge.

Bringing indices onchain: Prism, Spectrum 

The passive revolution that rewired traditional finance over three decades is landing onchain – and the infrastructure building it looks nothing like what came before. A TradFi index fund charges management fees, requires custodians and separates the holder from the underlying. The onchain version collapses the entire stack into a single contract.

Earlier onchain indexing projects, such as Index Coop and Set Protocol, largely relied on wrapper contracts and external rebalancing mechanisms to construct baskets. Prism, and its associated launchpad Spectrum, both built on Uniswap V4, take a different approach. Prism is a token protocol that integrates liquidity provision and fee rights directly into the token itself, while Spectrum allows users to create custom thematic index tokens (baskets of digital assets) through a simple auction mechanism.

Prism and Spectrum take this flexibility a step further by collapsing the ERC-20 token, the underlying liquidity position, and the associated fee rights into one integrated structure. Holding a whole unit of Prism automatically mints an NFT representing one share of the protocol’s V4 liquidity position. No staking transaction, no wrapper contract, no gauge approval - the LP share is native to the token by construction.

The scarcity mechanic is where it gets interesting. The supply ceiling is 5,000 tokens and 5,000 NFTs, but the live count structurally trends downward. NFTs only exist for integer balances – sell past a whole number and one burns permanently. Fewer NFTs in circulation means each surviving holder commands a larger slice of the fee pool. Supply compresses, per-token yield expands, and the dynamic is irreversible. It is a deflationary flywheel with no manual intervention required.

Spectrum extends the model to index creation. Any user can launch a thematic basket – a set of digital assets bundled into a single tradeable token – via a Dutch auction that opens at 1 ETH and clears over ten blocks. The index token’s price is bound to the weighted value of its constituents by construction, not by arbitrage or active management. Every swap the index token ever generates splits three ways: 30% to the creator as a perpetual royalty, 60% to holders as passive yield, and 10% to buy and burn Prism. The creator monetizes conviction; the holder earns without managing; and every index that ever trades applies permanent buy pressure to Prism underneath.

The contrast with TradFi indexing is structural rather than incremental. A conventional ETF charges basis points annually and separates management from ownership. Here the relationship inverts – the index creator earns from the market’s appetite for their thesis, and the holder receives yield rather than paying for access to it.

State of the experiment 

The caveats are real and worth sitting with. The integer mechanic which makes the NFT dynamic and elegant also makes Prism difficult to compose. Any protocol that moves balances – a lending market taking collateral, a DEX routing a partial fill, a fee contract skimming basis points – risks crossing an integer boundary and burning an NFT the holder did not intend to burn. This is not a minor technical inconvenience. It creates meaningful friction for users and limits how Prism can be used as collateral, as a tradeable asset in external venues, or as a building block inside other protocols. Until that composability gap closes, adoption outside direct holding remains constrained.

Adoption remains limited for now. Across Ethereum and Base, Spectrum currently hosts roughly 12 indexes with around $83k in assets under management, with the largest product being the Base AI Index. While these figures are a rounding error by any standard, they suggest the model is beginning to attract early experimentation and provide an initial testbed for the economics behind onchain indexing.

But the potential roadmap is considerably larger than the product today. If the model proves resilient, future iterations could extend beyond single-chain crypto baskets to encompass cross-chain indexes and eventually tokenized equities, significantly expanding the universe of assets that can be packaged into a single onchain index.

What the experiment demonstrates, more than any specific metric, is that the primitives now exist to build index infrastructure onchain without the management layer that made TradFi indexing extractive. The fee model works. The scarcity mechanic works. The composability problem is unsolved, but it is an engineering constraint on a live system, not a fundamental objection to the thesis. For institutional and sophisticated participants, this model offers the possibility of broad market exposure without ongoing management fees or separate custody arrangements – a structural shift from how indexing has traditionally worked in both TradFi and earlier onchain attempts.