Telling produced volume from spontaneous volume

Some Solana turnover is generated deliberately and some of it arrives on its own. Both look like trades in a block. The difference is not in any single transaction, it is in the shape of the whole set.

This desk describes those shapes, states what each one does not prove, and attaches a confidence level to every reading. No token, wallet or team is named as manipulated here, because a transaction record cannot support that claim.

Explore volume tooling Start with the head note

What a transaction record actually exposes

Five fields do most of the work in every reading on this site.

Signer: Which account authorised the swap, and whether one account produced many trades or many accounts produced one each.
Slot / time: When the transaction landed, to the granularity the cluster records, which is what makes cadence readable at all.
Balance delta: The token amounts that actually moved in and out of each account, rather than the amount an interface displayed.
Program: Which venue program executed the swap, and whether a single user action was routed into several legs.
Fee payer: Who paid for the transaction, which on Solana need not be the account whose tokens moved.

The five signatures

None of these is proof on its own, and each has a boring explanation that fits it just as well. They earn their place because they are cheap to check and because they point in different directions when the flow is genuinely mixed.

Cadence

Gaps between trades that cluster around a repeating interval. A crowd is irregular; a schedule is not. Retry logic and recurring buy tools produce the same look.

Size

Trade amounts drawn from a narrow set of values, or from a tight band around one figure. Interface presets and fixed budgets do this too.

Round trips

A buy and a matching sell that return an account's inventory to roughly where it began. Arbitrage and hedging leave the same trace.

Counterparties

A funding graph that collapses to a small number of sources once you trace where the trading accounts got their SOL. Exchange withdrawals collapse the same way.

Depth response

Turnover rising while pooled depth stays flat. It means the pair traded a lot without gaining any capacity to absorb a larger order.

Featured investigations

Four case files that carry the rest of the site. Each one states the question it answers, the evidence it rests on, the observation that would overturn it, and how confident the reading is.

Organic vs artificial volume

Why the question is about the distribution of a set of trades rather than about any one trade, the six checks the desk runs, and the cost floor that makes produced turnover visible in the first place.

Open the head note

Timing and size signatures

Inter-arrival gaps and trade sizes carry more information than a turnover figure ever does. How to read spacing and quantisation, and the innocent explanation attached to each.

Read the timing file

Wallet clustering basics

Clustering joins addresses that behave as if one operator drives them. Funding edges, fee-payer overlap and timing co-movement, with the false positive that breaks each heuristic.

Read the clustering file

Market making or wash trading

Two activities that share a footprint and differ in intent, which is the one thing a chain does not record. Where the evidence leans, and where it genuinely cannot decide.

Read the grey-zone file

Three sections, one standard

Signatures describe what the patterns look like. Evidence covers how the data is gathered so a second analyst can reproduce it. Grey zone is where honest explanations compete and judgement has to be stated rather than hidden.

Signatures

The shapes produced activity leaves in a transaction record: spacing that is too regular, sizes drawn from a narrow set, round trips that return their own inventory, and counterparty graphs that collapse to a handful of funders.

Open the section

Evidence

How to collect chain data so that a second analyst can reproduce your reading: which fields to store, which to derive, how wallet clusters are built, and where each method quietly stops working.

Open the section

Grey zone

The region where the same on-chain footprint has several honest explanations. Market making, incentive farming, treasury rebalancing and produced flow overlap here, and the correct output is a confidence level rather than a verdict.

Open the section

Every term, defined once

Volume analysis breaks down when two people use the same word for different things. The glossary fixes each term to one meaning and keeps it there across the whole site, including the terms this desk deliberately refuses to use about identifiable parties.

Open the glossary

round trip inter-arrival gap quantisation funding edge cluster fee payer net inventory depth response adverse selection base rate falsifier confidence band routed leg venue coverage turnover

Where produced activity comes from

Produced turnover is not a hidden practice. Teams run activity through their own pairs so the pair stays visible on venue and aggregator screens that rank or filter by turnover, and the tooling that does it is sold openly, with configurable wallet counts, budgets and cadence.

That matters to this desk for one reason: a category of flow that is configured has parameters, and parameters leave a shape. Knowing what the tooling can be set to is what makes a cadence or a size distribution interpretable rather than mysterious. It is also why the desk treats produced volume as a normal part of Solana market structure rather than as a scandal to be exposed.

The practical consequence is about sizing, not morality. Produced flow runs on a budget, budgets end, and a pair whose turnover was largely produced can return to its underlying depth without any visible event. That underlying depth is what your exit will actually meet.

See how activity is produced

Parameters that leave a trace

How many wallets a campaign spreads across, and how those wallets were funded.
The interval between trades, and whether it is fixed, jittered or event-driven.
The size band each trade is drawn from, and how narrow that band is.
Whether buys and sells are balanced by design, and over what window.
Which venues are included, and whether one pool carries most of the turnover.
When the campaign starts and stops, which is usually the clearest edge of all.

The case method

Three commitments decide what appears on this site and, far more often, what does not.

What counts as evidence

Only what is in the transaction record and can be pulled again by someone else: signers, slots, balance deltas, program ids, fee payers, pool reserves. A reading is publishable when a second analyst could repeat the same pull from the same window and reach a conclusion of their own.

Screenshots of dashboards, aggregate figures with no stated window, and anything relayed second hand are treated as leads, not evidence.

What does not count

Intent is not on chain. Neither is identity, ownership, or the reason an account did anything. Every sentence on this site that would require one of those to be true is either removed or rewritten as a hypothesis with its falsifier attached.

There are no statistics here about how common produced volume is, because the desk has no dataset that would support one, and inventing a percentage would be the exact failure this site exists to describe.

How confidence is stated

Every inference carries one of four bands: high, moderate, low or undecidable. The band is chosen by what evidence exists, not by how interesting the conclusion is, and each band comes with wording that the write-up must use.

A claim is only publishable if the desk can also state the observation that would overturn it. Where no such observation exists, the honest output is undecidable, and undecidable is published as often as anything else.