← Back to blog

Crypto Portfolio Concentration Analysis: How to Measure It

November 22, 2026
Crypto Portfolio Concentration Analysis: How to Measure It

Portfolio concentration describes how much of a portfolio's value sits in a small number of positions, venues or price drivers. In crypto it has to be measured at three levels: the weight of each asset in your portfolio, the share of value on each exchange or wallet, and, for individual tokens, how supply and liquidity are distributed. RatioNexFlow is a crypto portfolio tracker in beta with rule-based risk flags. It imports positions from several exchanges and wallets into one read-only view and states the method behind each concentration figure. This article explains how those measures work. It does not tell you what to do with them, because that decision belongs to you.

RatioNexFlow
rationexflow.com
See Your Portfolio Concentration Clearly
RatioNexFlow brings positions into one read-only view and shows concentration alongside equity curve, drawdown and returns, with each method stated.
Open the demo

TL;DR

  • Concentration is measured from position weights: each holding as a percentage of total portfolio value, including stablecoins and assets left on exchanges.
  • The Herfindahl-Hirschman Index (HHI) squares each weight and adds them up. Dividing 1 by the HHI gives the effective number of positions, which is often much lower than the number of tickers.
  • Many crypto assets move together, so counting tokens understates concentration. Group holdings by price driver, venue and chain as well.
  • Token-level holder and liquidity concentration depend on how they are measured. The same ledger can give very different figures under different specifications.
  • RatioNexFlow flags only show whether its own rules' conditions are met in the data it reads. They are not advice, forecasts or a verdict on any token.

Table of Contents

What portfolio concentration means in crypto

A concentrated portfolio puts most of its value in a few assets, so the result depends heavily on those assets. A diversified portfolio spreads value across holdings that behave differently from one another. The two terms describe a spectrum, not a switch, and where a portfolio sits on it is a personal decision tied to conviction, research and tolerance for drawdown (the fall from a portfolio's previous peak to a later low).

Crypto complicates the picture because the number of holdings and the number of independent bets are different things. An analysis by Digital Estate Group covering Bitcoin, Ether and 37 altcoins through May 2026 found Ether-Bitcoin correlation of 0.81 to 0.86, an effective number of positions of about 1.8, and an average of 96% of altcoins falling on days Bitcoin fell more than 5%. Correlation is a measure from -1 to 1 of how closely two prices move together. A high figure means two holdings add less variety than their separate tickers suggest.

For that reason, a useful concentration analysis looks at weights first and then asks how many distinct price drivers sit behind them.

Step 1: Build one complete list of holdings

Concentration figures are only as good as the list they are computed from. Holdings spread across several exchanges and wallets are easy to undercount, and each omission changes every weight.

  1. List every asset in every place it is kept: each exchange account, each wallet, and any positions entered by hand.
  2. Value each at the same moment in one currency, for example USD.
  3. Include stablecoins (tokens designed to track a currency such as the dollar) and staked tokens. Leaving them out inflates the weight of everything else.
  4. Add the values to get the total, then divide each holding by that total.

RatioNexFlow supports this with a read-only API key per exchange (a key that can view balances but cannot trade or withdraw), a statement file import, or manual entry of trades. Exchange connections are in beta. RatioNexFlow keeps no custody of funds and has no trading or withdrawal rights. A per-exchange page, such as the Kraken portfolio tracker page, describes how a given connection works.

Step 2: Calculate the weights and the largest positions

The simplest measures read directly from the weight list:

  • Largest position: the single biggest weight.
  • Top-3 share: the combined weight of the three largest holdings.
  • Stablecoin share: the portion of value held in tokens that track a currency.
  • Share of the two largest assets: useful because many other tokens tend to follow them.

There is no official threshold that separates concentrated from diversified. Published rules of thumb differ between sources and between investors, and they describe preferences, not facts about your portfolio. A more useful use of these measures is comparison with your own earlier readings: how the largest position and top-3 share change as prices move and as you add or remove positions.

Prices alone change weights. A holding that doubles while the rest stay flat becomes a larger part of the total without any transaction, so concentration is worth measuring repeatedly, not once.

