Making Sense of Electronic Money
Dr. Manfred Kohlbach
Lecturer, Institute for European Legal Development, University of Graz
This article offers (1) a comparative overview of electronic money regulation in America and Europe as well as (2) a second-level questioning of the value of such regulation. The FSA’s recent decision to classify PayPal (Europe) Ltd. as an electronic money issuer is critically discussed, revealing some of the shortcomings of the European regulatory regime (definitional uncertainty and, paradoxically, limited consumer protection; lack of financial incentives for European ‘start-ups’). It is in light of these shortcomings that the American system is presented as a superior approach to regulating, and making sense of, electronic money.
Keywords: Electronic Money, Electronic Commerce, Digital Cash, PayPal, Micropayments, Smart Cards, Relevance of Regulation.
This is a refereed article published on 30 April 2004.
Citation: Kohlbach, 'Making Sense of Electronic Money', 2004 (1) The Journal of Information, Law and Technology (JILT). <http://elj.warwick.ac.uk/jilt/04-1/kohlbach.html>. New citation as at 15/07/04: <http://www2.warwick.ac.uk/fac/soc/law/elj/jilt/2004_1/kohlbach/>.
Electronic money products and electronic money regulation have been around for years, but scholarly interest in the subject is a much more recent phenomenon and has only just matured. Much of the existing academic commentary traces the genesis of the EU’s Electronic Money Directives (Batalla 2001; Chuah 2000; and Vereeken 2000), or provides detailed accounts of subsequent Member State implementations (for the UK see Long & Casanova 2002 and 2003; and, more recently, Bamodu 2003). What is still missing is a comparative, international analysis of these efforts as well as a second-level questioning of the value of these efforts.
This article seeks to address both perceived shortcomings. A short historical sketch of two divergent electronic money paradigms (section 2) will pave the way for a comprehensive presentation of European, British and American regulatory regimes (sections 3-5). This threefold presentation will, in turn, provide the basis for a critical perspective on the value of regulation in Europe (sections 6 and 7). The text will focus on the UK’s recent decision to classify PayPal (Europe) Ltd. as an electronic money issuer. The PayPal classification is instructive and will illuminate a number of issues – some of them serious – that face European regulation as it stands. Three main problems will be identified. They will consolidate (section 8) the article’s implicit thesis that the American system offers a superior approach to regulating, and making sense of, electronic money.
When electronic money (or ‘e-money’, for short) was first presented as an integral part of the ‘digital revolution’ in the 1990s (Levy 1994; Steinert-Threlkeld 1996) most products were based on the metaphor of the ‘electronic wallet’. ‘Digital coins’, stored offline on smart cards or on user’s hard disks, were the ruling paradigm. ‘Micropayments’, small value purchases for online content (newspaper articles, say; or music clips), were seen as electronic money’s defining application. Despite the technological hype, consumers were apathetic, merchants were unimpressed, and most schemes disappeared as quickly as they had surfaced. Still, a number of risks were identified, and possible legal regulation based on ‘digital coin’ metaphors and smart card technology was debated.
Today, payment systems (such as PayPal) and services offered via mobile phones are electronic money’s new paradigm. The technology is online and predominantly account-based. Payments are ‘macro’ rather than ‘micro’. And consumers and merchants seem much more impressed with available services and solutions. Electronic money has come of age.
This shift has affected e-money’s definitional characteristics. Under the old paradigm it was natural to define electronic money as ‘monetary value charged and stored on an electronic support, in the form of a smart card or incorporated into the memory of a computer’ (Batalla 2001, p. 81). The emphasis was on electronically stored value – value in opposition to information to substitute one contractual debt for another, as in the case of credit cards; or value in opposition to information to instruct a bank to transfer money from an account, as in the case of debit cards (Robertson 1999, pp. 250-1; Solomon 1997, pp. 64-5). Definitions under the new paradigm, by way of contrast, are often technologically neutral and focus on structural characteristics of electronic payment systems. According to the structural view an e-money issuer sells tokens of electronically stored monetary value upon receipt of funds from a customer, a bearer. The bearer buys these tokens using any payment system other than the e-money system in question (cash, cheque, credit card, debit card; a different e-money system even), and may then purchase goods or services from any vendor who accepts the tokens. Upon receipt of the tokens the vendor can either use them to buy goods or services herself, or ask the issuer to redeem them (i.e. to exchange them for physical cash, for a cheque, etc.). The issuer, in turn, has four principal ways to make a profit in this system: (1) she may add a surcharge to any sale of tokens, (2) she may decide to redeem tokens at a discount, (3) she may invest the funds she receives (the interest earned is called seigniorage), or (4) she may offer the funds she receives as credit to a third party.
Not all of these possibilities are without risk. Depending on the nature of the investment in (3) the issuer might generate a loss; the risk here is that she might not be able to redeem tokens. (This is only one possible scenario, of course; the issuer might fail to redeem tokens for other reasons also. In any case, consumer confidence would be shattered.) Offering credit in (4) increases an economy’s money supply; the risk here is that if the sums involved are not regulated inflation may occur.
