Alberto Franco Pozzolo

I would like to thank Vincenzo de Stasio, Maria Iride Vangelisti and participants at the Conference on “L’attuazione della seconda direttiva sui servizi di pagamento e ‘open banking’”, Bergamo, 18-19 October 2019 for comments and suggestions. All remaining errors are my own responsibility.

Introduction

The provision of payment services is one of the most important activities performed by the financial sector.^[1]See De Bonis, R. and Vangelisti, M.I. (2019), La moneta – Dai buoi di Omero ai bitcoin, il Mulino, for a comprehensive account of the evolution of the payment system. Traditionally, most of these services have been offered by banks, although in the last decades non-bank financial institutions have gained a prominent role in some specific segments, for example in the credit card business. In the past, despite being a relevant source of revenues for banks and other financial intermediaries, payment services provision has ranked relatively low in the interests of policy makers and academics alike, with most of the attention on the soundness of the system rather than on its efficiency.

In recent years, this landscape has changed dramatically. The developments in information and communication technologies have made it possible to perform traditional activities more efficiently and to devise innovative services which were not available before. New players have entered the market, in some cases increasing the degree of competition to the benefit of customers, in other cases exploiting network economies of scale and their better and exclusive know-how to gain significant market power.

Policy makers have reacted to the innovation in the payment industry. In Europe, the first Payment Services Directive, ^[2]Directive 2007/64/EC of the European Parliament and of the Council of 13 November 2007 on payment services in the internal market, amending Directives 97/7/EC, 2002/65/EC, 2005/60/EC and 2006/48/EC … Continue reading published in 2007, has set the scene, defining the rights and obligations for a broad spectrum of payment services, such as credit transfers, direct debits and card payments. Moreover, it has defined compulsory information requirements, especially on costs. But the market has evolved rapidly, and yet in 2013 the European Commission has published the proposal of a new Directive on payment services, which was eventually published in 2015 and is widely known as Payment System Directive 2 (PSD2).^[3]Directive 2015/2366/EU of the European Parliament and of the Council of 25 November 2015 on payment services in the internal market, amending Directives 2002/65/EC, 2009/110/EC and 2013/36/EU and … Continue reading

The PSD2 intervenes extensively in the regulation of many aspects of the payment industry, defining new rights and obligations. One of its aims is to foster competition among service providers, to cut monopoly rents, reduce costs, and improve efficiency.^[4]For a thorough comparison of PSD1 and PSD2, see Sciarrone Alibrandi, A. (2020), Impostazione sistematica della PSD2, in Paglietti, C. and Vangelisti, M.I. (eds.), Innovazioni e regole nei pagamenti … Continue reading The adoption of the Directive in the various European countries, and the definition of regulations related to its implementation, has taken some time. For this reason, its impact on the industry has not yet fully shown. Nonetheless, while some effects are becoming visible, stronger ones are certainly to be expected in the coming years.

The purpose of this chapter is to discuss how some of the innovations introduced by the PSD2 might impact the business model of payment service providers in coming years, what the effect on the equilibrium in the industry will be, what new risks may emerge, and what actions should be taken to counter possibly undesired outcomes.^[5]See Zeno-Zencovich, V. (2020), Prefazione, in Paglietti, C. and Vangelisti, M.I. (eds.), Innovazioni e regole nei pagamenti digitali: il bilanciamento degli interessi nella PSD2, RomaTre Press, for … Continue reading The Directive has introduced numerous innovations, but the analysis in this chapter will focus mainly on the potential impact of the introduction of three new types of intermediaries, known as Third Party Providers (TPPs): Payment Initiation Service Providers (PISP), Account Information Service Providers (AISP) and Card-Based Payment Instrument Issuers (CBPII).

The rest of the chapter is organized as follows: the next Section briefly describes the activities of the new types of intermediaries that have been introduced by the PSD2. Section 3 describes the main sources of market power in the payment industry. Section 4, the core of the chapter, discusses the possible impact of the PSD2 on the degree of competition in payments. Section 5 argues in favor of some policy interventions and concludes.

