“People don’t talk about money” – or do they? Recently, at least, there has been increasing talk about money, especially digital fotrms of money such as the digital euro or crypto-assets such as e-money tokens (aka “stablecoins”) and, in this context, the so-called Commercial Bank Money Token and its regulatory classification.
Background
The German Banking Industry (“DK”) has published a comprehensive working paper on the further development of scriptural money in the form of tokens (also known as tokenized scriptural money or commercial bank money tokens – “CBMT”). [1] The regulatory classification of scriptural money tokens is sometimes controversially discussed in the literature, as depending on the design of the scriptural money token, it can either be
– fall under the provisions on e-money tokens in Title IV of the Markets in Crypto-Assets Regulation (“MiCAR”) as a crypto asset or
– as a crypto asset outside MiCAR (see Art. 2 para. 4 MiCAR) if it is classified as a deposit within the meaning of the Deposit Guarantee Schemes Directive (Directive 2014/49/EU) or
– be classified as e-money within the meaning of the E-Money Directive 2 (“EMD2”).
Legal classification of scriptural money tokens
The DK working paper calls for the planned CBMT to be classified as a new technical variant of commercial bank money and not as e-money (p. 12): “We believe that the CBMT needs to be seen as commercial bank money in an alternative technical form, i.e., not provided as electronic money on current accounts but as digital money provided on token accounts (‘addresses’).” The risk of classification as an e-money token is also recognized: “At the same time, it (meaning the MiCAR) regulates e-money tokens which might capture tokenized commercial bank money.”. However, there is no legal justification in the working paper.
In December 2024, the European Banking Authority (“EBA”) published its report on tokenized deposits, which, among other things, deals with the comparison between bank money tokens and e-money tokens (for more details, see our article: EBA’s Report on Tokenised Deposits).
In the meantime, a legal opinion has been published[2] which, among other things, addressed the regulatory classification of the CBMT (“DK scriptural money token”) presented by the DK in its working paper.
The main result of the legal assessment was that the DK CBMT is to be regarded as a deposit rather than an e-money token, meaning that MiCAR does not apply (see Art. 2 (4) b) MiCAR). Classification as e-money is probably ruled out as a result of the exemption for limited networks pursuant to Section 2 (1) no. 10 of the German Payment Services Supervision Act (“ZAG”).
Key points from the expert opinion
According to the regulatory opinion, the DK token would qualify as a deposit. Classification as e-money within the meaning of Section 1 (2) sentence 3 ZAG would exist (contrary to the view expressed in some literature), but would ultimately fail due to the exemption for limited networks. According to Section 2 (1) no. 10 a) Var. 2 ZAG, services based on payment instruments “which … can be used within a limited network of service providers under a business agreement with a professional issuer” are not considered payment services. The limited network of service providers here refers to “all participants in the cryptocurrency token ecosystem” (p. 30 of the expert opinion). In the “multi-issuer” concept of the DK, however, the number of participating banks – as with conventional scriptural money – is not limited in principle. It is not the number, but the close “business relationship of cooperation between the system participants” (p. 31 in the expert opinion) that should – according to the view expressed in the expert opinion – fulfill the requirements of a limited network.
The interpretation of the “limited network” exemption under PSD2/EMD2, according to which “service provider” obviously also or only includes the participating payment service providers (“PSP”), cannot be accepted. The scope exception of the limited network refers to the limitation of the acceptance of the payment instrument. Admittedly, the term “service provider” at this point (see Art. 3 k) PSD2) is not a happy choice, as it is intended to refer to providers of goods and/or services. However, there is no doubt that the term “service provider” refers precisely to those who accept the payment instrument as a means of payment. The relevant EBA guidelines on Limited Network Exclusion (“LNE”) also leave no doubt about this. They refer to “providers of goods and services” [3]. To benefit from the exclusion, this group of acceptance points must be limited, for example to a specific retail chain or region (e.g. via zip code). BaFin specifies a number of criteria in this regard, which can be found in its information sheet – Notes on the Payment Services Supervision Act (as at: 05.07.2024) (under C.X.).
