Last updated:

Compare banking access by jurisdiction

Two questions, kept apart. Whether an account is realistically available at an institution in a jurisdiction, and whether the business can be banked at all counting the multi-provider stack the network places. A jurisdiction can fail the first and pass the second, and that is the normal case rather than a contradiction.

Not sure which column is you? Ask Emma. She compares these jurisdictions in seconds, in your language.

Results

0 of 0 match
Non-matching jurisdictions stay visible, dimmed, with the reason, so you can see what relaxing a filter would buy you. Ratings marked not rated are exactly that: no rating exists yet, which is not the same as poor access.
Each rating also carries how we arrived at it. read means we placed it from a real case or from our own published cost tables. inherited means we wrote it for a neighbouring vertical and carried it across: directionally sound, not read for this profile. inferred means it was placed by rule relative to another tier: treat it as a starting point and confirm it before you rely on it. Most stack bands are currently inherited, and saying so is the point.
[  About  ]

About this comparator

This tool compares 55 jurisdictions on banking access. It exists because the same question was previously answered in two places on this site that did not agree with each other, and because banking access is the one thing a licensed operator has no honest published source for.

Because they answer different questions. Local route is whether a business domiciled there can realistically open an account at an institution in that jurisdiction. Workable stack is whether the business can be banked at all, counting the multi-provider architecture of local accounts, EMIs elsewhere and correspondent relationships.

A jurisdiction whose local route is rarely available while its stack is workable is the ordinary case, not an inconsistency. Collapsing the two into one number is exactly how the earlier contradiction on this site arose.

That no rating exists yet for that cell. It is not a poor score and should not be read as one. Coverage is published as it stands rather than filled in with estimates, because a comparator whose gaps are invisible is worse than one whose gaps are stated.

From the firm’s banking matrix, which is also what the jurisdiction table on the banking overview renders and what Emma reads. All three are generated from one file, so they cannot disagree.

Each record carries the month it was last updated, shown on its card.

Not on its own. These ratings describe the market; placement depends on the specific business, its licence status, its processing history and its documentation. The assessment is where that is established, and it is free.

Because those are recorded for 13 of 55 jurisdictions so far, and a filter over incomplete data silently hides jurisdictions that simply have not been rated. They are shown on the cards where known and will become filters when coverage supports it.