Step 3: Calculate the HHI and the effective number of positions

The Herfindahl-Hirschman Index is a standard concentration measure. Express each weight as a fraction of 1, square it, and add the squares. The result runs from near 0 (many equal positions) to 1 (everything in one position).

Portfolio weights combine into a concentration measure

The effective number of positions is 1 divided by the HHI. It answers the question: how many equally sized holdings would produce the same concentration?

A worked example with five hypothetical assets (A to E), as an illustration of the arithmetic only:

AssetWeightWeight squared
A0.400.1600
B0.300.0900
C0.150.0225
D0.080.0064
E0.070.0049
Total1.000.2838

HHI is 0.2838, so the effective number of positions is 1 / 0.2838, about 3.5. Five assets behave, in weight terms, like three and a half equal ones. Ten assets with 60% in one of them would show a similar result. The count of tickers is not the measure; the distribution of weights is.

The method has a stated limit: it uses weights only. It does not know that two assets are correlated, so it can understate concentration when holdings share a price driver.

Step 4: Group by price driver, venue and chain

After the weight-based figures, group the same list in other ways and recompute the shares:

  • By price driver: assets that depend on the same underlying network or sector, or that track the largest assets closely, can be grouped together. This is a judgment you make and state, which is why the grouping should be written down.
  • By venue: the share of total value on each exchange or in each wallet. This is a custody concentration measure: it shows how much depends on one platform's availability and operation, not how the assets are priced.
  • By chain: the share of value on each blockchain network.

Each grouping answers a different question, so each produces a different figure. A portfolio can look spread out by asset and still have most of its value on one exchange. This is the case for a combined view across venues, since no single exchange's balance page shows it.

Token-level holder concentration

The second kind of concentration concerns a single token: how its supply is distributed among holder addresses. Common measures include:

  • Top-N share: the percentage of supply held by the N largest addresses.
  • Gini coefficient: a 0 to 1 measure of inequality across all balances, sensitive to many near-empty addresses.
  • Nakamoto coefficient: the smallest number of holders whose combined balance exceeds a chosen share of supply.
  • HHI on balances: the same index as above, applied to holder balances, and most sensitive to the very largest.

These measures have a limit that matters in practice: an address is not a person. An exchange address can represent a great many customers, and one operator can use many addresses. Liquidity pool and contract addresses hold tokens for functional reasons. Whether such addresses are excluded changes the result.

An academic paper by Liu, Ji and Lin shows how much. Four discretionary measurement choices moved one token's balance HHI from 109 to 2,336, a factor of 21, and each choice was defensible. The implication is that a holder concentration figure is a property of its method, not only of the ledger. A figure should therefore be read together with its parameters: the block height (the point in the chain's history at which balances were read), the smallest balance counted, the denominator used for supply, and which addresses were excluded.

Liquidity concentration

Liquidity is how easily a token can be exchanged without moving its price much. On decentralized exchanges (DEXs), it sits in pools: pairs of tokens deposited by liquidity providers. Concentration appears in two forms.

Across pools. An OECD working paper using Uniswap v2, v3 and Curve data from December 2020 to August 2023 found that 20% of pools on some DEXs account for over 90% of that DEX's volume, and in about half of the sample a single liquidity provider controls 90 to 100% of a pool's liquidity. The authors read this as growing concentration of trading power.

Within a pool. In pools where liquidity providers choose a price range, the total value locked (TVL, the dollar value of assets in the pool) is an incomplete depth measure. Researchers Zhu and Moallemi separate depth into a TVL channel and a concentration channel around the mid-price, because the same TVL can offer different depth depending on where it is placed.

For a portfolio, liquidity concentration is relevant to position size: a holding that is large relative to the token's pool depth and daily volume is measured differently from one that is small relative to it. That comparison is a measure, and what to do about it is your decision.

Liquidity concentration

Reading concentration with drawdown and returns

Concentration describes the structure of a portfolio at a point in time. Drawdown and returns describe what happened. Reading them together shows how the structure and the history relate:

  • A peak-to-trough drawdown on the equity curve (the portfolio's value over time) shows how far the portfolio fell in a given period.
  • Concentration at the start of that period shows how the weights were arranged when it fell.
  • Returns by position show which holdings contributed to the result.

