Custodial vs Non-Custodial Swaps — the Real Risk

When people evaluate a swap, they fixate on privacy claims and skip the question that has cost more users their funds than any logging policy: does the service hold your money at any point? Custody is where exit scams, freezes, and honeypots live. It deserves more attention than it usually gets.

The distinction in one line

That holding window is the entire risk.

Why custody is the real risk

Almost every way a swap goes catastrophically wrong requires the service to hold your funds first:

Privacy claims are important, but a perfectly private service that holds your coins can still lose or freeze them. Custody risk sits above privacy risk in the hierarchy of what can hurt you.

Why non-custodial also helps privacy

The two are related. A service that never holds a balance and never requires an account simply has less to collect and less to retain. There is no stored wallet to log, no deposit history accumulating under your identity, no withdrawal profile building over time. Non-custodial design reduces both the custody risk and the amount of data a "no-logs" promise even has to cover.

How to tell them apart

You can often recognize custodial behavior by its shape:

The takeaway

Judge a swap on custody first, privacy second. A non-custodial flow removes the entire category of risk where the worst outcomes live — exit scams, freezes, honeypots, hacks — because the service never holds the funds that those failures require. It also happens to minimize the data any privacy promise has to protect. Verify the structure (no deposit-then-withdraw, no stored balance, destination given up front), and you have addressed the risk that actually matters most.