Fake volume bots: dashboard activity is not the same as independent demand

The short answer

A fake-volume bot automates trades intended primarily to create the appearance of market activity rather than express independent demand or manage genuine inventory risk. It can increase displayed volume, buys, sells, transactions, makers, holder activity, or chart motion by moving the same economic capital repeatedly through one or many wallets.

Bot activity is not automatically fake. Arbitrage, market making, routing, rebalancing, liquidation, and DCA are routinely automated. The investigation asks what the activity accomplished economically: who funded it, how positions changed, which costs were borne, whether outside participants joined, and whether the pattern persisted when the suspected wallets stopped.

Start by defining what the dashboard counts

Fix the chain, pair, quote asset, venue, and time window. Record whether “volume” sums the notional value of swaps, whether buys and sells refer to the base token, how routed transactions are assigned, and whether “makers” means signers, fee-payers, token accounts, or another unique address field.

DEX Screener's public pair schema exposes pair-level transaction, volume, liquidity, and price-change fields. Those are useful indexed observations; the schema does not turn an address into a person or certify that volume represents independent demand. Use the DEX Screener field guide to preserve pair identity and field definitions.

One economic loop can inflate several metrics

• Gross-volume loop: related wallets alternate buys and sells, counting notional on every pass while combined exposure changes little.
• Maker multiplication: a controller funds fresh wallets that each perform one or more swaps before consolidating funds.
• Transaction splitting: one desired notional is divided into many small trades to raise transaction and activity counts.
• Chart shaping: timed or sized trades create candles, apparent buy pressure, or repeated prints around a target level.
• Incentive cycling: trades chase points, rebates, leaderboards, fee distributions, or trending visibility rather than independent token demand.

A pattern can target several outcomes at once. State which metric appears affected instead of using “fake volume” as a complete explanation.

Normalize routed swaps before counting behavior

One user trade can touch several pools and intermediary tokens. Aggregators may split a route across venues, and a transaction can contain multiple swap instructions. If each leg is treated as a separate participant or independent decision, both volume and transaction structure can be misread.

Decode the signer, fee payer, input and final output, route plan, pools, inner instructions, transfers, and balance deltas. Group legs that belong to one user intent while retaining pair-level volume as the dashboard records it. Solana transactions explicitly contain one or more instructions, so instruction count and user count are different measurements.

Reconstruct entities before calculating participation

Trace wallet creation, funding source, transfers between participants, gas or SOL top-ups, fee payer, transaction construction, settlement destinations, and later consolidation. Apply the entity-inference evidence ladder and keep service wallets, exchange withdrawals, routers, and common launch tools from becoming false ownership links.

Report raw unique addresses and evidence-adjusted entity estimates separately. Do not replace an unknown beneficial owner with a confident “bot cluster” label simply because several wallets are fresh or mechanically active.

Gross activity can hide a nearly unchanged position

For each supported entity set, reconcile opening token and quote balances, external deposits and withdrawals, swaps, internal transfers, closing balances, network fees, priority fees, tips, transfer fees, and pool fees. Measure gross volume beside net token change, net quote change, realized loss, and remaining inventory.

Include LP ownership. A controller that supplies the pool may recover part of the swap fees paid by its own trading loop, although it still faces execution cost and external arbitrage. Cost does not prove organic demand; it shows the price paid to manufacture the signal.

Look for activity that survives outside the suspected cluster

• New independently funded entities enter and remain after the bot window.
• External net quote inflow grows rather than cycling back to common funders.
• Hold times and position outcomes broaden beyond repetitive short loops.
• Volume appears across independently controlled pools and venues.
• Price and liquidity remain resilient when the suspected wallets stop.

Compare equal windows before, during, and after the activity. Also compare the token with launches of similar age, liquidity, volatility, and attention. Research on crypto volume warns that statistical and behavioral regularities can reveal anomalies, but a token-level conclusion still needs transaction and entity evidence.

Fake-volume evidence and wash-trading evidence overlap but are not identical

A dashboard can be distorted by route overcounting, incentive farming, or maker multiplication without enough evidence to show the same controller took both economic sides. Conversely, self-financed round trips may strongly support common control even when the displayed volume increase is small.

Use the wash-trading investigation for the common-control and independent-risk analysis. Avoid making a legal classification from a bot signature or dashboard anomaly alone; product and jurisdiction matter.

A fake-activity investigation workflow

1. Freeze the metric. Save pair, venue, quote, window, dashboard, field definition, displayed values, and ranking or incentive context.
2. Collect underlying transactions. Decode signers, fee payers, instructions, route legs, pool amounts, balance changes, fees, and timestamps.
3. Normalize user intents. Separate one routed swap from multiple independent trades and distinguish pair volume from token-wide activity.
4. Build entity hypotheses. Trace funding, transfers, common construction, counterflow, consolidation, and service explanations.
5. Reconcile economics. Compare gross volume with combined net exposure, external flow, fee burden, LP recovery, and retained participants.
6. Write a metric-specific conclusion. Separate automated trading, suspicious repetition, maker inflation, recycled volume, supported common control, and unresolved activity.

What belongs in the journal

Record pair and token addresses, venue and quote asset, window, dashboard and metric definitions, screenshots or API values, all underlying swaps, route normalization, signer and fee payer, wallet funding and entity confidence, repeated sizes and cadence, gross volume, opening and closing positions, external net flow, fees and LP ownership, participant retention, comparison windows, incentive or ranking context, alternative automated strategies, and the narrow metric-level conclusion.

Primary sources