Make the
exchanges
light up.
Last-mile traceability to the exchanges.
“Which on-chain addresses actually belong to the exchanges?” Readers kept asking after the last report. Nobody had mapped them, and a forensic trace goes cold the moment funds reach an exchange. Your tracer slips in and moves silently through its wallets; a non-fungible token cannot commingle, so the last mile finally has an exact map.
Cardano has no map of exchange wallets
A question readers kept asking after the last report: which on-chain addresses actually belong to the exchanges?
“Which on-chain addresses actually belong to the exchanges?”> readers, after the last report
Nobody has mapped them. Every exchange’s wallets are an unlabelled web: a hub, a few hops, a question mark over who owns it. No comprehensive map exists, so this study builds one, in the open.
Drop a tracer, watch it light up
- 1Mint a tracerOne unique native token. It cannot be split or duplicated.
- 2Deposit it to your exchangeSend it to your own deposit address; we build the tx for you.
- 3Follow it on-chainOne UTxO carries it at every hop. We follow it and update the live map.
Why it works: unlike fungible ada, a non-fungible tracer cannot commingle.
coins mix in a UTxO, the thread is lost
one UTxO carries it at every hop
Wherever the tracer stops, a wallet gets tagged
Deposits sweep to a hot wallet for liquidity, then to cold storage for safety. The non-fungible tracer rides every move.
Sits, swept, or stored, every stop is a wallet we tag.
A public map of which wallets belong to which exchange
Each label is inferred from independent voluntary deposits. On-chain footprint only, not a confirmed entity.
A reusable, public map of exchange-controlled wallets, built to close the loop for any forensic study: when a trace dead-ends at an exchange wallet, the map says which exchange, turning a dead end into an answer. For each cluster we publish the full tracer-label distribution; the agreement among independent depositors, plus whether the address reads as exchange infrastructure, sets its confidence. See the live map →
You mint to map.
This continues the report’s transparency work, from a new angle. A trace goes cold the moment funds reach an exchange, so you mint tracers to use them: drop them into your exchanges and they reveal which wallets the exchanges own. Tools, not collectibles. The map is the point, and it stays public.
We only go when the data will hold
We proceed only with enough volume and enough unique wallets. The same two axes set each published label’s confidence.
> thin data is never published.
> We publish the findings in the future, when the data is sufficient to stand behind and the timing is compatible with the wider investigation.
> This study will not run indefinitely: we update the map as appropriate, and once the findings stabilize it concludes. Because the data is on-chain, anyone can keep tracing exchanges from the tracer policy id and assets and update the findings afterward.
Built to resist poisoning
A bad actor can send a tracer to the wrong, or an arbitrary, address and report a false exchange.
On-chain topology is immutable, so the flow a tracer takes cannot be forged. The self-reported exchange label is a soft input, and poisoning it is the main attack to defend against.
Each cluster shows every exchange its tracers claimed, with counts. The dominant label and the agreement among independent depositors set its confidence; a lone false report stays visible as an outlier, not silently dropped.
A label only counts if the address independently reads as exchange infrastructure: heavy throughput, many distinct depositors, the token junk-drawer, third-party explorer labels. A fabricated target address fails this test.
We seed exchanges with our own honest deposits. Reports that contradict those anchors are caught and down-weighted.
Converging depositors that share a funding origin or one burst of activity count as a single voice, so no lone actor can manufacture agreement.
Every false report burns a tracer, a fee, and the min-ADA that rides with it. Out-numbering honest depositors on a real exchange gets expensive fast.
Every label carries its depositor count, the converging txids, and its corroboration. Thin or contested clusters are withheld. On-chain footprint only, not a confirmed entity.
> A malicious actor can mint tracers and deposit them with a false label. Data integrity holds as long as honest depositors outnumber the malicious ones, which is cheap for us and expensive for an attacker. see how false reports surface on the map →
Four steps, start to finish
Send 50 ADA. We proceed if there is enough interest, full refund if not.
We mint your 50 tracers and hand-deliver them to you.
We build the tx, you pick the exchange and sign.
Weekly analysis as tracers flow through, labelling the exchange wallets.
50 ADA buys 50 tracers, minted FIFO. Anything over 50 is a donation. Want more than 50? Send separate, independent 50-ADA transactions. Mint and delivery happen off-chain to stay ada-efficient, a trust step, stated plainly. Tracers are unique by serial only: no art, no rarity, no perks.
It traces the exchange, not you.
A tracer follows the exchange you send it to, not your wallet history. To keep your part in the study unlinked from your personal funds, do not mint from your personal wallet.
To avoid tying this study to your personal wallets:
- Create a fresh Cardano wallet you use only for this study.
- Fund it with a small withdrawal from a centralised exchange, not a transfer from your existing wallet, so there is no on-chain link back to you.
- Send your 50 ADA to the study from that fresh wallet, and deploy the tracers from it.
Pick a tracer, point it at your exchange
We format your form input into a structured, on-chain metadata record, so anyone can reproduce the study.
{ exchange: "A" }
Only your exchange pick goes in the metadata; the tracer and the deposit are already in the txid and asset name, so the label is permanent and anyone can reproduce it. You just sign.
Get the most from your tracers
This is why a mint is a batch of 50: do not send them all to one address at once. Spread them, and they fan out across more of the exchange's wallets.
If your exchange lets you generate a fresh deposit address, do it and send tracers to several. A new deposit address may be served by a different aggregation wallet, so more of the consolidation layer lights up.
Send in batches over days, not all at once. Deposits get swept on different schedules into different hot and cold wallets, so staggering reveals more of the plumbing.
Map the most, top the board
We keep score: contributors are ranked by the agreed exchange wallets their tracers map. Spread your batch to climb it.
| # | Study wallet | Agreed nodes | Branches | Exchanges | Tracers |
|---|---|---|---|---|---|
| 1 | addr1qxmapp…e1f2 | 41 | 14 | 5 | 50 |
| 2 | addr1qymapp…e2f3 | 38 | 12 | 4 | 50 |
| 3 | addr1qzmapp…f3g4 | 33 | 11 | 3 | 50 |
| 4 | addr1qwmapp…g4h5 | 27 | 9 | 3 | 50 |
| 5 | addr1qvmapp…h5j6 | 22 | 8 | 2 | 50 |
| 6 | addr1qumapp…j6k7 | 19 | 6 | 2 | 50 |
| 7 | addr1qtmapp…k7l8 | 14 | 5 | 2 | 50 |
| 8 | addr1qsmapp…l8m9 | 9 | 3 | 1 | 50 |
Only corroborated nodes score, so the board rewards mapping the wider set agrees on, not raw deposit volume or lone claims. Spreading a batch across deposit addresses and over time is how you climb it.
> example board, invented data; it fills in as tracers land once the study runs.
We follow them in
Exchange wallets are black holes for NFTs: tracers flow in, never out. Each one settles into the exchange’s infrastructure, and we label every wallet it touches.
Grab a tracer
Mint your tracers, drop them in your exchanges, and watch them light up. Full refund if interest is short of the line.
Support the investigation from any chain. It funds the tracing, not a treasury, and it does not mint tracers (those are ADA-only).
bc1qnxx3xg8snykvsj0nufvzj6v4x8dgswthe7t8mp0x89eEc99361B65A73ddd717B322146a3a7776Df876dWNSFmqoauuMBZYtzSspfNuZDTTPa16811SvFz3abJzAn exchange label is an inference from voluntary deposits, on-chain footprint only, not a confirmed entity. Reaching an address shows contact and sweep behaviour, not custody of any named party’s funds. The tracers are research tools, not an investment or a collectible. What this proves, and what it doesn’t →