The new players at the heart of PSD2

The main impact of PSD2 on the market for payment services will most likely be caused by the introduction of three new types of players: Payment Initiation Service Providers (PISP), Account Information Service Providers (AISP) and Card-Based Payment Instrument Issuers (CBPII).^[6]As is well explained by Porta (2019), Obiettivi e strumenti della PSD2, in Maimeiri, M. and Mancini, M. (eds.), Le nuove frontiere dei servizi bancari e di pagamento fra PSD 2, criptovalute e … Continue reading

With the burgeoning diffusion of online purchases, PISPs are expected to become important players in the payment procedure, because they link the website of the merchant with the banking platform of the payer. Namely, they are responsible of guaranteeing the payee that the payment has been initiated, so that the merchant can safely transfer the good or service purchased, and the operation can be concluded. Aside from crucial issues related to consumer protection and security, the PSD2 requires that any payment service provider is permitted to offer payment initiation services, therefore guaranteeing fair competition in the market. To achieve this goal, the Directive posits that PISPs only need the consent of the account holder to access his banking platform, without being required to use any particular business model, or entering into any contractual relationship with the payer’s bank. This requires that common and open standards of communication be implemented, according to guidelines that are provided by the European Banking Authority (EBA). In practice, financial intermediaries holding customers’ accounts are mandated to make them easily accessible through Application Program Interfaces (APIs). Importantly, PISPs cannot hold accounts in the name of their customers, but their activity is limited to initiating the provision of payment services.

AISPs cover a different part of the market. The increasing number of relationships that corporations and individuals have with banks and other financial intermediaries allows a better management of liquidity and payment services. However, this also makes a unified picture of the position across all different accounts difficult to obtain, which can lead to organization and synchronization problems. AISPs provide aggregated online information on one or more payment accounts held with one or more other payment service providers. Since better information helps fostering competition, the PSD2 has made it easier for AISPs to access customers’ accounts, along similar lines as those defined for PISPs. Indeed, the only difference is that AISPs are not allowed to initiate a payment operation, but they can only gather and provide information to those entities that the account holder has authorized. To this purpose, similar common and open standards of communication are to be implemented. Obviously, like PISPs, AISPs must observe specific data protection and security requirements.

Finally, the PSD2 also allows financial intermediaries to issue card-based payment instruments, such as credit-cards and debit-cards, that directly debit a holder’s account held with a different financial intermediary. This allows to break the connection between the issuer of the card and the bank where the account that the card debits is held. As with PISPs and AISPs, holders of customers’ accounts are mandated to make them easily accessible to CBPIIs through standardized APIs.^[7]For a thorough legal analysis of PISPs, AISPs and CBPIIs see also: Profeta, P. (2019), I third party provider, profile soggettivi e oggettivi, in Maimeiri, M. and Mancini, M. (eds.), Le nuove … Continue reading

To assess how the entry of these new players will change the payment industry, first we need to understand what the main levers of competition in this market are, which is the topic of the following section.

Competition in the payment industry

In the past, competition in the market for payments has been low, allowing some players to earn significant extra profits at the expense of their customers. Three main factors favor the emergence of monopoly rents: the possibility to bundle payments with other (banking) services, such as current account services; the presence of large network externalities, due to the fact that the incentive to adopt a new means of payment increases with the number of other individuals adopting it; and the importance of reputation, that hinders the entry of new players in the market. To better understand the role of each one of these three factors in harming competition, it is useful to analyze the steps involved in a typical payment.^[8]On this issue, see also Geerling, M. (2018), E-commerce: A merchant’s perspective on innovative solutions in payments, in Journal of Payments Strategy & Systems 12, pp. 58-67.On this issue, see … Continue reading

Payments are normally the counterpart of a purchase of goods or services, that can take place either physically, for example in a shop, or through the internet. In the case of online purchases of goods, and some services such as travel tickets and hotel rents, the consumption is deferred from the moment of payment.^[9]For an analysis of the legal implications of payment services, see De Stasio, V. (2016), Ordine di pagamento non autorizzato e restituzione della moneta, Giuffré, Milano. However, an increasing number of services are accessed and transferred through the internet, such as video and music files, but also online teaching or consulting services.

From the point of view of its payment, and leaving cash aside, each purchase involves four main steps. First, the customer enters a physical brick and mortar or a virtual shop and chooses the good or service that he wants to buy from the merchant. Second, he transfers an accepted mean of payment from his account to the hands of a financial intermediary providing the payment service (Payment Service Provider, PSP). Third, the PSP transfers the mean of payment to the account of the merchant. Fourth, the merchant delivers the good or service acquired, and the purchase is finalized. Although often some steps are even further segmented, as in the case of credit-card payments, we can focus on this simplified framework to understand the determinants of competition in the market, without much loss of generality.