In light of these considerations, it is plain why legislators might take an interest in regulating the issuance of e-money. Both e-money spending individuals and the economy as a whole are at a potential risk. Indeed, the European Union thinks that the risks involved are critical enough to justify a complex regulatory regime – at the heart of which are two Directives, Directive 2000/46/EC (‘the E-Money Directive’, or simply ‘the Directive’) and Directive 2000/28/EC. 
The way this regime is set up might seem confusing at first. Directive 2000/28/EC creates a two-pronged definition of ‘credit institution’ to be inserted as Article 1 (1)(a) and (b) into an existing Directive, The Banking Co-ordination Directive 2000/12/EC. Let us, for the sake of simplicity, refer to the first type of credit institution, which includes banks and building societies, as ‘type (a)’ and to the second type of credit institution as ‘type (b)’. With this distinction in place Article 1 (3)(a) of the E-Money Directive proclaims that it shall apply to ‘electronic money institutions’, which are defined as
an undertaking … other than a credit institution as defined in Article 1, point 1, first subparagraph (a) of Directive 2000/12/EC which issues means of payment in the form of electronic money
The upshot of this is that type (a) credit institutions, such as banks and building societies, may issue electronic money, but are regulated under existing provisions. The E-Money Directive itself applies to new type (b) institutions only. Conversely, inclusion of type (b) institutions under the Banking Co-ordination Directive’s definition of ‘credit institution’ is slightly arbitrary  (and somewhat awkward  ), but it allows the E-Money Directive to be brief, since many provisions pertaining to reserve requirements, money laundering, and the prudential operation of business are already enacted in existing regulations that apply to credit institutions in general. Article 1 (3)(a) is exhaustive: paragraph (4) stipulates that any institution that does not fall under either type (a) or type (b) is to be prohibited from issuing electronic money.
Electronic money itself is defined in Article 1 (3)(b) as
monetary value as represented by a claim on the issuer which is:
(i) stored on an electronic device;
(ii) issued on receipt of funds of an amount not less in value than the monetary value issued;
(iii) accepted as a means of payment by undertakings other than the issuer.
The wording in (ii) is designed to deter issuers from creating artificial value by giving out more e-money than customers pay for. The formulation (‘funds … not less in value’) allows issuers to give out less e-money than customers pay for however. This was identified in (1), above, as one of the ways for an issuer to make a profit (we will return to the other ways in a moment). The criterion in (iii) is designed to demarcate ‘electronic money’ in the EU’s definition from similar products such as tube tickets, phone cards, photocopy cards and ski passes, most of which are only accepted by one party – the issuer herself.
With the two initial definitions in place, the Directive next covers the applicability of the Banking Directives and asserts, in Article 2 (3), that a receipt of funds will not constitute a deposit within the meaning of Article 3 of Directive 2000/12/EC, ‘if the funds are immediately exchanged for electronic money’. This provision is important because special requirements (and special insurance covers) pertain to deposits under the latter Directive. Given its importance, some scholars (Chuah 2000, p. 183; see also Long & Casanova 2003, p. 11) have suggested that more guidance is needed with respect to the meaning of ‘immediately exchanged’. How is immediacy affected if there is a time lag between a consumer purchasing a smart card and the e-money tokens being activated for instance? What if e-money issuers wish to wait for payments to clear before they issue tokens in return? How immediate is ‘immediate’? The Directive is reticent. It is submitted that the most sensible approach to answering these and other questions would be a functional one. If there is a functional connection between a delay and a purchase the exchange should be viewed as ‘immediate’ within the meaning of the provision. Thus, a time lag between purchase and activation should not have any effect on Article 2 (3) if the lag is a result of the technology in use. In this case any delay involved is a function of the very act of purchasing. The same applies to any waiting period for payments (by cheque, for instance) to clear. The reason for such a delay is to make sure that the funds are actually received by the e-money issuer. Again, the delay is a function of the act of purchasing; and the exchange is ‘immediate’ in that (and as long as) it is effected as soon as the payment goes through and the funds are actually received.
The Directive continues with a number of restrictions and requirements. Article 1 (5) restricts all business activities of electronic money institutions outside of the issuing of e-money to (a) the provision of closely related non-financial services and (b) the storing of data on behalf of other undertakings or public institutions.  In addition, electronic money institutions are not allowed to have any holdings in other undertakings except where these undertakings perform operational or ancillary functions related to e-money. Logically connected to these requirements are strict initial capital and ongoing funds requirements in Article 4. Initial capital must be at least EUR 1 million. Ongoing own funds must be at least two per cent of the institution’s financial liabilities related to outstanding electronic money.  Surprisingly, investment limitations are stricter still. Electronic money institutions must have investments of an amount of no less than their financial liabilities related to outstanding electronic money; and these investments must be in assets with a zero credit risk weighting and with sufficient liquidity.  What’s more, the assets are valued conservatively at either cost or market value, whichever happens to be lower.  It is apparent that all this severely restricts an issuer’s potential return on investments, which was identified as an important source of income in (3), above.