The reference in the expert opinion (p. 31) to a French ruling by the Conseil d’État from 2013 with regard to the application of the “limited networks” exception does not lead to a different approach. According to this judgment, the existence of capital relationships and the closeness of the business relationships between the parties involved[4] could justify an LNE. However, this case concerned the relationship between the issuer of the payment instrument (here: the fashion department store Printemps as the issuer of a gift voucher instrument) and the accepting merchants (here: the Citadium and FNAC store chains belonging to the Printemps group). It was not a question of the relationship between several PSPs that jointly operate a payment system and issue payment instruments that are then used, i.e. accepted, by the respective users as payers or payees (as in the DK’s planned CBMT system). However, the French ruling by the Conseil d’État from 2013 should not be decisive for the interpretation and application of the LNE, but primarily the guidelines on the LNE published by the EBA in 2022 (not mentioned in the expert opinion).[5]
The CBMT system planned by DK could now make use of the LNE if the acceptance of the tokens as a payment instrument is limited accordingly, e.g. only within a chain of stores or a franchise system. However, no such limitation is currently planned. The CBMT is to be used in particular in the B2B sector for various use cases based on distributed ledger technology (“DLT”), such as delivery-versus-payment, pay-per-use and smart contract-initiated payments.[6] However, the use of the token as a means of payment should not be limited: “There should be no programmability of the token (in the form of inherent properties) and a possibly associated purpose of limitation since the universality as a monetary property is then possibly no longer fulfilled and differing economic values due to different risk profiles between the deposit tokens are excluded.”[7]
The limitation of system participants in a payment system and their cooperation relationship are not criteria for the application of the LNE, even if – contrary to the prevailing interpretation of the legal materials, literature and financial supervision – the view is held that the area exception of the limited network should be regarded as a “moving system of abstract-general, mutually complementary or substitutive criteria” (p. 31 in the expert opinion). Assuming that it depends on the limitation and the cooperation relationship between the system participants, this would mean that an e-money system operated by only one institution (such as PayPal) would ultimately be exempt from authorization, as in this case the exception for limited networks would have to apply all the more. Whether multi-issuer or single-issuer, the decisive factor for any application of the LNE is limiting the acceptance of CBMT as a means of payment for the purchase of goods and services.
Conclusion
This means that the discussion on the regulatory classification of CBMTs issued by banks or comparable products remains open and is presumably still in its infancy: it is therefore questionable how CBMTs should be classified under regulatory law – as a technical variant of deposits, e-money or e-money tokens? If the LNE is ruled out, but the e-money criteria are met, the CBMT would be classified as e-money. However, the expert opinion also assumes that the CBMT of DK is to be classified as a deposit. This raises the (follow-up) question of how e-money differs from bank money or e-money tokens from CBMT. This question has not yet been sufficiently addressed. Initial thoughts on this can be found in our Article.
[1] Version 1.51, Status: July 2023, available at https://die-dk.de/themen/pressemitteilungen/digitalisierung-von-bankeinlagen-deutsche-kreditwirtschaft-veroffentlicht-neue-version-des-working-paper-zum-commercial-bank-money-token/ (last accessed on 17.07.2025).
[2] The legal opinion was published via LinkedIn, but could not be found via search engines. The key statements of the opinion for this article were reproduced in the article by Omlor/Heine “Aufsichtsrechtliche Einordnung von Giralgeldtoken”, WM 2025, Issue 17, 729-738.
[3] See EBA, Final Report on the Guidelines on the Limited Network Exclusion under PSD2, 24 February 2022 (EBA/GL/2022/02), p. 19-20.
[4] “l’importance des liens capitalistiques entre ses membres, ou l’étroitesse de leurs relations commerciales”, see: Conseil d’État, 9ème et 10ème sous-sections réunies, 24/04/2013, 354957 (https://www.legifrance.gouv.fr/ceta/id/CETATEXT000027353547; last accessed on 17.07.2025).
[5] The expert opinion takes the view that the case law of the Conseil d’État – without taking into account the more recent EBA Guidelines on LNE – “can at least be used as a persuasive authority” (cf. p. 31 of the expert opinion).
[6] See DK, Working Paper on Commercial Bank Money Token, Version 1.51 (2023), p. 10.
[7] DK (2023), p. 17
