Quantifying Concentration Risk: Wallet Distribution Analysis as a Due Diligence Tool

A token called SQUID was trading at about a penny on October 26, 2021. Five days later, it briefly touched $2,861 before collapsing to less than a tenth of a cent within minutes. Contemporary reports estimated that its developers extracted approximately $3.3 million from investors who discovered, too late, that the contract’s own code prevented them from selling.

The SQUID collapse demonstrates why wallet concentration cannot be reviewed in isolation. A token may appear distributed across many addresses while contract logic gives only selected parties the practical ability to exit.

That's the case for treating concentration risk as a number you calculate, not a feeling you have. This article is one piece of the broader risk-scoring framework used to evaluate crypto platforms before allocating capital, and it builds directly on the supply-side metrics covered in tokenomics forensics, where top-holder concentration is one of three scoring inputs. Here, we go deeper into exactly how to measure it: the Gini coefficient, the top-10-wallet threshold, and how both benchmark against documented collapse cases like SQUID.

How Concentrated Is Too Concentrated?

Ask ten people in crypto what counts as dangerous concentration, and you'll get ten different gut-feel answers. The point of this article is to replace the gut feeling with two specific, checkable numbers.

The first is a fast screening heuristic based on top holder concentration. The second is a statistical measure, the Gini coefficient, which summarizes inequality across the specific address population included in the calculation.

Used together, they give a due diligence process something to actually check against, instead of an impression of whether a project feels decentralized.

The Top-10-Wallet Threshold

This is the fastest concentration check available, and it requires nothing more than a token's Etherscan Holders tab, filtered to remove exchange wallets, liquidity pool contracts, and burn addresses, which was covered in detail in the tokenomics forensics piece.

While full chain-explorer data from the top 10 non-exchange, non-contract wallets provides a useful starting point, address-level data alone cannot account for entity clustering, beneficiary attribution, or cross-chain balances. The following bands are screening heuristics, not validated industry standards or regulatory thresholds. Their practical interpretation depends heavily on token design, project maturity, circulating supply, vesting schedules, liquidity, and governance structures.

The advantage of this metric is speed. The limitation is that it only looks at ten wallets. A token could pass a top-10 check cleanly while still having a dangerously unequal distribution further down the holder list, which is exactly the gap the Gini coefficient is built to close.

Classifying Addresses vs. Simple Removal

Rather than deleting non-wallet addresses from analysis, a rigorous methodology classifies them into functional categories:

A complete audit must distinguish between address-level concentration

($\text{Address-level concentration}$)

and entity-adjusted concentration

($\text{Entity-adjusted concentration}$),

while explicitly disclosing the percentage of supply residing in unidentified addresses.

The Gini Coefficient, Explained Properly

The Gini coefficient is a statistical measure of inequality, originally built for measuring wealth distribution across a population, and it has been adapted directly to measure token distribution across wallet addresses. It ranges from 0 to 1.

1. A Gini coefficient of 0 means every wallet holds an identical amount, perfect equality.
2. A Gini coefficient of 1 means a single wallet holds the entire supply, total concentration.

Academic literature on cryptocurrency wealth distribution defines the Gini coefficient using the Lorenz curve, a curve plotting the cumulative share of tokens held against the cumulative share of holders, ranked from smallest to largest. The Gini coefficient is calculated as the area between the Lorenz curve and the line of perfect equality, divided by the total area under that line of equality.

In plain terms: the further the actual distribution bends away from an even split, the higher the score.

To maintain mathematical precision, the Gini coefficient must be calculated across positive-balance addresses after separately classifying burn, exchange, bridge, liquidity pool, treasury, and vesting addresses.

A publishedNature Humanities and Social Sciences Communications study on Ethereum supply distribution applied Gini methodology to measure concentration at Ethereum's consensus layer, illustrating how address inclusion boundaries directly impact calculated scores. The Gini coefficient summarizes inequality across the included sample but does not automatically account for common off-chain ownership, custodial aggregation, or lost wallets.

Where Real Token Gini Numbers Actually Land

Numbers without a benchmark are hard to interpret, so it helps to know where established, liquid tokens actually score. According to Glassnode data cited in recent academic research, raw address-level metrics for Bitcoin recorded Gini values exceeding 0.92, while Ethereum registered values above 0.89 during specific observation windows (2020–2025). However, these values are highly sensitive to UTXO handling, exchange wallet inclusion, dust account filtering, and circulating supply definitions. They serve as context for unadjusted address distributions rather than absolute benchmarks.

That's a genuinely useful, if uncomfortable, benchmark. Even the two most established, most liquid, most widely distributed cryptocurrencies in existence carry Gini coefficients well above 0.89. This tells you that a Gini score in isolation, without context, can be misleading. A brand-new token with a Gini coefficient of 0.85 isn't automatically safer than Bitcoin's 0.92, and a token scoring above 0.92 isn't automatically an outlier disaster. What matters more is the trend over time and how the number compares to similar tokens at a similar stage, plus whether the top holders are known, locked, or vested versus anonymous and freely tradable.

A separate cross-cryptocurrency study notes that while some tokens demonstrate increased dispersion over time, concentration trajectories vary significantly by asset type, economic incentives, and measurement methodology. Direct comparisons between token Gini metrics and national wealth distribution statistics are methodologically flawed, as national wealth metrics measure individuals or households rather than raw blockchain addresses.

A related metric used in decentralization analysis is the Nakamoto coefficient. The Nakamoto coefficient measures the minimum number of independent entities required to reach a specified control threshold in a defined subsystem (such as consensus, block production, or governance voting power). For token governance, it represents the minimum number of distinct entities needed to pass or block a proposal, rather than a general price manipulation index.