What about other sources of income? In (2), above, we identified a discount at redemption as a possible source. It turns out that this, too, is restricted under the EU regime. Article 3 stipulates that redemption must be at par value in bank notes or coins, or by transfer to an account. The e-money issuer may only stipulate two conditions: she may charge for any costs necessary to carry out the operation, and she may set a minimum threshold for redemption (up to a maximum of EUR 10). The last source of possible income, identified in (4), above, is credit lending. Is it an option? Despite the fact that electronic money institutions are ‘credit institutions’ within the meaning of Directive 2000/12/EC the answer is a definite no. The relevant provisions in the Banking Directives are expressly excluded – electronic money institutions are not, Article 2 of the E-Money Directive stipulates, allowed to extend credit.
In light of these rather severe restrictions, it is almost paradoxical that within the same framework electronic money institutions should benefit from what is known as the ‘single passport’ – mutual recognition arrangements that enable a credit institution established in one Member State to operate throughout the EU. It almost seems as if the conjunction of investment restrictions, funds requirements and passport freedoms gives e-money issuers who can’t turn a profit in their own Member State a ‘licence’ to not make a profit in the rest of the Union either.
There is an exception to all of this though. Article 8 expressly allows Member States to waive the application of some or all of these provisions, including the application of Directive 2000/12/EC to electronic money institutions, provided that the electronic device in question is restricted to a maximum storage capacity of EUR 150 and either (a) total business activities of the issuer do not normally exceed EUR 5 million and never exceed EUR 6 million, or (b) the e-money issued is only accepted by the issuer’s subsidiaries, by the issuer’s parent undertaking or any subsidiaries of that parent undertaking, or (c) the e-money issued is only accepted by a limited number of undertakings (which are within a limited local area, or in a close financial and business relationship with the issuer). The flip side to this waiver is that the single passport is not available to those (smaller) e-money issuers that benefit from a laxer regime. 
The remainder of the Directive is quite inconspicuous. Article 6 provides for compliance checks by competent authorities on own funds and investments. These checks are to be made at least twice a year. Article 7 demands that electronic money institutions be soundly and prudently managed – a demand that seems superfluous given that e-money issuers are within the scope of Directive 2000/12/EC which stipulates similar requirements already. Article 9 deals with ‘grandfathering’, i.e. rules on how Member States are to treat electronic money institutions that commenced business activity before the Directive’s entry into force. Finally, Article 11 stipulates that the Directive is to be reviewed by April 2005.
According to the Recitals that precede the main text, the Directive is devised to achieve three main purposes: to assist electronic money in delivering its ‘full potential benefits’ (Recital 5), to ‘ensure bearer confidence’ (Recitals 4, 9 and 10, this includes consumer confidence), and to ‘preserve a level playing field between electronic money institutions and other credit institutions issuing electronic money’ (Recital 12).  It is submitted that the Directive successfully achieves the second and third objective. Redeemability on par value and strict investment regulations strengthen bearer confidence (but strengthened confidence does not necessarily equal strong legal protection; more on this in section 7, below). Restrictions on business activities and strong own funds requirements place newcomers on a level playing field with established credit institutions. Incorporation into the larger framework of the Banking Directives may not be ideal in a conceptual sense (it is confusing to define e-money issuers as ‘credit institutions’, yet deny them the one activity that would seem to make a ‘credit institution’ a credit institution). But still, and legally speaking, a level playing field is created.
A serious issue with this Directive is the question whether a level playing field is what is needed. Large credit institutions are much more flexible in their investments and have a large and established customer stock to work with; they comply with the Banking Directives already, and need not change any of their business practices. Small-scale e-money issuers will be exempt from many of the onerous restrictions and may be successful on a local level. But what is missing is a growth path for small issuers that want to go transnational, and a framework that provides financial incentives for ‘start-ups’ to launch new, Union-wide payments products. The Directive would hence appear to fail in its first objective, which is ‘to assist electronic money in delivering its full potential benefits’.
There are many things to say in response, of course. The United States take a very different approach. But is it necessarily better? Some payments may depend on a more flexible scheme. But does regulation really matter? Does it matter that much? These issues will be dealt with in later sections of this paper. Right now it is more sensible to remain focussed on strictly legal questions, implementation in particular.