The three factors favoring monopoly rents mentioned above can intervene at different stages in the process described above. Consider the first, the possibility of bundling payments with other services. An obvious case is when the PSP coincides with the financial intermediary either of the customer or of the merchant, which allows to reduce the number of intermediaries involved in the process and, in turn, the cost of the transaction. Financial intermediaries able to offer both payments and deposit services can thus benefit from the economies of scope of carrying both activities. This increases the optimal size of operation, making it more difficult for new players specialized only in payment services to enter the market. Eventually, a merchant could even be in the position to agree with his bank to force customers to use a specific PSP, making it impossible for him to choose an alternative. In fact, this is happening increasingly often in the case of some e-commerce websites.

The second factor favoring the emergence of monopoly rents in the payment industry is the presence of network externalities. As already argued above, reducing the number of intermediaries that take part in the payment process is likely to reduce the costs. In principle, if all customers and all merchants held their accounts with the same financial intermediary, that would be the only PSP that is needed, because all payments would involve just the record of a transfer from the account of the customer to that of the merchant. While this is an unlikely outcome in practice, since it would imply the existence of a single monopolistic bank, it remains true that financial intermediaries with a larger number of clients are more likely to be able to settle payments within their own accounts, and therefore to lower costs. This is a clear case of existence of strong network economies, which favor market concentration.

The third factor, reputation, relates to a key characteristic required of any payment process: that it is secure, i.e. that the transfer is certain. A customer transferring a given amount of money to a PSP wants to be sure that it will then be transferred to the account of the merchant, so that he delivers the good or service that the customer has acquired, and the purchase is finalized. Although PSPs are subject to regulations and supervision, in principle their position allows them to divert the money received, eventually robbing the customer. As a less fraudulent alternative, PSPs can economize on their investment in information technologies, reducing their standards of security and therefore increasing the probability that the information on the payment is lost. This makes it harder for an unknown financial intermediary to enter the market for the provision of payment services, because customers are unlikely to trust transferring their money in his hands. PSPs need therefore to gain a reputation for being reliable, something that typically requires time and is often associated with size. Again, this limits entry in the market for payments, favoring concentration and monopoly rents. In fact, customers in many countries seem to have a preference for making their payments through traditional banks, possibly because of their better reputation.

PSD2 and competition in the payment sector

Judging from the picture described above, PSD2 will foster competition by favoring the entry of new players in the second and third phase of the payment process: the transfer from the account of the buyer to the PSP and that from the PSP to the account of the merchant. By allowing PISPs to directly initiate a payment process and AISPs to rationalize the management of payments from different accounts, PSD2 aims at favoring the entry of players which may be able to innovate the existing services, exploit better technological know-how, accept lower profit margins – offering better and cheaper services overall. Reassuringly, PSD2 also requires PSP to be subject to a set of regulations aimed at guaranteeing the security of the services that they offer. Will this work?

A payment is very similar to what economists define as a commodity good, because its characteristics make it easily exchangeable across different suppliers. A key characteristic of competition in markets for commodity goods is that it is only based on prices. If a purchase can be paid with two different payment cards, and one has a lower cost, a rational economic agent would never use the other card. However, this analysis is correct only if we focus exclusively on the payment, i.e., on the activity of transferring a mean of payment from the account of the buyer to that of the seller. In fact, without taking into account riskiness, payments can be extremely different depending on at least two additional characteristics: how easy it is to make them and how quick the transfer is. Overall, entrants can gain market shares if they are able to provide a payment experience that improves on the existing one along one of four dimensions: because it is cheaper, it is easier, it is quicker, or it is less risky. On which one of these characteristics are entrants more likely to have a competitive hedge?

Two main reasons suggest that price competition is likely to be relevant only on the merchant side of the market. First, the structure of interchange fees is such that the costs of a payment is mainly born by merchants. In fact, in most countries, the cost of a good or service cannot be different depending on the mean of payment that it used to pay for its purchase.^[10]There are many reasons why this is the case. From an economic perspective, the literature on two-sided markets has shown that charging merchants is in many cases the profit-maximizing strategy of … Continue reading Second, the cost of the single payments is already relatively low, and it is unlikely that possible reductions allowed by better technologies will be so large to motivate consumers to switch from one PSP to another: it is therefore the merchant that may have an incentive to push for the adoption of cheaper means of payment.

