How Crypto Indices Work: 4 Ways to Weight the Market

On 10 September 2026, total crypto market capitalization fell 4.27% in twenty-four hours. On the same day, bitcoin dominance rose to 58.57%.

So how much did "the crypto market" fall?

It depends on which index you ask. Rising dominance on a red day means bitcoin fell less than everything else. A market-cap-weighted index, where bitcoin carries most of the weight, would have looked relatively calm. An equal-weighted index, which gives a mid-cap altcoin the same vote as bitcoin, would have looked worse. An index with bitcoin removed entirely would have looked worse still.

None of those numbers would be wrong. Each one correctly answers a slightly different question. Knowing which question your index answers is most of what it takes to use one well.

A crypto index is a rules-based basket of digital assets that tracks how a defined part of the market performs. The rules decide which coins are included, how much weight each one carries, and how often the basket is updated. Those three choices, far more than the name or the number of coins, determine what an index actually measures.

This guide explains how crypto indices are built, why serious ones disagree, and how traders put them to work.

An Index Is a Rulebook Before It Is a Number

The word "index" means three different things, and most confusion around crypto indices starts there.

The first is the rulebook. This is a written methodology that decides which assets belong in a basket and how much weight each one gets. The Nasdaq CME Crypto Index, the CoinDesk 20, and CCi30 are rulebooks before they are anything else.

The second is the published series, the number you see on a chart. It is recalculated continuously and shows what a basket built to those rules has done over time. It is information, not an instrument.

The third is the product: a fund, an ETF, or a derivative that tracks the series and that you can actually hold or trade.

You can invest in the S&P 500 because hundreds of products track it. In crypto, that last layer is thin. Many of the best-designed indices have no fund attached. The ones that do often hold something slightly different from their benchmark, because custody and listing rules also shape what a fund can own.

So when you look at any crypto index, the useful question is not what level it is at today. It is what rules produced that number.

Three Decisions Behind Every Index

Strip away the branding and every index comes down to three choices.

Who gets in: Top 10 by market cap? Top 100? Only assets a regulated custodian can hold? Stablecoins in or out? Memecoins? Each filter is an editorial decision, and a technical methodology document is mostly a list of them.

How much each one gets: Market cap, equal weight, capped, square root, factor-based. This is the choice that shapes behavior more than any other, and we will spend most of this article on it.

How often it changes: Monthly, quarterly, or continuously. Plus the buffer rules that stop an asset hovering around rank 20 from entering and leaving every cycle.

Two indices with the same number of constituents can behave completely differently if they answer these three questions differently. That is why "top 20" tells you almost nothing about how a benchmark will move.

The Bitcoin Problem

Every crypto index designer runs into the same wall.

Bitcoin dominance has sat around 57 to 59% through 2026. Strip stablecoins out of the denominator, since they are dollars rather than risk assets, and Bitcoin's share of the risk-taking market was closer to 65% in early September.

Weight an index by market cap in a market shaped like that and you get a bitcoin tracker with decorations. The evidence is public. In its Form 10-Q for 31 March 2026 , the Bitwise 10 Crypto Index ETF reported bitcoin at 76.73% of the fund and bitcoin plus ether at 91.04%. Ten holdings, two of which are the portfolio.

That is not a flaw in anyone's product. It accurately reflects the market. But accurate and diversified are different goals, and each major index answers the bitcoin problem in its own way. There are four basic approaches.

Nasdaq CME Crypto Index: Gate the Universe

Nasdaq CME Crypto Index live chart on cryptoindex.ai

The Nasdaq Crypto Index launched in February 2021 and has since been rebranded as the Nasdaq CME Crypto Index through a partnership with CME Group. It is one of the most institutionally engineered benchmarks in the market, and its answer to the bitcoin problem is to leave the weights alone and be strict about who gets in.

Bitcoin and Ethereum are permanent constituents. The rest rotate from a set that has included Solana, Chainlink, Cardano, Stellar, Litecoin, Bitcoin Cash, and Polkadot. Weights follow free-float market capitalization with no caps.

The interesting part is the gating. An asset must trade on vetted Core Exchanges and be supported by vetted Core Custodians before it can qualify, and it must reach at least 0.5% of the full market capitalization of all eligible assets. The logic: any product tracking the index should sit on investment-grade infrastructure, and an asset no regulated institution can hold has no business in a benchmark that funds reference.

Notice what is missing from the name: a number. NCI has no fixed constituent count. It expands and contracts with however many assets clear the bar each quarter.

Its governance is among the strongest in the space. Nasdaq owns and administers the index, CF Benchmarks calculates it, and an oversight committee must approve any material methodology change. The trade-off is speed. A fast-growing asset can sit outside the index for months simply because custody support has not caught up yet.

CCi30: Square-Root Weighting

CCi30 is the academic answer, and one of the most elegant pieces of design in crypto indexing.

A team led by Igor Rivin, a professor of mathematics at Temple University, and economist Carlo Scevola built it in 2017, with a base value of 100 set at the start of 2015. Its thirty constituents are weighted by the square root of market capitalization.

Why square root? Market-cap weighting lets Bitcoin swallow the index. Equal weighting overcorrects and gives the thirtieth coin the same influence as bitcoin, which floods the index with illiquid noise. The square root sits between the two: it keeps the market's size ordering while shrinking the distance between the top and the bottom.