Regulations that implement the E-Money Directive in the UK are scattered over a number of provisions (see Bamodu 2003 for a thorough overview) and include three principle sources: (a) the Financial Services and Markets Act 2000 (Regulated Activities) (Amendment) Order 2002, (b) the Electronic Money (Miscellaneous Amendments) Regulations 2002 and (c) the Financial Services Authority’s (‘FSA’) Handbook of Rules and Guidance (‘the Handbook’), which contains a new ‘ELM’ manual for e-money (other parts of the Handbook, including ‘AUTH’, have been modified also). The Electronic Money (Miscellaneous Amendments) Regulations 2002 are brief and, as their name would suggest, merely amend the wording of miscellaneous primary and secondary legislation. The main implementation issues concern the Financial Services and Markets Act 2000 (Regulated Activities) (Amendment) Order 2002, which amends the Financial Services and Markets Act 2000 (Regulated Activities) Order 2001 (‘the Order’), and the Handbook.
As a rule, the Order stipulates that the issuing of electronic money is to be a regulated activity under the Financial Services and Markets Act 2000 (‘FSMA’). Electronic money issuers are thus subject to supervision by the FSA. Only small-scale e-money schemes are exempted from this rule, but require a special certificate in turn. The Order’s Article 3 (1) defines ‘electronic money’ as
monetary value, as represented by a claim on the issuer, which is—
(i) stored on an electronic device;
(ii) issued on receipt of funds; and
(iii) accepted as a means of payment by persons other than the issuer;
A direct comparison with the Directive’s definition shows that this only implements (i) and (iii) fully; implementation of (ii) is limited to an issuing of funds on receipt, but the Directive’s additional criterion – that the funds be ‘of an amount not less in value than the monetary value issued’ – is dropped. Article 9H (1) adds that
The [Financial Services] Authority may make rules applying to authorised persons with permission to carry on an activity of the kind specified in Article 9B, prohibiting the issue of electronic money having a monetary value greater than the funds received.
The UK government chose this particular construction because it (rightly) thought that implementing (ii) in full would create a legislative loophole where institutions that issue tokens with a greater monetary value than the funds received in return would not, contrary to the EU’s intentions, be subject to the Directive’s regulatory framework. The government’s proposed solution involves a broader definition of ‘electronic money’, so as to bring potential issuers under FSA authority. Once these issuers are under the FSA regime, it is then left to the Handbook (ELM 4.4.1R) to prohibit the issuance of e-money having a monetary value greater than the funds received. 
Other parts of the Order make no modifications to the Directive’s source text and copy select passages verbatim. Article 9A states that ‘[a] sum is not a deposit for the purposes of article 5 [of the Order] if it is immediately exchanged for electronic money.’ This carries into English law a problem that was already considered, namely that of interpreting ‘immediately exchanged’. It has been suggested elsewhere (Long & Casanova 2003, p. 11)  that the problem is not so acute in England since Article 5 (2)(b) of the Order defines a deposit as a sum of money paid to an institution on terms which, amongst other things, ‘are not referable to the provision of property (other than currency) or services or the giving of security’. As long as the funds received are referable to a service (the issuing of e-money), it is further argued, they would not be seen as deposits. The Handbook’s AUTH App 3.2.16G supports this reading, but leaves the theoretical problem of interpreting the expression ‘immediately exchanged’ intact. It is suggested that this problem is best solved in a general way by drawing on the functionality test introduced in section 3, above. The Handbook offers a more limited solution by offering guidance in two specific cases only (AUTH App 3.2.18G and 3.2.19G), but it is reasonable to assume that other cases would be dealt with by applying the Handbook’s specific guidance by analogy.
Implementation of the Directive’s Article 8 waiver provisions is another interesting topic. The Order’s Article 9C allows so-called ‘small issuers’ to apply for a certificate which expressly excludes their business activities from being regulated as an Article 9B-‘specified kind of activity’. A certificate is to be granted if the electronic devices in question are restricted to a maximum storage capacity of EUR 150 and either (a) total business activities of the issuer do not normally exceed EUR 5 million and never exceed EUR 6 million, or (b) total business activities do not exceed EUR 10 million and the e-money issued is only accepted by the issuer’s subsidiaries or other members of the same group as the issuer, or (c) the e-money issued is only accepted by not more than one hundred persons (and these persons are within the same premises or in a limited local area – a shopping centre, an airport, a railway station, a university campus; or any area that does not exceed four square kilometres – or where the persons are in a close financial and business relationship with the issuer). Long & Casanova (2003, p. 14) suggest that it is likely that these (unwieldy) provisions will prove problematic. They consider the restrictions imposed under (c) to be arbitrary.
It is true that the restrictions are more meticulous than the Directive’s own. The Directive does not mention the one hundred-person limit, nor does it specify a ‘limited local area’ down to a size of four square kilometres. But at least the latter restriction will not have any impact besides offering guidance; for the Order expressly qualifies that the area specifications are ‘illustrative only and are not to be treated as limiting the scope of [(c)]’.  On the other hand, the restriction to one hundred persons is not in any way qualified and does not match the wider wording of the Directive. The same applies to the EUR 10 million ceiling in (b) which is not part of the Directive’s formulation either. Despite these deviations from the original text, it is suggested that this part of the Order does not pose the problems other scholars see. The Directive’s waiver provisions stipulate minimum restrictions that electronic money issuers must meet. The objective here is to avoid that Member State regimes are too lax. This reading is supported by the fact that, according to the first sentence of Article 7 (1), Member States may, implementing these reduced restrictions, allow their competent authorities to waive certain provisions. All of which suggests that Member States are not barred from imposing stricter conditions in exchange for a ‘small issuer’ certificate (i.e. a waiver from standard restrictions and requirements).
