CME CF Crypto Indices Alternative: CCi30 vs CME CF Crypto Indices

The count decides the case before any other criterion is examined. Twelve constituents and ten constituents both sit below the threshold at which a sample becomes a statistically significant estimator of its population. At n=12 the index is not measuring the cryptocurrency market; it is expressing a view about twelve assets. At n=10, less.

Thirty is the minimum count at which the top of the market represents the whole at a 99% confidence level with a 1.11% margin of error, and the maximum before liquidity deteriorates. CME landed at forty per cent of it.

The Ground Rules define the market capitalisation against which coverage is assessed as the sum of the market capitalisations of every digital asset for which a CF Benchmarks reference rate exists. The denominator is the administrator’s own product catalogue.

An index constructed this way cannot, as a matter of arithmetic, discover that it is missing anything. Whatever the sponsors decline to price is not underweighted; it is definitionally outside the market. Percentile coverage computed against such a denominator will always look complete, and will always be measuring the wrong population.

For years, providers excluded assets through vendor screens and called the result a methodology. CF Benchmarks states the mechanism plainly. An asset enters the investible universe only if at least one constituent custodial services provider publicly offers safekeeping for it, and only if it is listed for trading to users from major financial jurisdictions on two or more constituent exchanges. The custodian list runs to five US-regulated firms. The exchange list, as of August 2026, runs to seven venues.

Monero cannot satisfy either test, and no market capitalisation it ever achieves will change that. Neither can Zcash or Dash. The CME CF roster of single-asset crypto reference rates contains no privacy asset of any kind.

This is a coherent rule. It is simply not a rule about the cryptocurrency market. It is a rule about which cryptocurrencies American custodians and a short list of venues have decided to touch, which is a fact about compliance departments, not about money. An index whose universe is defined by permission measures permission.

Free-float weighting is imported from equities, where it is defensible because a shareholder register exists. Digital assets are bearer instruments and no such register exists. The Ground Rules concede the point and then proceed anyway, replacing the missing register with a chain of estimates.

For coin-centric assets, any coin that has not generated a UTXO output in the preceding seven years is deemed unavailable, and the document describes the seven-year cut-off as arbitrary and states it will need future revision. For account-centric assets, balances are discounted on a five-step schedule, by 20% to 100%, according to the fraction of supply a single account holds, with exemptions applied to accounts a blockchain-analytics vendor identifies as service providers. Where classification is unavailable, the free float is set to the median ratio observed across the other constituents. Any resulting change in free-float supply is then capped at 5% per month, with the remainder carried forward.

Each step is reasonable in isolation. Together they produce a weighting input that is not observed but constructed, from a heuristic, a discount table, a vendor’s labels, a median substitution and a smoothing cap, each with parameters chosen by the administrator. The output is presented as market capitalisation.

The CCi30 uses observable supply and applies one published function to it: an exponentially weighted moving average, then a square root. Two transformations, both disclosed, neither of which requires knowing something about ownership that no one can know.

CF Benchmarks circulated an indicative research portfolio for the emerging index on 13 August 2026: BNB at 32.0%, XRP at 27.6%, SOL at 18.5% and HYPE at 14.6%. Four assets, 92.7% of the index. The remaining six constituents shared 7.3% between them.

These were research estimates rather than the confirmed launch file, and should be re-verified against the live composition. But the figures follow from the methodology, not from a data error: free-float capitalisation weighting applied to a ten-asset universe produces a four-asset bet with six decorations. An investor who buys the bottom six is paying index-licensing economics for an average of 73 basis points of exposure each.

Add bitcoin and ether back, as the broad index does, and the ten emerging names collapse further still. The twelve-constituent index is a bitcoin-and-ether tracker carrying ten rounding errors.

The emerging index excludes bitcoin and ether by construction. This is not a diversification measure; it is the removal of the asset class’s monetary base from a benchmark for the asset class.

An index of crypto except crypto’s two settled monetary assets answers no allocator’s question. An institution that wants broad exposure cannot use it, because it omits most of the market. An institution that wants alt exposure already sizes that position against a whole-market benchmark, which this is not. What remains is a sector product, useful for constructing a spread trade against instruments the same exchange lists, which is a legitimate commercial purpose and an illegitimate claim to measurement.

Constituents are reviewed twice a year, on the first business day of June and December. In an asset class where the composition of the top ten changes materially inside two quarters, that means the index can carry a constituent list the market abandoned five months earlier, and can exclude an asset that entered the top ten in July until the following December. The CCi30 reconstitutes quarterly and rebalances weights monthly, precisely because a six-month lag in this market is not conservatism, it is error.

CF Benchmarks discloses its position without evasion, and the disclosure is the problem. CF Benchmarks belongs to the Crypto Facilities group, which belongs to the Payward group, which owns and operates the Kraken exchange. Kraken is a constituent exchange supplying input data to the indices, and Kraken Custody is one of the five custodians whose product decisions determine which assets are eligible in the first place.

The CME position is the mirror image. CME lists regulated futures on a number of the constituents, distributes the index data commercially, and co-brands the benchmark. The exchange that sells the instruments is a sponsor of the ruler by which the instruments are judged.

Neither firm has done anything improper. The structure simply places both sponsors on the field and in the officiating booth at once, which is the condition an independent benchmark exists to avoid.

The Ground Rules contain a constituent-selection route reserved for CME CF indices under which eligible assets are restricted to those covered by CME CF single-asset benchmarks and those with regulated futures or options listed on CME.

The two largest weights in the emerging index, BNB and HYPE, together 46.6% of the indicative portfolio, do not appear on the CME CF roster of single-asset crypto reference rates, and CME lists futures on neither. The roster published on 31 August 2026 covers AAVE, ALGO, APT, ARB, AVAX, AXS, BTC, BCH, TAO, ADA, LINK, CHZ, ATOM, CRV, MANA, ENA, ETH, FIL, HBAR, ICP, LTC, NEAR, ONDO, DOT, POL, SOL, XLM, SUI, SNX, XTZ, UNI and XRP. Ten of the twelve index constituents appear on it. BNB and HYPE do not.

Either that selection route does not govern these indices, in which case the published constituents and the published rulebook describe different instruments, or the roster is incomplete on launch day. For a firm whose entire proposition is regulated process integrity, a benchmark should not go live with its rulebook and its constituent list requiring reconciliation by the reader.

Both indices were launched on 31 August 2026. CF Benchmarks records an inception date of 1 June 2026 at a value of 1,000 and states that all values before that date are back-tested. Published history therefore begins on the launch date, with three months of pre-launch calculation behind it.

Any longer series presented for either index will be back-tested, constructed with the benefit of hindsight over a period whose survivors are already known. The CCi30 has computed values under rules fixed in advance since 1 January 2015, through two full bear markets, and carries no back-tested day at all.

Credit where it is earned. The Ground Rules exclude every pegged asset, stablecoins included, from the investible universe as a matter of rule. That is correct, and it is the founding principle of the CCi30, adopted here nine years later. The single-asset reference rates underneath these indices remain among the best price-fixing engineering in the asset class, and the regulatory pedigree behind them is earned.

None of that transfers. A rigorous price for twelve assets, selected from a custody-permitted universe and weighted by a modelled float, is a rigorous price for twelve assets.