A price-based competitive advantage is therefore more likely to emerge if a PSP has a better ability to process payments. A crucial example is that of closed loop systems, i.e., “systems set up and operated by a single payment service provider”, in which consumers and merchants hold an account with the same financial intermediary. Examples of these systems are US-based PayPal, Chinese Alipay, and Italian Satispay. Within a closed loop system, a payment entails only two opposite records: in the accounts of the consumer and of the merchant. Consequently, it does not require the transfer of information across different accounting systems, as in the case of a transfer of funds from the account of one bank to another. The competitive advantage of closed loop systems is thus the result of their ability to process payments at lower costs, which may give them a competitive advantage with respect to more traditional processes involving banks or credit card systems. Clearly, the incentive for merchants to hold an account with a closed loop system is increasing in by the number of consumers using it as a mean of payment, and symmetrically the incentive for consumers is increasing by the number of merchants that offer it as a mean of payment. This is a major source of network economies: the larger the number of users, the lower the costs.^[11]In addition, closed loop systems may be in the position of earning an interest rate margin, if they are allowed to invest the funds of their customers at higher rates than those that they pay on the … Continue reading 

Remarkably, PSD2 excludes closed loop system from the obligation to allow access to AISPs, PISPs and CBIIPs through standardized APIs, thus giving them a significant regulatory advantage with respect to other PSP. I will come back to this issue below.

As already argued above, risk can be an important factor influencing the choice of the PSP. However, in all developed countries payment institutions are heavily regulated and supervised along all crucial dimensions: technological, organizational and financial.^[12]European Banking Authority (EBA) has been required to define the technical standards of secure communication between Third Party Providers (TPPs) and the financial intermediaries where the accounts … Continue reading For example, financial intermediaries issuing electronic money, as well as PISPs and CBPIIs, cannot accept deposits from users and can only use funds received from users for payment services. Moreover, they must respect initial capital requirements. Since regulation creates a level playing field, it is unlikely that riskiness may become a crucial competitive factor in the supply of payments services within the same country. Riskiness can indeed be a crucial characteristic in the case of cross-border payments, but it is likely that once a given threshold of safety is guaranteed, competition will be on other features.

Aside from prices and riskiness, competition will most likely be based on how easy and quick it is to make a payment, limiting the number of steps required to make it. Let’s consider in-shop purchases first. With cash payments, the transfer is typically very swift, with the only nuisance of requiring to hold the right amount of change. With electronic payments, readiness has increased significantly in recent years. In the past, all electronic payments required to insert a plastic card into a device that connected remotely, typically via a phone call, and printed a receipt that had to be signed in order to authenticate the transaction. The process could easily last around one minute. Some improvements have initially come with the substitution of the signature authentication with that based on the personal identification number (PIN). More recently, contactless payments of small amount, with no requirement of authentication, have made the payment process as swift as that of cash payments, and without the problem of having the right amount of change. Contactless payments using a smartphone with a biological identification device can be as swift as that with a card, but with the advantage of being potentially less risky.

Only cheaper and better technologies seem capable of offering a better payment experience in the case of in-shop purchases. By using a smartphone it is possible to transmit a large amount of data, thus allowing to perform two crucial steps of an in-shop payment: authentication and transfer. The smartphone industry is very concentrated: four corporations (Samsung, Huawei, Apple and Oppo) are responsible of half of total world production of devices, and only two operating systems are used in practice: Android and iOS. But these corporations do not offer payment services autonomously, preferring to partner with different financial intermediaries. Since producers of smartphones do not have incentives to make an exclusive partnership with just one provider of financial services, technological innovation stemming from hardware improvements is unlikely to be a key competitive factor for payment service providers, and the new players introduced by PSD2 are unlikely to have a major role.

A different case is that of software improvements. Although only two operating systems are used in the market, competition could stem from the ability of payment service providers to interface with these systems, making their apps swifter and more user-friendly. By allowing PISPs and CBPIIs to enter this sector, PSD2 may alter competition dynamics also in the case in-shop payments.^[13]A different issue, that is also part of PSD2 but is not analyzed in this chapter, is the reimbursement of fraudulent transactions. Better authentication systems, leading to a lower number of frauds … Continue reading

In the case of online purchases, payments can at times be rather cumbersome: customers may be required to type their cards’ codes: the PIN, and the one time password (OTP) that they receive through different means of communication – typically small text messages (SMS) or, less commonly, e-mails. Completing these tasks may require a few minutes and can be rather complicated, with the risk that a purchase is not completed or the payment is made to the wrong counterpart. It is therefore no surprise that new services are being provided, allowing to register in smartphones or computers a set of information that make electronic payments easier and quicker.