The UK’s implementation may seem complex – a result of the stringency and complexity of the underlying EU Directives. The US approach, by way of contrast, is refreshingly simple. It is simple because it is an approach not to regulate at all. There are presently no special restrictions pertaining to the issuing of e-money, and there are, so far, no government proposals for any future restrictions either. Some academics (Good 2000, chapter 6; Solomon 1997, chapter 5) raise the issue of risk and regulation, but policy makers including Chairman of the Federal Reserve Alan Greenspan  do not see any need for intervention. It is ironic that Europeans point to the so-called ‘free banking era’, a brief period in 19th century America where private banks issued their own dollar bills, where inflation was high and bank failures were common, in order to justify a strict e-money regime (Long & Casanova 2002, pp. 242 ff.). For American policy makers not only remain unimpressed; they unashamedly point to the very same era in order to justify today’s non-intervention. Drawing on recent historical research, they suggest that ‘the problems of free banking had little to do with banking’ (Greenspan 1997), and conclude that money regulation works best when government does not intervene. They argue for less regulation rather than more, and talk of ‘freeing up the system’ (Macintosh 1999) for novel technologies to flourish.
There are two qualifications to be made to the uncomplicated picture presented. The first qualification is that while the US have not enacted any specific electronic money regulation some electronic payment schemes may nevertheless fall under existing banking and/or payment service regulations. Since these regulations are enacted on the State level, as opposed to the Federal level, the question whether a given scheme is, or is not, subject to (any kind of State) regulation may have up to 50 different answers. The second qualification is that some existing Federal legislation has been specifically extended to cover e-money. There have been at least two published legal opinions pertaining to the topic – one by the Federal Deposit Insurance Company (‘FDIC’), and one by the Federal Reserve Board (‘FRB’). The FRB has proposed an application of Regulation E – which provides rules for the provision of certain ‘access devices’ such as debit cards; requires documentation in the form of regular receipts and account statements; and limits consumer liability – to smart card-based electronic money products.  According to this proposal, all cards with a maximum storage capacity of USD 100 or less are exempted from compliance with Regulation E. Cards with a higher storage limit are placed in a three-tier system, depending on whether they are online or offline (and, if they are offline, whether they are accountable or unaccountable  ). Offline unaccountable systems must provide initial disclosure, but, like sub-USD 100-cards, are otherwise exempted. Offline accountable and online systems are more stringently regulated. Issuers will be subject to liability limits and error resolution procedures, must make initial disclosures, and are to provide immediate receipts. The smart card’s current balance and a complete transaction history are to be made available on request. Meanwhile, the FDIC has issued a legal opinion  on the question of whether smart card-based e-money products are covered by deposit insurance. The gist of the opinion is that, in general, card systems will not be covered by FDIC deposit insurance. However, it is stated that insurance cover would be offered to solutions that are linked with money stored on a customer’s bank account (more on this in section 7, below).
Both the FDIC opinion and the FRB proposal are very reserved in tone and restricted in their application. And both of them focus on a limited field of technology (smart cards). This is in line with the prevailing non-interventionist stance in the United States. An obvious consequence of this fixation on a particular technology is that newer technologies, such as online payment systems and mobile phones, are not covered.
The American FDIC opinion and FRB proposal are based on electronic money’s ‘old paradigm’. The European Directive’s structural definition of e-money contrasts with this position. It is much more flexible, covering old and new technologies alike. Is this flexibility (within a stricter regime) an asset? It is tempting to assume that it is, but the FSA’s recent decision to declare PayPal (Europe) Ltd. an electronic money issuer may expose some of the problems a wider definition (and a broader regulation) of electronic money entails.
PayPal, in its own self-image, ‘enables any individual or business with an email address to securely, easily and quickly send and receive payments online. PayPal’s service builds on the existing financial infrastructure of bank accounts and credit cards’.  Users typically use a credit card to open a PayPal account and ‘load’ it with funds. These funds, which PayPal deposits in regular bank accounts  , may then be sent to virtually any valid email address in the world. Once the email’s recipient accepts the payment, her own PayPal account will be credited. If she does not yet have a PayPal account the email she receives will advise her on how to open a new account in order to accept the payment. Once the funds are on her account, they may be withdrawn from PayPal at any time.  The system operates simply and very successfully, but do PayPal funds constitute electronic money? PayPal (Europe) Ltd.’s User Agreement states at § 2.1 that
The Service is an e-money payment service rather than a banking or escrow service, and we are not acting as a trustee with respect to balances that you choose to keep in your account. We are acting only as an issuer of electronic money (stored value). 
