DEXONOMICS

How the Dexonomics on-chain health index is calculated

Every token page on this site carries a number from 0 to 100 and a grade from A to F. This page is the full recipe: every signal, every threshold, every penalty. If you disagree with a weighting you can see exactly which one it is, and you can recompute any score yourself from the figures we publish on the page.

What this index is not.

It is not an audit. We do not read source code, we do not verify teams and we do not certify that anything is safe. It is a mechanical reading of what the chain shows, and a token can score well and still go to zero the same afternoon. Nothing here is financial advice.

The starting point

Every pair starts at 50, a deliberately neutral score, and moves from there. A pair we know nothing about stays near the middle rather than being flattered or punished by default.

1. Liquidity depth

The single most useful anti-rug number there is. Liquidity determines how much you can sell before you move the price against yourself, which is the difference between a position and a trap.

Pool liquidityEffect on score
$5,000,000 or more+22
$500,000 to $5M+17
$100,000 to $500k+12
$25,000 to $100k+7
$5,000 to $25k+2
Under $5,000-20 and a flag

2. Volume measured against liquidity

Volume on its own says nothing, because volume is the easiest number on a chain to manufacture. What carries information is volume relative to the pool that produced it. A pool with $255,000 of liquidity reporting $599 million of daily volume is not a busy market, it is a number being made up, and we have seen exactly that.

24h volume vs liquidityEffectReading
1x to 20x+12Healthy turnover
Above 20x-8Possible wash trading
Above 100x-20Strong wash trading signal
Above 500x-35Not physically plausible in real trading
Some volume, under 1x+3Quiet but alive
No volume at all-5Dead pool

3. The balance between buys and sells

Applied once a pair has at least 20 transactions, because below that the sample means nothing. If people are buying and nobody is selling, that is the classic honeypot shape: the contract lets money in and not out.

PatternEffect
More than 10 buys and zero sells-25 and a flag
Sells between 20% and 80% of trades+8
Skewed in either direction+2

4. Age of the pair

Surviving the first hours already filters most rug pulls, so age is evidence rather than nostalgia.

Pool ageEffect
More than 180 days+10
30 to 180 days+7
7 to 30 days+4
1 to 7 days+2
Under 1 hourFlagged as unproven

5. What the deployer wallet has done before

Where the launch platform exposes it, we look at the wallet that created the token and at what else it has launched. A wallet on its tenth token where eight collapsed after pumping is telling you something no price chart will.

Deployer historyEffect
10 or more prior launches with 80% collapsed-20 and a flag
5 or more prior launches-8
First token from this wallet+5

6. Whether the project exists off-chain

A token with a site, a published account and a written description is not safer, but it is less anonymous, and anonymity is a cost the buyer pays if something goes wrong.

Public presenceEffect
A description plus two or more official channels+6
No site, no accounts, no description-5
Flagged as banned by its own launch platform-30

The ceiling nobody can pass without being checked

A pair we have not been able to verify off-chain, meaning we have no deployer history, no contract check and no published description, cannot score above 82 no matter how deep its pool is. Without that ceiling every liquid pair came out at 98 and the number stopped saying anything. A high score has to mean something was verified, not merely that a lot of money is sitting in a pool.

Two things that disqualify a pair outright

These are not penalties, they are gates. A pair that fails either one is served but kept out of our index pages entirely, because the figures we would be publishing are not figures we trust.

From number to grade

GradeScoreWhat it means
A88 to 100No mechanical red flags and verified off-chain
B74 to 87Solid on the measurable signals
C58 to 73Mixed, worth reading the individual signals
D42 to 57Several signals against it
FBelow 42Serious problems in the basic numbers

Where the data comes from

Prices, liquidity, volume, transaction counts and pair ages are read from decentralised exchange pools through the CoinGecko on-chain API. Launch details and deployer history for tokens created on Solana bonding curves come from the pump.fun public API. Contract checks shown on token pages are produced by GoPlus Security and attributed to them, because those are their checks and not ours. Contract and pool addresses on every token page link straight to the block explorer for the chain, so nothing here has to be taken on trust.

How often it is recalculated

Price, liquidity, volume and trade counts refresh continuously. The slower parts of a page, meaning the description, the chart history, the contract checks and the market list, are rebuilt in the background on a rolling basis, with the busiest pairs first. Every page states the figures it was built from, so you are never reading a score without also seeing its inputs.

If you think a score is wrong

Every input is on the page next to it. If a pair looks mispriced by our index, the signal breakdown will tell you which rule produced it, and the tables above tell you what that rule is worth. That is the point of publishing this: a proprietary number that nobody can audit is just an opinion with a decimal point.

Start reading

All Solana tokensEvery pair we track on Solana, each with its index and full signal breakdownAll Ethereum tokensEvery pair we track on Ethereum, each with its index and full signal breakdownAll Base tokensEvery pair we track on Base, each with its index and full signal breakdownAll BNB Chain tokensEvery pair we track on BNB Chain, each with its index and full signal breakdown Tokenized stocksTokens using the ticker of a listed company, and how to tell them apart