CCi30 adds a second idea. Market caps are computed as an exponentially weighted moving average rather than a single snapshot, which filters out flash spikes and makes the index harder to game on rebalance day. Constituents reconstitute quarterly; weights rebalance monthly.

CI100: Factor-Based Weighting

CI100 index live chat

CI100 , the flagship index published by Cryptoindex.ai, starts from a different observation: market capitalization is only one of the things that make an asset index-worthy, and in crypto it is often the least reliable.

A token can carry a large market cap that is little more than circulating supply multiplied by a thin order book. A conventional market-cap index treats that asset exactly like a deeply traded one. CI100 does not.

Its weights blend four factors:

Size at 40% keeps the index anchored to the actual shape of the market. Volume and liquidity together carry 50%, which is the deliberate part: an asset has to trade, not just exist, to earn weight. Low volatility at 10% works as a damper, small enough that CI100 stays a market index rather than a risk strategy, large enough to soften the effect of a single coin in a violent repricing. The same four factors also select the hundred constituents, so one consistent view of quality runs through the whole index.

The rules are public. Anyone can read the formula on the methodology page and check the number against it.

Two design choices are worth being clear about. CI100 prints every five minutes from live exchange prices with a base of 1,000, and its composition and weights refresh continuously rather than locking for a month or a quarter. That makes it an unusually current picture of the market. It also means it is a display index: an observation of market structure, not a fund, and not something a real portfolio could replicate trade for trade.

Zooming In: Ecosystem and Sector Indices

Broad indices tell you what the market did. They cannot tell you where the money went. For that, you need narrower baskets.

Ecosystem indices measure a chain's economy rather than its token. SOL's price and the Solana ecosystem are related, but they are not the same thing. When capital rotates into Solana applications, the ecosystem can outrun the base asset, and the reverse happens on the way down.

Cryptoindex.ai publishes two of these: CI-Solana tracks the most liquid projects in the Solana ecosystem plus a curated SOL core, weighted by market cap. CI-Base does the same for Coinbase's Base network. Put the two series side by side, and you get an answer neither chain's own token can give you: which ecosystem's applications the market is pricing more generously right now.

Sector indices do the same across themes rather than chains. AI tokens, DePIN, DeFi, real-world assets, perpetual DEX tokens, prediction markets. Each has its own index on the site, with equal weighting where one giant would otherwise swallow the basket and market-cap weighting where the sector's internal hierarchy is itself the information.

This is where indices become most useful for traders, because sector and ecosystem rotation drives most of crypto's relative performance.

How Traders Actually Use Indices

Four uses come up again and again.

Benchmarking your own results: If your portfolio rose 20% while a broad index rose 25%, you did not beat the market. You held beta and lost some of it. Without a benchmark, every green month feels like skill.

Reading the regime: Compare a broad index with an ex-bitcoin one. When they move together, the market is broadening. When the broad index holds up while the ex-bitcoin one sinks, capital is hiding in bitcoin. Pair that with bitcoin dominance and the altcoin season index, which has sat around 30 in recent readings, well below the 75 that defines an altseason, and you get a fast, data-based answer to "what kind of market is this."

Spotting rotation: Line up sector and ecosystem indices over the same window. Narratives show up in the spreads between them before they hit headlines.

Checking a narrative: Someone says "AI tokens are ripping." A sector index answers whether the group is moving or whether one name is carrying a story.

From Reading an Index to Building One

Every index above encodes someone else's view of the market. At some point, most active traders want to test their own.

That is what Cryptoindex Pro was built for. It is an index studio that runs on the same engine as the CI indices. You pick the assets, choose equal, market-cap, or custom weights, set a rebalance rule, and see how the basket would have moved through history, with total return and maximum drawdown alongside. Then you can compare it with bitcoin, ether, or CI100 to see whether the idea adds value, and publish it with your thesis attached if you want others to follow it.

Nothing is bought or sold. No wallet and no order. It is analysis, and you can build an index before creating an account.

Two cautions apply to any backtest, here or anywhere else. Historical simulations ignore trading costs, and someone who already knows how the story ended builds them. Treat them as a description of how a rule behaves, not a promise of what you would have earned.

[Image 4: Index studio on cryptoindex.pro | Alt text: Custom crypto index builder on cryptoindex.pro compared with bitcoin] Build a basket, set the weights, compare it with bitcoin. Hypothetical historical performance, analysis only.

Five Questions Before You Trust Any Index

Whatever index you use, run it through five checks. You can answer all of them from public documents in a few minutes.

1. Who runs it? An independent administrator, a fund issuer, an exchange, or a data site. Each has different incentives.

2. Who is allowed in? Read the eligibility rules. They tell you what the index will never hold.

3. How is it weighted? This decides behavior more than the constituent count ever will.

4. How often does it change? And what buffer rules stop it from churning?

5. How much of the history is real? Many indices show years of backtested data before their launch date on the same continuous line. Only the post-launch period is a record.

The Takeaway

Crypto does not have one index the way US equities have the S&P 500. It has a dozen serious benchmarks, each giving a well-reasoned answer to the same hard question: what to do about a market where one asset is most of the value.

Nasdaq gates the universe, and CCi30 compresses it. Meanwhile, CI100 blends size with liquidity and stability. Each is built to answer a different question, and the traders who get the most from them are the ones who know which question they are asking.

You can compare all of them side by side at cryptoindex.ai , and when you are ready to test your own view, build it at cryptoindex.pro .