Bitcoin in B2B (Role & Limits)

Whether bitcoin holds up as a payment rail is not decided by the technology. It is decided by an ordinary invoice — and by who bears the difference when the rate moves between quote and receipt.

The question is rarely asked in principle

It usually arrives through a supplier. A vendor outside the EU offers to settle invoices in bitcoin because the usual correspondent banking route takes days at their end. That triggers an assessment which has nothing to do with the investment question: not whether holding pays off, but whether the payment process fits your own operations.

The distinction is worth making. Bitcoin as a balance sheet item — reserve, backed lending, valuation, custody — is a separate decision with its own reasoning. It is covered in Bitcoin in a Corporate Context. Here it is only about paying.

The base layer was not built as a payment rail

The Bank for International Settlements described the reason soberly in 2022. Permissionless blockchains can achieve only two of three properties at once: scalability, security and decentralisation. Sustaining participants' incentives through fees buys congestion — and congestion limits volume.

From this follows a property that is unfamiliar in payments. The fee depends on how busy the network is, not on the amount. A small payment therefore becomes expensive precisely when many people want to pay. The BIS draws a broader conclusion. More users here mean more congestion rather than a network benefit. That tendency towards fragmentation is, in its view, the real weakness of the approach as the basis of a monetary system.

How finality comes about in such a network at all, and what it costs in energy, is covered in Consensus Mechanisms (PoW/PoS & More). For the payment question the consequence suffices: on the base layer, a payment is neither fast nor predictable in price.

What Lightning changes — and what it demands in return

The Lightning network starts exactly here. Two parties lock funds in a shared channel and then settle any number of payments between them without each one being written to the chain. Only opening and closing the channel touch the base layer. Payments between parties without a channel of their own travel through the channels of third parties.

This genuinely solves the volume and fee problem for the individual payment. In return it demands something that does not occur elsewhere in payments: liquidity has to be committed in advance, and it is directional. A channel through which money arrives cannot readily send in the same direction again. A company that pays regularly is therefore running a rail it pre-funds and keeps in balance itself.

There is also a dependency that is hard to steer. Whether a payment finds a route depends on the liquidity of other parties' channels, which cannot be inspected. For day-to-day operations that means Lightning is not a setting you switch on but an operational asset with its own ownership.

The difference to stablecoins is not a technical one

Both run over open networks, and both work without a bank in the middle. The difference sits in supervisory law, and for payments it is the whole point.

BaFin classifies bitcoin as an “other crypto-asset”; the relevant MiCAR rules have applied since 30 December 2024. An e-money token (EMT), by contrast, is a crypto-asset whose value stability references an official currency. Its issuer must already be authorised as a CRR credit institution or an e-money institution. Behind a regulated euro stablecoin there is therefore a supervised institution. Behind bitcoin there is no one — which is not a flaw in the design but the design.

For an invoice this is decisive. The invoice is denominated in euros. A means of payment whose value references the euro transfers that amount. A means of payment with its own exchange rate transfers a counter-value that shifts between quote and receipt. Which of the two you want is not a matter of taste but the question of who carries the price risk. What else sets euro stablecoins apart in payments is covered in Stablecoins – Fundamentals.

Where day-to-day operations draw the line

An invoice for 84,000 euros is to be settled in bitcoin. The supplier quotes an amount in BTC, converted at a rate that by its nature holds only for a moment. Between internal release and receipt at the supplier lie minutes to hours. If the rate moves two per cent in that time, around 1,700 euros are missing — or 1,700 too many have been sent.

That question cannot be solved technically. It is a contract clause: which rate applies, at what moment, within what tolerance, and who bears deviations beyond it. Without that clause every payment produces a small dispute, and that costs more handling time than the rail saves in fees.

Three further points show up in daily use, and all three sit outside the technology:

• Early payment discounts and instalments presuppose a fixed amount, which first has to be defined in a fluctuating unit.
• Unwinding a payment after a faulty delivery is not an operation in the network but a fresh payment by the supplier — there is no direct debit return.
• Every payment is also a disposal of an asset and therefore lands in accounting, not only in payments.

For the CFO the last point is the most awkward. A payment rail that triggers a valuation-relevant event on every use is no longer purely a payment rail. Added to that is the pre-funded liquidity in the channels: committed capital that does not arise this way in account-based payments.

Where it does hold up

The conclusion is not a verdict on bitcoin but an allocation, and the benchmark decides it. Against SEPA Instant inside the euro area the route loses clearly, being more expensive, slower and less settled in law. Against a multi-day correspondent banking route into a country with weak banking access it can win.

Two conditions carry that case. The counterparty holds bitcoin anyway and does not want to convert immediately — otherwise the on/off-ramp merely moves the problem. And the amounts are large enough for the rate clause and the operational effort to pay off. For many small payments inside the euro area, neither applies.

What remains

Bitcoin is not unsuitable as a payment rail, but it is a special route with its own operating cost. It does not replace a rail that works, and it solves no problem that exists inside the euro area. Where it holds up, it holds up against a poor alternative, not against a good one.

More useful than the question of principle is therefore the order of steps in the individual case. First comes the benchmark: what exactly is this route competing against? Then the rate clause. The technology is the part that causes the least work in the end.