This suggests that, in line with EU and UK legislation, PayPal funds constitute monetary value, as represented by a claim against the issuer (i.e. PayPal (Europe) Ltd.), which are stored on an electronic device (PayPal, Inc.’s servers, presumably), are issued on receipt of funds (through a credit card transaction initiated by the user), and are accepted as a means of payment by persons other than the issuer (other PayPal users).
But then PayPal [America], Inc.’s User Agreement for users outside of Europe states at a corresponding § 2.1 that
PayPal acts as a facilitator to help you accept payments from and make payments to third parties. We act as your agent based upon your direction and your requests to use our Services that require us to perform tasks on your behalf. PayPal will at all times hold your funds separate from its corporate funds, will not use your funds for its operating expenses or any other corporate purposes, and will not voluntarily make funds available to its creditors in the event of bankruptcy or for any other purpose. 
This is startling. PayPal [America], Inc. and PayPal (Europe) Ltd. both utilise the same underlying technology. Both offer the same kind of basic, account-based service. And yet the legal regime in Europe views PayPal (Europe) Ltd. as an electronic money institution while the American regime treats PayPal, Inc. as an ‘agent’ or ‘payments facilitator’. The European interpretation sees any transfer into and out of the PayPal system as a funds exchange – ‘real money’ is exchanged for ‘e-money’, and vice versa. The US interpretation, by way of contrast, views transfers as transparent – ‘real money’ is directly transferred into user accounts at third-party banks.  No funds exchange is involved. American users appear to remain the owners of any ‘real money’ in the system (arg: ‘your funds’).
How is such a different perception of the legal nature of a service possible? The European definition of electronic money cannot, on closer inspection, explain the two different interpretations. Recall the UK’s definition of e-money as
monetary value, as represented by a claim on the issuer, which is—
(i) stored on an electronic device;
(ii) issued on receipt of funds; and
(iii) accepted as a means of payment by persons other than the issuer;
‘Monetary value, as represented by a claim on the issuer’ cannot distinguish between divergent contractual approaches without giving up a technologically neutral reading of ‘value’. For both European and American User Agreements involve (some sort of, but not necessarily the same) legal claim against PayPal, and every legal claim may be treated as ‘monetary value’ in a loose sense of the term. Neither the Directive, nor the FSA’s Handbook has anything to say on this issue. ‘Issued on receipt of funds …’ cannot explain the differences between the European and American User Agreements either. For there is no criterion that can determine, in any pre- or non-legal way, whether an exchange has taken place or not. When transfers are electronic, there is no ontological difference between the funds ‘put into’ and the funds ‘taken out of’ the PayPal system. The difference can only be one of legal construction, but here, once again, no proper demarcation criterion is in place. Once again, the Directive is reticent; and so is the current edition of the Handbook.
A recent FSA consultation paper  – designed to offer perimeter guidance on electronic money systems – seeks to establish a demarcation criterion of sorts, stating that ‘[w]hen deciding whether a particular scheme involves issuing e-money or not, it is necessary to take into account the substance of the scheme’ (draft AUTH App 3.3.22G). Relevant factors in this assessment are to include the risks incurred by the bearer, the nature of the rights and obligations of the bearer, and what the scheme permits the bearer of the value to do (draft AUTH App 3.3.23G). The draft text adds, rather unhelpfully, that any ‘artificial features of the system that disguise, or try to disguise, the payment function’ will not likely prevent the scheme from being viewed as involving electronic money (draft AUTH App 3.3.24G).
The wording is not finalised yet, but the ‘guidance’, as it stands, leaves much to be desired. The question is not whether some features may disguise, or try to disguise, the ‘true nature’ of an e-money service. The question is how to ascertain the ‘nature’ of a payments system in the first place. What legal construction will bring payments systems within the scope of European regulation? What construction will allow systems to operate outside of the European regime? What constitutes ‘monetary value’? When are funds ‘exchanged’? The current ‘definitions’ offer no guidance. They would seem to allow for arbitrary FSA decisions.
But arbitrary or no, it is potential e-money issuers who must bear the costs of uncertainty (and actual issuers who must bear the certainty of a strict regime). Consumers, it is commonly assumed, can only benefit from such a system. The assumption is strong, given that the Directive’s Recitals 4, 9 and 10 mention bearer confidence as a central aspect of the European regime. But the assumption appears to be wrong.