In the case of internet-based transactions, purchases and payments are much more connected, and there is no reasonable alternative to electronic payments.^[14]Some goods can still be paid in cash to the delivery man, and services such as hotel rents booked in advance on the internet can be paid when checking-out with different means of payment from those … Continue reading Internet-based merchants have a stronger incentive to offer a better payment experience, because the cost of switching to a different merchant is much smaller on the web than across traditional brick and mortar shops. Internet based merchants are thus among those that could benefit the most from becoming a PISP, because in this way they could link the purchase with the payment, offering the best possible customer experience. Bundling purchases and payments can therefore be more effective than in the case of brick and mortar shops. Clearly, since setting up a PISP involves relevant fixed and sunk costs, big-techs like Amazon, which can dilute these costs across their large consumer base, may be more likely to exploit the possibilities offered by PSP2.

However, the previous discussion on cost advantages of closed-loop systems suggests that setting up a PISP may not be the best solution for e-commerce platforms. In fact, PayPal itself was initially created to make swifter payments on E-bay, the website specialized in consumer-to-consumer and business-to-consumer sales, and Alipay is linked to Alibaba, the largest e-commerce platform in China. PayPal and Alipay are closely linked to the e-commerce platform and are thus able to offer an easy and swift payment experience. But they are also closed loop systems, thus benefitting from lower costs and the potential to earn positive interest margins. Big-techs like Amazon may therefore prefer to set up their own closed loop system instead of creating a PISP, as Alibaba has already done with Alipay.

As mentioned above, PSD2 favors closed loop systems, arguing that “it would not be appropriate to grant third parties access to those closed proprietary payment systems”. Allegedly this choice is made “in order to stimulate the competition that can be provided by such closed payment systems to established mainstream payment systems”. But will the entry of these players foster competition in the payment industry? Since big-techs enjoy massive network economies of scale, it may indeed be the case that their closed loop systems allow swifter and easier payments at lower costs. At the same time, the optimal minimum scale of operations may be so large to allow for very few players in the market, thus hindering competition. As it is always the case with natural monopolies, finding an equilibrium between competition and efficiency is not easy. But it may be worth to consider granting third parties access to closed loop systems, since they already benefit from large network economies.

One additional argument stands in favor of not granting privileges to big-techs operating in the payment industry, and possibly limiting their expansion: the value of the information that they can collect. So far, our discussion has focused mainly on making a payment, i.e., on PISP and possibly on CBIIP. But with the explicit authorization of customers, these new types of intermediaries, as well as AISPs, can collect, store and elaborate information on each single payment for a potentially enormous number of consumers all over the world. If payments are managed by an in-house PISP or closed-loop system of an e-commerce platform, each payment can easily be linked with the purchase that caused it. Clearly, such information has an enormous value, for example for marketing and credit-scoring purposes. In Europe, the General Data Protection Regulation (GDPR) seems to be the only boundary to potential abuses in this field. Additional efforts in this direction seems warranted.

Conclusions

With few exceptions, I am generally in favor of any policy aiming at improving competition, because it increases aggregate economic welfare. In most cases, helping entrants in a market improves competition, and therefore I see it as a positive step. But markets in which bundling and network economies cause strong economies of scale is one of the exceptions to the general rule. Entry of large and powerful players can in this case reduce competition, with a negative impact on aggregate welfare.

The direction taken by the PSD2 is correct: favoring the entry of new players, reducing the benefits of holding proprietary information, dismantling the bundling of different services related to payments will improve competition and increase welfare. But attention must be paid to two possible risks.

The first risk is that big-techs exploit economies of scale related to network effects and the possibility of bundling different services to gain large market shares and hinder competition. Allowing third parties to access closed loop systems may be an important step to avoid this outcome.

