Reading a peel chain, and why it defeats most people
Peel chains beat investigators through patience rather than sophistication, which is precisely why they survive contact with a spreadsheet. Here is the structure, why manual tracing fails on it, and the signatures that give it away.
Ask most people how stolen cryptocurrency gets laundered and they will describe a mixer. Mixers get the attention because they are conceptually dramatic. In actual regional casework, the pattern we meet far more often is duller and considerably more effective: the peel chain.
The structure
Someone controls a large balance in a single address. Rather than moving it in one transaction, which is trivially followable, they send a small portion to a destination they care about, typically an exchange deposit address or an off-ramp. The remainder, the change, goes to a fresh address they also control.
Then they do it again. And again. Each transaction peels a modest amount off the main balance and pushes the rest one hop further along. A chain of a few hundred hops is unremarkable. Each individual hop looks like an ordinary payment with change returned, which on a UTXO chain like Bitcoin is exactly what almost every honest transaction also looks like.
The technique does not hide anything cryptographically. Every hop is public and permanent. What it does is convert a single obvious movement into several hundred boring ones, and it relies on the investigator running out of time, budget or attention before the chain runs out of hops.
Why manual tracing fails here
An analyst following a peel chain by hand faces a decision at every hop: which output is the payment and which is the change? Get that wrong once, early, and everything downstream is wrong. The analyst does not usually discover the error. They simply arrive at a confident conclusion about the wrong branch.
The danger of a peel chain is not that it cannot be followed. It is that it can be followed incorrectly with complete confidence.
This is why we treat change identification as a step that must be justified rather than assumed, and why a report that traces two hundred hops without stating how change was determined at each decision point is a report we would not sign.
The signatures
Peel chains leak information in several ways at once, and the reliable approach is to look for agreement between independent signals rather than trusting any single heuristic.
Value asymmetry. The change output is usually much larger than the payment output, and stays large across many hops while the peeled amounts remain small and fairly consistent. A long run of transactions where one output is two orders of magnitude larger than the other is a strong structural hint.
Address reuse and script type. Wallet software generates change to a fresh address of the same script type it spends from. A transaction spending from a native SegWit input to one native SegWit output and one legacy output is telling you something about which output the wallet made and which the user typed in.
Round number payments. Humans send round amounts. Change is whatever is left after fees, so it is almost never round. An output of exactly 0.05 BTC alongside an output of 1.83472916 BTC is not ambiguous.
Timing rhythm. Automated peeling produces suspiciously regular intervals. Manual peeling produces clusters during waking hours in a particular timezone, which is intelligence in its own right.
Fee behaviour. The same wallet software tends to make the same fee estimation choices. A consistent fee rate policy across a long chain supports the inference that one actor controls the whole thing.
Where the chain actually ends
The peeled amounts have to go somewhere, and that somewhere is usually a service. This is the practical point: the chain itself is rarely the objective. The objective is the set of deposit addresses at the far end, because those are the points where a real identity was presented to a business that keeps records.
In regional cases those endpoints are frequently not major international exchanges. They are peer to peer desks, local traders, and small services that convert to mobile money. Those endpoints often keep worse records, but they also often keep records that a well-framed request can actually obtain, and they are frequently more responsive than a large offshore platform.
What we tell investigators
Do not attempt to follow every hop manually. Establish the peel pattern, characterise it, and then reason about the endpoints as a set rather than a sequence. The question that matters to a prosecution is rarely "what happened at hop 147". It is "which services received value from this chain, in what amounts, over what period, and what did those services know about who was receiving it".
And record your uncertainty. If change attribution at a particular hop rests on a single weak signal, say so in the report. A defence expert will find that hop. It is far better for the court to hear about it from you first, in your own framing, than from them in theirs.
Written by the NuruTrace team
We publish what we learn from casework and from reading the regime closely. If something here is relevant to a matter you are handling, we are happy to talk it through.