Let us stay with the PayPal example. A comparison of PayPal [America], Inc.’s User Agreement with the corresponding agreement of PayPal (Europe) Ltd. shows that on at least three counts US citizens are offered better rights and protection than their European counterparts. A first, inconspicuous difference involves redeemability. In Europe there is a bottom limit on the amount of funds PayPal users may withdraw from their accounts. § 6 of PayPal (Europe) Ltd.’s User Agreement sets this threshold to GBP 6 and EUR 10, respectively. This is in line with the E-Money Directive’s Article 3 (3) on redeemability. No such threshold exists in the US under § 5.3 of PayPal, Inc.’s User Agreement – American users may withdraw as little as they like. A second difference involves the legal effects of a potential PayPal bankruptcy. If PayPal (Europe) Ltd. were to go bankrupt, users would retain their (worthless) e-money tokens. All legal claims would have to be made against the bankrupt’s estate. § 6 of PayPal (Europe) Ltd.’s User Agreement expressly warns of this possibility, stating
Please be aware that your account balance represents an unsecured debt of PayPal (Europe) Ltd. to you, is at risk in the event of PayPal (Europe) Ltd.'s insolvency and is not covered by the Financial Services Compensation Scheme or any other public or private insurance scheme. 
If PayPal [America], Inc. were to go bankrupt however, users funds would most probably remain outside of the bankrupt’s estate and be protected in full.  This is because the American User Agreement does not include any contractual exchange of ‘real money’ for e-money. A third difference flows directly from the second. If the banks that partner with PayPal (Europe) Ltd. were to go bankrupt, only PayPal (Europe) Ltd. would have claims against the bankrupt’s estate. Only PayPal (Europe) Ltd. would be eligible for deposit insurance – if PayPal (Europe) Ltd., as a legal entity, is eligible for deposit insurance at all.  If the banks that partner with PayPal [America], Inc. were to go bankrupt however, users who hold funds in USD with the banks in question would be eligible for FDIC pass-through insurance up to a value of USD 100,000. 
All three examples go to show that a strict regime does not automatically offer better legal protection to customers. Under the current e-money regime, European PayPal customers are actually worse off than their American counterparts.
Three main criticisms of current e-money regulation in Europe emerge from a comparison with the legal set-up in America.
(1) Regulation is too strict on issuers – section 3 argued that strong own funds requirements, limited investment opportunities, and heavy supervision place a serious burden on electronic money institutions. Financial incentives for new ‘start-ups’ that wish to operate on a Union-wide basis are lacking.
(2) Current e-money regulation is too vague in its definitions – section 6 showed that the EU and UK statutory definitions of electronic money are not able to determine whether technology, such as PayPal’s, constitutes e-money or not. In the absence of court rulings, demarcation is left to Member State authorities. These authorities have yet to develop objective, dependable criteria for their decisions. It should be possible to devise new electronic payment technologies that do not fall under the Directive, yet do not, at the same time, constitute an unlawful ‘evasion’ of the Directive.
(3) And while the European regime may have succeeded in creating bearer confidence, it still lacks adequate bearer protection – the analysis in section 7 showed that a regime that avoids excessive regulation and ‘piggy-backs’ on existing institutions (such as FDIC-protection in America) may offer a better framework for all parties involved.
The European Commission is to present a report on the application of the E-Money Directive in a year’s time. In line with the Directive’s Article 11 revisions may be proposed. If the comparative analysis in this article is correct the current EU legislation faces some serious problems. It should be reviewed and revised with care.
Brindle M & Cox R (eds.) (1999) Law of Bank Payments, Second Edition (London: Sweet & Maxwell).
Good B (2000) The Changing Face of Money: Will Electronic Money be Adopted in the United States? (New York and London: Garland Publishing, Inc.).
Solomon E (1997) Virtual Money: Understanding the Power and Risks of Money’s High-Speed Journey into Electronic Space (New York: Oxford University Press).
Bamodu G (2003) ‘The Regulation of Electronic Money Institutions in the United Kingdom’, The Journal of Information, Law and Technology (JILT).
Batalla E (2001) ‘Electronic Commerce’, Computer and Telecommunications Law Rev 80.
Chuah J (2000) ‘The New EU Directives to Regulate Electronic Money Institutions – A Critique of the EU’s Approach to Electronic Money’, Journal of International Banking Law 180.
Long W & Casanova J (2002) ‘European Initiatives for Online Financial Services, Part 1: The Regulation of Electronic Money’, Butterworths Journal of International Banking and Financial Law 242.
––(2003) ‘European Initiatives for Online Financial Services, Part 2: Financial Services and the Regulation of Electronic Money’, Butterworths Journal of International Banking and Financial Law 8.
Macintosh K (1999) ‘The New Money’, Berkeley Technology Law Journal 652.
Robertson P (1999) ‘Internet Payments’, in Brindle M & Cox R (eds.) (1999) Law of Bank Payments, Second Edition (London: Sweet & Maxwell).
Vereecken M (2000) ‘Electronic Money: EU Legislative Framework’, European Business Law Review 420.