None of these predicts the future. Correlations in particular change: the Digital Estate Group analysis cited above found that altcoins fell together with Bitcoin on large down days, which is why a concentration reading from a calm period may not describe a stressed one. The portfolio risk analytics guide covers drawdown, returns and concentration methods in more detail, and the token metrics guide explains the token-level data points.

How RatioNexFlow presents these measures

RatioNexFlow's Portfolio Analytics covers the equity curve, drawdown, returns and concentration, each with its method stated. Positions from connected exchanges, imported statements and manual trades appear in one view, so the weight list in Step 1 is built from the data you provide.

Threat Protection (beta) is separate. It runs 17 rules on token contract, liquidity, pool, volume and holder data read from GoPlus, DexScreener and CoinGecko, and shows only whether each rule's conditions are met in that data. A flag is therefore a statement about data at the time it was read: it is not a verdict on a token, a forecast, or a statement that anything is safe or unsafe. RatioNexFlow does not predict, and does not guarantee to detect, any hack, exploit, rug pull or scam. Data gaps exist in any scanner, since the underlying sources cover some pool types and chains more fully than others. The risk disclosures set out the limits in full.

Price and percent-change alerts use levels you set. Email and push alerts are planned, not available. "Ask your portfolio" answers plain-language questions about your own holdings with AI. RatioNexFlow gives no advice, trade ideas or price forecasts, and RatioNexF OÜ is not a crypto-asset service provider and is not authorised or supervised as one.

Our perspective on measuring concentration

We build RatioNexFlow around one idea: a measure is useful when you can see how it was produced. A concentration figure without its method, such as the denominator, the treatment of stablecoins or the grouping used, invites the reader to trust a number they cannot check. For that reason each Portfolio Analytics measure states its method, and each Threat Protection rule states the conditions it tests.

We also keep the tool read-only. It has no custody of funds and no trading or withdrawal rights, so a view across exchanges and wallets does not require giving anyone control of assets. A combined view is what makes venue and asset concentration visible in the first place.

What the figures mean for a particular portfolio is a decision for its owner. We measure and explain, and we leave the conclusion to you.

Magnus Christiansen

One read-only view of the whole portfolio

The Free plan has no time limit and needs no card. Pro costs €29 a month or €290 a year, with the first 30 days free. Pro adds enhanced analytics, exchange syncing, wallet tracking and unlimited alerts. It also exports a CSV of closed trades and reward income, which is a file of records and not a tax report. Current terms are on the pricing page, and you can start from the RatioNexFlow homepage.

RatioNexFlow

FAQ

How do I measure crypto portfolio concentration?

List every holding, including stablecoins, as a percentage of total value. Then read the largest position, the top-3 share and the HHI, which is the sum of squared weights. One divided by the HHI gives the effective number of positions. Repeat the calculation grouped by exchange, wallet and chain.

Is owning many tokens the same as being diversified?

Not by the measures above. Weights and correlation both matter. Equal weights across many tokens give a low HHI, but if the tokens tend to move together, the number of independent price drivers is smaller than the number of tickers.

What does token holder concentration show, and what does it miss?

It shows how supply is split across addresses, usually as a top-N share, Gini coefficient or HHI. It cannot show who owns the addresses, and exchange, pool and contract addresses may be included or excluded depending on the method. The same data can give figures that differ widely, so read the stated parameters.

Does RatioNexFlow tell me whether a token is risky?

No. Threat Protection (beta) shows whether each of its 17 rules' conditions are met in the data read from GoPlus, DexScreener and CoinGecko. It gives no verdict, advice or forecast, and any decision stays with you.

Sources

Created with BabyLoveGrowth technology

This article was written with AI assistance and is general information only. It is not investment, tax or legal advice. RatioNexFlow measures and explains; it never advises, and RatioNexF OÜ is not a crypto-asset service provider. Crypto-assets are volatile and you can lose money.