The second risk is related to data collection and management. As emphasized by Gammaldi and Iacomini (2019),^[15]Gammaldi, D. and Iacomini, C. (2019), Mutamenti del mercato dopo la PSD2, in Maimeiri, M. and Mancini, M. (eds.), Le nuove frontiere dei servizi bancari e di pagamento fra PSD 2, criptovalute e … Continue reading “[t]he Directive indeed introduces the concept according to which data are available to the customer who ‘generated’ them”. This allows all players in the payment industry – including big-techs, if they are willing to enter this market – to collect information on payments linked with those on purchases, in addition to what they already collect, for example through social media. The use of information credit risk evaluation and price discrimination can have a disruptive effect on the financial market. Moreover, individuals may misperceive the immediate benefits that they obtain by releasing their data, and the often uncertain future costs related to their use on the part of specialized operators. In this context, “the explicit consent (freely expressed by users) may no longer be sufficient to guarantee respect for privacy”, as argued by Menzella (2019).^[16]Menzella, R. (2019), Il ruolo dei big data e il mobile payment, in Maimeiri, M. and Mancini, M. (eds.), Le nuove frontiere dei servizi bancari e di pagamento fra PSD 2, criptovalute e rivoluzione … Continue reading 

The risks caused by data mismanagement should be forcefully contrasted, within and outside the financial industry. For this reason, each specific piece of legislation should coordinate with the provisions of the GDPR.^[17]Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of … Continue reading Moreover, to guarantee all players using individual data and derived elaborations a level playing field, it should be considered to allow third parties to access the information collected by big-techs, according to the same philosophy which led the regulator to permit AISPs to access information collected by financial intermediaries, when users gave their consent. Of course, any such policy should be mirrored by initiatives aiming at making citizens aware of the value of their personal information – with a specific focus on financially-related information – for example offering “privacy education” in addition to “financial education”.

Introduction

In the last years, after decades of lethargy, the retail payment industry has started to thrive. Means of payment which were unthinkable of in the recent past are now available, and innovative services are offered by old and new players. The revolution in information and telecommunication technologies has taken some time to shake what was a very dull and traditional business activity. But with fintech, it has finally arrived.

This revolution has hit all segments of the industry. Not only transferring value has become as easy and cheap as it has never been before, but also what is behind this value has changed substantially.

The arrival of new and disruptive technologies has caused some mayhem in the market. The possibility to offer innovative services has created new profit opportunities, and incumbents in the payment industry are now facing unprecedented threats by aggressive entrants with better technological know-how. As it is always the case in periods of turmoil, it is not necessarily true that innovation leads to welfare improvements. New technologies allow to offer better services, but they may also favor the emergence of monopoly positions, due for example to network economies of scale and the possibility to bundle the supply of goods and services. New types of risks may emerge, that are not fully appreciated by consumers, therefore biasing their choices with respect to what would be socially optimal.

Technology adoption displays very strong path-dependence. Players that will manage to build a strong position in the industry in the coming years will very likely shape its evolution in the future. For this reason, it is of paramount importance to act now to assess the benefits and risks of the new opportunities, and tackle the possible negative externalities with adequate policy interventions.

The environment

If we think about how retail payments are generally made, we come out with a handful of alternative options. The most traditional way, and the most common in the majority of countries, is with cash. But debit- and credit-card payments are becoming more and more common. In some countries, smart-phones are supplanting or have already supplanted plastic cards as a support containing the information required to make a payment, and they may also be offering different types of services, not necessarily fully transparent at the eyes of end-users. Finally, we can directly make an order to the financial intermediary where we store our assets to transfer some of them on our behalf to another account, through cheques or money transfers.

In synthesis, cash, debit- and credit-cards, and bank orders make the vast majority of retail payments, even if some of them go through our smartphone. But what are we transferring when we use these means of payment? Cash is fiat-money, a liability of the central bank that we may call central-bank-money. Although it is technically a claim on the central bank, it is outside money and its real value is a function of the credibility of the central bank.

When we do not pay with cash, we are very likely to be transferring bank-money. Obviously, this is the case when we use a cheque or we make a bank transfer. But it is also very often the case when we use a debit- or a credit-card. Bank-money is a liability of a commercial bank, which is one-to-one convertible in cash. It is inside money, whose value is formally guaranteed by the assets of the bank – and by the ample set of regulations that assure that these assets are worth at least as much as the amount of bank-money issued by the bank, i.e. its deposits.