3. Statutes & Statutory Instruments
Financial Services and Markets Act 2000
Electronic Money (Miscellaneous Amendments) Regulations 2002 (SI 2002/765)
Financial Services and Markets Act 2000 (Regulated Activities) Order 2001 (SI 2001/544)
Financial Services and Markets Act 2000 (Regulated Activities) (Amendment) Order 2002 (SI 2002/682)
Directive relating to the taking up and pursuit of the business of credit institutions (2000/12/EC)
Directive amending Directive 2000/12/EC relating to the taking up and pursuit of the business of credit institutions (2000/28/EC)
Directive on the taking up, pursuit of and prudential supervision of the business of electronic money institutions (2000/46/EC)
Federal Deposit Insurance Company (1996) ‘General Counsel’s Opinion No. 8 – Stored Value Cards’, Federal Register 40,490.
Federal Reserve Board of Governors (1996) ‘Official Staff Commentary on Regulation E Electronic Funds Transfer’, Federal Register 19,696.
Financial Services Authority (2002) ‘CP 117: The Regulation of Electronic Money Issuers’
Financial Services Authority (2003) ‘CP 172: Electronic Money: Perimeter Guidance’
Financial Services Authority, Handbook on Rules and Guidance
HM Treasury (2001) ‘Implementation of the Electronic Money Directive: A Consultation Document’ < http://www.hm-treasury.gov.uk/mediastore/otherfiles/e_money.pdf >
HM Treasury (2002) ‘Implementation of the Electronic Money Directive: A Response to Consultation’ < http://www.hm-treasury.gov.uk/media//49D42/emoney_response.pdf >
6. Online Articles (last visited April 4, 2004)
Greenspan A (1997) ‘Regulating Electronic Money’
Levy S (1994) ‘E-Money (That’s What I Want)’
Steinert-Threlkeld T (1996) ‘The Buck Starts Here’
7. Links (last visited April 4, 2004)
* I would like to thank the two anonymous reviewers for their very helpful criticism of a previous draft of this paper. I am grateful also to Mr Andrew Murray from the London School of Economics for his encouragement.
 Long & Casanova (2002, pp. 245-7) provide a detailed account of the Directive’s genesis.
 Vereecken (2000, p. 419) offers a cynical explanation of the paradoxical integration of electronic money institutions into Directive 2000/12/EC. He suggests that the European Central Bank, operating from behind the scenes, wanted full control of the reserve holdings of electronic money institutions; this necessitated their inclusion within the definition of the latter Directive.
 Awkward: as will be seen later on, one of the things type (b) credit institutions are not allowed to do is actually extend credit. This is confusing and unnecessarily complicates matters.
 Despite the Directive’s technologically neutral definition of e-money in Article 1 (3)(b), Article 1 (5)(b) seems to have been devised with mainly offline, smart card based systems in mind.
 The details are quite complex. The two per cent are calculated from (a) current financial liabilities or (b) average liabilities of the preceding six months of business operation. Special rules exist for companies in their first six months in business. See Article 4 (2) and (3) in particular.
 Article 5 (1) and (a). A number of tightly regulated exceptions exist; see Article 5 (1)(b) and (c), (2) and (3).
 Ibid., at paragraph (5).
 Chuah (2000, p. 184) rightly notes that this approach ‘appears incongruous with the EU’s push for interoperability in payments systems’.
 See also Vereecken (2000, p. 417: ‘2. Objectives of the E-money Directive’).
 ELM 4.4.2G and 4.4.3G offer (non-binding) guidance on circumstances where this prohibition might not apply. The two main conditions demand that all e-money with a higher monetary value than the funds received be (a) part of a promotion and (b) always covered by an appropriately enlarged float.
 For a general overview on the issue of deposits see also Chuah (2000, especially p. 183).
 Financial Services and Markets Act (Regulated Activities) Order 2001, Article 9C (7). See also ELM 8.4.12G.
 Greenspan <http://www.cato.org/pubs/policy_report/cpr-19n2-1.html> writes: ‘I am specifically concerned that we not attempt to impede unduly our newest innovation, electronic money, or more generally, our increasingly broad electronic payments system’.
 Federal Reserve Board of Governors (1996, pp. 19,696 ff.). See also Good (2000, at pp. 88 ff.).
 ‘Accountable’ here means that a purchase is linked to an individual bearer; the transaction is not anonymous.
 Federal Deposit Insurance Company (1996, pp. 40,490 ff.).
 American users may place their funds into PayPal, Inc.’s money market fund also. But this option is not relevant for our further discussion.
 Certain restrictions apply but do not impact the main point of the argument.
 See note 18, above, for a qualification.
 This claim is based on PayPal Inc.’s own legal assessment. See <http://www.paypal.com/cgi-bin/webscr?cmd=p/gen/fdic-outside >
 The question whether or not PayPal (Europe) Ltd. is protected under any kind of deposit insurance critically depends on the applicable jurisdiction (third party banks!) and cannot be answered in a general manner.
 See <http://www.paypal.com/cgi-bin/webscr?cmd=p/gen/fdic-outside >This protection is not available to users who decide to invest their funds in PayPal, Inc.’s money market fund.