SQUID Case Study, Due Diligence Sequence & Framework

The SQUID token collapse, independently documented by NBC News and CoinMarketCap, serves as a primary case study in transfer-control risks rather than simple holder concentration.

The core failure mechanism was contract logic that prevented retail holders from selling while allowing privileged addresses to liquidate positions. As detailed in CoinMarketCap's retrospective, the token dropped from $2,861.80 to $0.0007926 in under five minutes. SQUID highlights why wallet distribution metrics must always be evaluated alongside smart contract execution permissions.

The lesson for concentration analysis specifically: a token can look reasonably distributed by wallet count while still concentrating all effective control, the ability to sell, in a small number of addresses through contract logic rather than pure token balance. This is why concentration due diligence has to include checking the contract's actual transfer and sell functions, not just the raw Holders tab numbers. A clean top-10 or Gini reading on paper doesn't offset a contract that structurally prevents most holders from exiting.

Putting the Two Metrics Together

Neither metric replaces the other. The top-10 check is fast and catches the most obvious cases. The Gini coefficient is slower to calculate but captures concentration across the full holder base, and it's the number that reveals whether a token's supply is becoming more or less distributed over time, which matters more than any single snapshot.

Standard Due Diligence Workflow

1. Data Extraction & Snapshot Definition: Export holder balance data at a specific block height using node APIs or indexed chain data rather than reliance on basic explorer web views.
2. Wallet Classification & Clustering: Categorize addresses (burn, exchange, pool, vesting) and perform clustering heuristics (funding sources, synchronized transfers, shared multisig signers) to estimate entity-level ownership.
3. Metric Calculation: Calculate the top-10 entity share alongside the sample-defined Gini coefficient.
4. Multi-Category Differentiation: Evaluate three distinct concentration vectors:

• Ownership Concentration: Who holds the tokens.
• Governance Concentration: Voting weight and Nakamoto threshold.
• Liquidity Concentration: Top holder size relative to order book depth.

1. Trend & Snapshot Analysis: Compare monthly snapshots at consistent block times to evaluate 30-day and 90-day concentration changes.
2. Smart Contract Verification: Audit transfer restrictions, freeze rules, minting privileges, and sell permissions.

A token scoring poorly on multiple fronts, high top-10 share, high and rising Gini coefficient, and restrictive contract logic, is a fundamentally different risk than a token with a single elevated metric and an otherwise clean profile. This combined read feeds directly into both the overall risk-scoring framework and the position-sizing decisions built on top of it, since concentration is one of the clearest signals for how much capital a given allocation can reasonably absorb.

Comprehensive Risk Assessment Matrix

Override Red Flags

Regardless of favorable Gini or top-10 concentration scores, immediately escalate risk ratings if:

• Smart contract logic restricts standard transfer/sell operations.
• A single entity controls upgradeability keys or unmanaged minting functions.
• Top holding wallets were funded from a single primary source transaction.
• Holder distributions change drastically immediately before promotional events.

Bottom Line

Concentration risk is measurable, not a matter of impression. The top-10-wallet check gives a fast first read. The Gini coefficient gives a fuller picture across the entire holder base, benchmarked against real data showing even Bitcoin and Ethereum carry high concentration scores by design. And documented cases like SQUID show why concentration analysis has to extend into the contract's actual logic, not stop at the wallet balances alone.

Frequently Asked Questions

Can you calculate a token's Gini coefficient yourself, or does it require specialized software?

Export full holder balances at a specified block number using standard RPC nodes or analytics indexing platforms. For a sample of $n$ holders with balances $x_i$ and mean balance $\bar{x}$, the standard relative mean-difference Gini formula is:

$G = \frac{\sum_{i=1}^{n}\sum_{j=1}^{n}\vert{}x_i-x_j\vert{}}{2n^2\bar{x}}$

For ordered balances ($x_1 \le x_2 \le \dots \le x_n$), the computationally efficient formula is:

$G = \left( \frac{2\sum_{i=1}^{n}i x_i}{n\sum_{i=1}^{n}x_i} \right) - \frac{n+1}{n}$

For tokens with tens of thousands of holders, on-chain analytics platforms that already aggregate wallet balance data make the calculation faster, but the underlying math is fully transparent and independently verifiable either way.

Does high concentration always mean a token is more likely to be a scam?

No, and treating every high-concentration reading as a red flag on its own leads to false positives. A newly launched token with 60% held by a handful of wallets could mean venture investors and the founding team hold large, publicly disclosed, contractually locked allocations, which is a very different situation from the same percentage held by anonymous wallets with no lockup and full ability to sell today.

What matters is the identity and lockup status behind the concentration, not just the percentage itself. This is why the practical due-diligence sequence in this article pairs the concentration number with checking who the top holders actually are, not just how much they hold.

Why might two different concentration tools show different numbers for the same token?

The most common reason is wallet clustering, or the lack of it. Sophisticated holders and even scam operators sometimes split their actual position across many separate wallet addresses, which makes each individual wallet look small even though the same entity effectively controls all of them. Basic explorer-based concentration checks count each address independently and can understate true concentration as a result.

More advanced tools apply clustering heuristics, tracing funding sources and transaction patterns, to group wallets likely controlled by the same entity before calculating concentration, which is why a clustering-aware tool can show meaningfully higher effective concentration than a simple top-10 wallet count.

Disclaimer: This article is for educational purposes only and is not financial, legal, investment, or tax advice. Concentration metrics are one input among many and do not guarantee a token's future price behavior or safety. Review all primary on-chain data and seek professional advice before investing.