Ask a bank or a VASP how it checks where a client’s crypto wealth came from, and the answer is usually two tools. A crypto tax calculator pulls the client’s exchange and wallet history. A blockchain analytics tool screens the addresses. An analyst puts the two outputs side by side and assembles the case by hand.
ChainComply has seen this setup at banks and at VASPs, in most of the institutions it works with. It is not a lapse of judgement. Until recently nothing better existed, and the tax calculator had the one thing the desk needed first: API connections to the exchanges.
The setup carries simple cases. It starts to fail at roughly the point where crypto wealth becomes worth reviewing.
What the tax calculator shows the desk
A tax calculator is built to answer what gain was realised, when, and what tax it triggered. For a reviewer that is useful: realised gains per year, and in most tools a simplified capital-gains view of how the position performed over time.
It was not built to show the route of the money. Where did a transfer arrive from, and where did it go? Was a movement a transfer between two of the client’s own accounts, or a payment to someone else? A tax engine needs only enough of that to match disposals against acquisitions. The Source of Wealth question needs all of it. Only a few calculators handle the fiat side well, and growth that was never sold appears, when it appears, as one global figure.
What blockchain analytics adds, and where it stops
The analytics tool fills in the chain: addresses, counterparties, labels, exposure scores. Its output is an indicator for the analyst, not a verdict on the client.
With one exchange and a couple of self-custody wallets, the two tools together can carry a case. Across two VASPs it gets hard. Past three, the combined picture no longer fits in one analyst’s head, and a hard case can involve fifteen to thirty accounts and wallets.
The reason is that exchanges are the stopping point. Once funds reach an exchange they are mixed with those of millions of other clients, and the on-chain trail stops at the deposit address. The trades, the conversions, and the fiat deposits and withdrawals live in the exchange’s own records: the data the tax calculator imported, without the originating and destination addresses the analytics tool needs. Unless those addresses are added back, the movements in and out of exchanges are never screened, and counterparty risk is checked only on the part of the history that happened to stay on-chain.
And even with the trail complete, a blockchain analytics tool does not calculate gains. It sees value arrive and leave, not what it was bought at or what it earned in between, so the investment half of the story, how the wealth was actually made, is invisible to it.
The artisan’s table
So the case is rebuilt by hand, for as long as the review lasts, from pieces held in two systems built for two different purposes. The tools meet at exactly two points. Realised gains appear on the tax side and matter to the Source of Wealth review. Risk labels appear on the analytics side and matter to it too.
Everything else is carried by the analyst:
- The same history entered twice, once into each tool, with nothing to guarantee that the two copies agree.
- Balances reconciled against flows by hand, because neither tool does it. In crypto the current balance is the initial investment plus gains plus net flows: a current account and an investment account in one.
- Missing addresses on exchange movements looked up and added, or left as a gap that screening never reaches.
- Missing wallets discovered from the story as a whole. It is the fastest route to the right question for the client, and neither tool is designed to prompt it.
- Every transfer classified as internal or external. Count the client’s own moves as inflows and outflows and the case confuses itself. The classification is needed in every case and is the hardest step to get right.
When the analyst is lucky, most of the information exists somewhere. It is still scattered, and each of those steps is a place where data goes missing, errors enter and control of the file loosens.
What falls between the tools
Put the two outputs side by side and the gap has a shape. The fiat that came in and went out: sometimes in the calculator, often not. The growth that was never sold. Which transfers were the client’s own. The client’s account of how the wealth was built, and whether the transaction data supports it.
None of this is exotic, but putting it together in a defensible decision package is a very complex task. Between them, these pieces are the Source of Wealth question itself, and the part no tool on the table owns.
What the table costs
Collecting the data is the small part. Analysis is around ninety percent of the job, and on the table it is the slow part: a medium-risk crypto client takes 4 to 20 analyst hours when reviewed manually, a high-risk one 20 to 200. Then comes the synthesis, turning two tools’ output and a stack of notes into one document someone else can read and decide from. The familiar failure is the file that documents every flow for weeks and that nobody can read end to end.
The consequence is structural. In ChainComply’s experience the table runs at about a tenth of the efficiency the work allows, so Source of Wealth is checked rarely, and mostly when a case is already large or alarming enough to justify the hours.
A file built for the question
The alternative is not a better table. It is one data set that holds both halves from the start: exchange and wallet histories reconstructed once, transfers classified, fiat flows included, and exchange movements carrying their addresses so that screening covers them. Built that way, the two meeting points stop being reconciliations. The realised gains agree with the tax side because they come from the same records, and the risk labels sit on counterparties the file already knows.
Every counterparty in that data set, including the ones behind exchange deposits and withdrawals, is then screened automatically with blockchain analysis tools, and its labels and risk scores are carried into the file. The analyst starts from a screened picture instead of assembling one.
What reaches the reviewer can then answer five questions directly: what the client’s Source of Wealth story is, what data was reviewed, what explains the current balance, what risks or gaps remain, and why the conclusion is defensible. Perfection is not the bar. A balance around ninety percent explained can be acceptable, provided the rest is named and bounded.
The same three circles, drawn for the crypto holder rather than the desk, are published by CashoutReady, the consumer service ChainComply runs. The two roles are kept apart. ChainComply works for institutions, on the reviewer’s side of the desk. CashoutReady prepares the client’s side: a claim the client commissions, which reaches the reviewer as evidence to test rather than as a verdict. From the holder’s side the question is which document answers the bank. From the desk, it is how much of the case the tools leave for the analyst to carry.