But since recently, when we make a payment with a card or using our smartphone we might also be transferring electronic money. This is a liability of a financial intermediary different from a bank. It can be outside money, as bitcoins, or inside money, as the majority of electronic moneys issued by financial intermediaries around the world. Often, electronic money is also convertible one-to-one into cash, but this is not always the case. Remarkably, from the point of view of a customer, the using experience of paying with a card of a bank is identical to that of paying with a card of a financial intermediary issuing electronic money, although what is behind the value that is transferred can be very different.

The choice among the different alternatives

How do we choose among different means of payment? A rationale choice is based on the characteristics of each instrument in terms of its cost of usage, its effectiveness, and its riskiness. The cost of usage and the effectiveness are technological characteristics. Apart when we use cash, making a payment involves the transfer of a certain amount of information. This can be done on paper, as in the case of a cheque, or of the old payments with the credit-card, when a paper slip was produced and sent to the interested counterparties. But in all other cases it is done electronically. More efficient information procedures to match the two sides of a transaction, higher computing power, more user-friendly procedures or interfaces to link purchases and payments are all characteristics that make it more likely that a given instrument is less expensive and more effective.

The riskiness of a means of payment depends instead on a number of different features. A crucial feature is the value that is promised by the issuer. In the case of central-bank-money such as cash, the real value is guaranteed by the central bank and by how well it maintains price stability. Crucially, price stability is expressed in terms of the legal tender, that undertakes the function of unit of account. In the case of bank-money, perfect convertibility with cash is assured by contractual and normative provisions. The nominal value is therefore guaranteed and the real value is the same as that of cash. An identical case is that of electronic-money, which is one-to-one convertible into cash or into bank-money. On the contrary, holders of electronic-money whose value is linked to that of a pool of assets – as for example in the case of the Libra, the currency proposed by Facebook – incur the risk of imperfect convertibility with cash or bank-money. Of course, whether the real value of such an electronic-money is more or less stable than that of central-bank-money depends on the conduct of monetary policy, on money supply by the electronic-money issuer and on the relevant rate of inflation for each consumer. In fact, the real value of an electronic-currency may turn out to be more stable than that of the legal tender, if the central bank is not good at maintaining price stability. The same, at least in theory, could happen with outside electronic-money, such as bitcoins.

A related issue is how this value is guaranteed. In fact, promising one-to-one convertibility into cash is not the same thing as being able to perform it. In the case of bank-money, convertibility is guaranteed by deposit insurance and by bank supervision, which assures that the value of bank’s total assets does not fall below that of deposits. With electronic-money, the value is also guaranteed by the portfolio of assets of the financial intermediary. If electronic money issuers are required to hold assets for a value equivalent to that of the money that they have issued in the form of bank deposits, the guarantee is identical to that on bank-money. If this is not the case, holders of electronic-money may incur higher risks than holders of bank-money, even in case of formal one-to-one convertibility.

An additional source of riskiness is the liquidity of each means of payment. Clearly, even in the case of fully guaranteed one-to-one convertibility into cash, discovering that the money that we hold is not accepted to make a payment, for example because our counterpart fears it is counterfeit, will be a source of problems. Clearly, the risk of liquidity is going to be higher for moneys that do not offer, or guarantee, convertibility into bank-money or cash.

Finally, the riskiness of a means of payment is also related to that of the technological infrastructure that it adopts. The security of the plumbing, as the payment system is known in the jargon of central bankers, is a crucial issue, as the recent frauds at bitcoins exchanges have recently made clear. For this reason, in all countries payment systems are supervised by regulatory authorities.

The implications

In the environment described above, the market equilibrium is unlikely to be fully competitive and efficient. A number of factors work against such an outcome. For obvious technological reasons, transferring value within a network of users sharing the same technological platform – what is known in the jargon of the payment system as a closed loop system – is less expensive and more effective than transferring bank-money. These economies of scale emerge because, within a single network, any transfer of value from customer A to customer D amounts to two records in their positions within the system. A money transfer requires instead to record changes in the position of customer A with respect to his bank, call it B, in the position of bank B with respect to the bank of customer D, call it bank C, and in the position of customer D with respect to bank C. If banks B and C want to make the transfer in central-bank-money, a record in their reserve positions with the central bank is also necessary. This requires to standardize information protocols and to connect different computers and servers, all activities that require time and money. Clearly, network-economies-of-scale favor the concentration of the market in the hands of few corporations.

