The rail that scales
Everything on this site points in one direction: the route is expensive on small amounts and cheap on large ones. A bank account is the starting point that makes the large end reachable, because its ceilings are set by a bank relationship rather than by an app tier.
| Sent | Reaches the table | Lost | Best case |
|---|---|---|---|
| ₦50,000 | ₦43,104 | 13.8% | ₦48,770 |
| ₦100,000 | ₦90,147 | 9.9% | ₦98,020 |
| ₦250,000 | ₦231,278 | 7.5% | ₦245,770 |
| ₦500,000 | ₦466,496 | 6.7% | ₦492,020 |
The rows a wallet tier will often refuse are exactly the rows where the route costs least per naira.
At ₦500,000 the flat network fee has shrunk to between 0.1% and 0.8% of the amount, and essentially the entire cost of the route is the spread. That is a cost you can shop for, which makes it a different kind of problem from a fee you simply pay.
What the bank step actually costs
Almost nothing, and the range this site uses — 0% to 0.5% — covers both a free instant transfer and one carrying a small charge. On the amounts where a bank account is the right starting point, that is a rounding error next to everything else.
| Amount | This step at 0.5% | The spread at 4% | Ratio |
|---|---|---|---|
| ₦100,000 | ₦500 | ₦4,000 | 1 : 8 |
| ₦250,000 | ₦1,250 | ₦10,000 | 1 : 8 |
| ₦500,000 | ₦2,500 | ₦20,000 | 1 : 8 |
Eight to one, in every row. Time spent shopping for a bank transfer fee is time not spent on the number that is eight times larger.
Which is the same point this site makes about wallets, and it is worth repeating because the instinct runs the other way: the visible, itemised, receipted charge gets the attention, and the unlabelled gap in a P2P quote does not. The spread page is where that attention belongs.
What the bank rail asks in return
Setting it up takes longer, and the account is subject to its own scrutiny — a sequence of transfers to an exchange is a pattern a bank may ask about. Nothing on this site suggests obscuring that, or spreading transfers to avoid attention, or presenting the purpose of a payment as something other than what it is. Where a bank asks, the answer is the true one.
Why the return leg argues for a bank account
Most of the case for a bank rail is about the deposit, but the stronger half is about the withdrawal. Two things meet there. The first is that money should return to the same account, in the same name, that sent it — the single most reliable way to avoid a held payout. The second is that the flat network fee is paid once per movement, so a ceiling that forces a withdrawal into pieces multiplies it.
| Withdrawing ₦200,000 | Movements | Flat fee paid | Worst-case fee |
|---|---|---|---|
| Under a high ceiling | one | once | ₦4,000 |
| Under a ₦50,000 daily cap | four | four times | ₦16,000 |
| Under a ₦25,000 daily cap | eight | eight times | ₦32,000 |
Nothing else on the ladder changes between these rows. The only variable is how many times the same fee is charged, and that is set by a ceiling.
A wallet tier is the usual source of those ceilings, and a bank relationship is the usual way past them. The detail is on limits on the way out, and the fee arithmetic on chains and flat fees.
Bank first, or wallet first?
| Wallet | Bank account | |
|---|---|---|
| Time to set up | minutes | days |
| Cost of the first step | 0–0.5% | 0–0.5% |
| Typical ceiling | tier-bound | relationship-bound |
| Best for | ₦5,000–₦100,000 | ₦100,000 and up |
| Returning a withdrawal | same account, same name | same account, same name |
Neither is better in general. The wallet suits the amounts most people actually start with, and the bank suits the amounts where this route stops being expensive.
Nearby on this route
Three pages that pick up where this one stops: The P2P spread, measured, Starting from OPay, Naira to the table.