What these corporations will be, is related to a large extent to the effectiveness of the payment experience. Payments are increasingly generated by purchases taking place within platforms offering a large array of services, often owned by bigtech such as Amazon, Alibaba, Facebook, WeChat, Google. Suppliers of such services are in a better position to offer means of payment specifically tailored to the characteristics of their platforms. These corporations have thus strong incentives to create their own electronic moneys and exchange them within the network of their customers, as Alibaba has done with Alipay and WeChat with Tencent, and Facebook is willing to do with Libra. Having few players controlling the entire payment system, and with the ability to bundle purchases and payments, is likely to generate severe anti-trust concerns.

In addition, if such moneys are one-to-one convertible into bank-money or cash, monetary policy will remain in the hands of the central bank. But if this were not the case, a large diffusion of such means of payment might indeed hinder central bank’s ability to conduct monetary policy.

Finally, a subtler issue relates to the value of the information generated by payments. Electronic-money transactions are not anonymous, and can be linked to single purchases. They have therefore enormous value for marketing and lending purposes. How they would be stored, treated and used is indeed a major cause of concern.

The problems listed above are likely to be even more of a concern considering that customers choosing among different means of payment are likely to assign a prominent weight in their choices to cost and effectiveness, underestimating the risk that they incur.

All this, calls for regulatory intervention.

What type of regulation

The digital technology advancements we have experienced in the last fifteen years will produce epochal transformations and, if properly directed, will deliver enormous benefits. Payments systems are not immune to these transformations with a process that will have to be governed for the reasons discussed above. In addition, both the payments system and the digital markets that may be bundled with, such as information (Google), social interactions (Facebook), market places for commodities transactions (Amazon and Alibaba) and entertainment (Netflix and Spotify), they all share a common property, namely the presence of strong network externalities. This force has already displayed its effect selecting winners-take-all champions, and levels of concentration and market power that have rarely seen in market economies.

In this environment, we think there is no doubt a need of ex-ante and ex-post interventions to guarantee that innovations in payments systems produce their benefits, still guaranteeing stability.

In particular, operating ex-ante, regulation must grant at least three main objectives. First it must be able to induce transparency for users on the values of the different means of payments. Second, it should let monetary policy authorities to continue lead an effective monetary policy. Third, it must grant effective competition among the different means of payment.

As for the first objective of regulation, it should be noticed that lacking transparency on the value of means of payments, moral hazard issues may emerge with a bad and risky money winning over good and safer ones. This may in the end lead to a complete break-down of the new payment systems. One possibility to obtain value-transparency is that regulation requests any electronic money to be fully backed by perfectly convertible assets such as banks and/or central bank reserves.

As for competition, both ex-ante regulatory intervention and ex-post antitrust activities are important. Regulation should incentivize if not impose interoperability among the different payment systems so that customers can easily transfer their money across different systems. Standardization of protocols to transfer bank and electronic money across financial intermediaries is key for smooth interoperability. An alternative to full standardization that has been put forward in markets with network externalities is multi-homing, where users find it preferable to be active on several possibly competing networks. In the case of payment system this would require having wallets with different digital payments. Clearly, obtaining these outcomes is not easy and may require significant costs, such as duplicating costs of networks, which however will grant significant future benefits, which explains why regulation is needed.

Ex-post antitrust intervention should instead take care of the dominant positions that naturally emerge in markets characterized by strong network externalities. As mentioned above, companies already relying on large customer base will have incentives to expand new payment systems. They will be able to do so also both leveraging on the bundling of their core business with internal payment systems and also on the processing and use of the large amount of information they have on their customers. As shown by past antitrust interventions, bundling is a very powerful strategy that dominant firms tend to use in order to extend their power in adjacent markets. The risk that payments systems become quickly monopolized by the already dominant players is concrete and antitrust authorities must be on the monitor and promptly act if needed. As with regulation this is more easily said than done. Consider for example the recent acquisitions of Instagram and WhatsApp by Facebook. These mergers have been greenlighted on the ground that these companies where active in different markets. However, it is now clear that the business models evolve very quickly in digital and network markets and that possibly unrelated companies may be formidable competitors in a few months-time horizon. Independent Instagram and WhatsApp could have been now competitors to Facebook and Google on the advertising market.

Authorities and regulators should team up to identify the right trade-off between tackling the dangerous issues outlined above and at the same time guaranteeing that the technological developments on payment systems deploy all their windful to users allowing for a modernization of money markets.