No Free Lunch: A Primer on Interchange and Debit
SEE LAST PAGE OF THIS REPORT Howard Mason
FOR IMPORTANT DISCLOSURES 203.901.1635
March 28, 2014
No Free Lunch: A Primer on Interchange and Debit
This note comments on the WMT suit filed yesterday against Visa in the context of the broader impact of interchange on the structure of the card payments industry, particularly in debit. The dynamics are more involved than in many industries and it is plain, not least from the extraordinary comments in last Friday’s appellate Court ruling (see our “Quick Thoughts” note of this Monday), that merchants have not been able to convey an adequate understanding of the competitive and economic realities of the US card system through the Court system. Our object is to provide investors with a sense of how these competitive and economic realities work in practice, of how they have caused the US payments system to evolve in a dysfunctional way (particularly with respect to signature debit) in absolute terms and relative to overseas markets, and how the partnership of the merchants’ payments consortium, MCX, with FIS (which is a partner of the more secure and efficient debit networks in both Canada and Australia) is likely looking to shape the US market.
- Yesterday, WMT filed suit against Visa with a key allegation being that network-set interchange rates, particularly in combination with the honor-all-issuers rules, represent horizontal price-fixing between issuers. This claim has been advanced in the past and the Courts have consistently ruled that interchange fees are reasonable and essential for a payment card system.
- This Court position is puzzling given experiences in other markets. Canada has a zero-interchange debit model in the (not-for-profit) Interac payments system which performs more strongly than the US system on several important measures: it is real-time, has fraud rates just one-third of those in the US (because, with issuer incentives aligned around reducing cost rather than maximizing interchange, it has adopted chip-and-PIN), and runs on network fees of less than 1 cent versus 6-7 cents for US.
- Critics of zero-interchange systems suggest consumers pay more for debit cards (and, indeed, in Canada there is tiered pricing for checking accounts based on customer use of debit card transactions) and receive fewer benefits (and, indeed, in Canada there are no charge-backs). The merchant counter that lowering US interchange would provide an equal-or-better consumer benefit in the form of savings at point-of-sale is dismissed as if interchange generates a (merchant-subsidized) free lunch to consumers.
- The economic reality in the US is that, through interchange, some consumers are getting a benefit they do not pay for (because either merchants do not raise prices or, if they do, the higher prices are typically borne by all consumers even those offering other tender). It is hard to see how this creates a more efficient allocation of payments resources than the Canadian system where consumers pay more transparently for the debit transactions that they use.
- We understand the role of transfer prices, such as interchange, in two-sided markets such as payments. Our issue is not with network-set interchange in theory as with its US implementation in practice. The root problem is that V and MA exploit their power in the credit card market and hence (at one time through the now-defunct honor-all-cards rules and now through FANF) in the debit market through setting supra-competitive interchange with object not of optimize the payments system as a whole but maximizing their volumes through locking-up the issuer distribution channel.
Importance of Interchange
The key pricing variable in the US card business, fundamentally affecting industry structure, is interchange which is effectively a transfer payment between the merchants accepting cards for payment and the banks issuing them to consumers. Transfer payments are common in “two-sided” markets where the aggregate benefit depends on the participation of two different types of customer in some optimal ratio: the textbook example is of the cover charge assessed by some bars on women, but not on men.
In the case of the card business, the interchange “transfer” payment is technically made by the “acquiring” banks representing merchants in the V or MA systems to the “issuing” banks representing cardholders. In practice, acquiring banks pass the interchange cost through to their merchant clients and it is by far and away the largest portion of the overall cost to the merchant of accepting credit or debit cards even after the Durbin amendment capped debit interchange for regulated issuers.
Specifically, debit interchange for regulated issuers is ~24 cents/transaction compared with the other two fees paid by the merchant: the network fee to V or MA of ~6 cents/transaction and the acquiring fee to the acquiring bank which varies depending on the type and size of the merchant. WMT, for example, probably pays an acquiring fee of less than half-a-cent/transaction while a typical retailer will pay 2-3 cents/transaction. Unlike debit interchange which is assessed per transaction, credit interchange is assessed more on the size of the transaction and runs at ~1.5% for a non-rewards credit card and ~2.4% for a rewards credit card. The acquiring fees are similar for debit and credit while the network fee is higher for credit at ~10-12 cents/transaction.
The above figures are approximate averages and conceal a great deal of complexity and variation. There is an important distinction between card-present rates (when a card is swiped at the merchant’s point-of-sale terminal) and card-not-present rates (when card credentials are entered online or conveyed via a mobile device); the card-not-present rates are higher. There is also wide variation between merchant categories with supermarkets paying lower interchange than restaurants, for example; to give a flavor for the complexity, Visa’s current credit interchange table, typically revised each April, is included in Appendix A1 and the debit interchange rates for unregulated issuers for different networks included in Appendix A2.
Impact of Interchange on US Debit Industry
Signature debit (where a consumer authenticates a transaction by signing for it just like a credit card transaction) is an inferior product to PIN debit (where a customer authenticates a transaction by entering a Personal Identification Number just like an ATM transaction) because there is higher fraud content. Globally, the fraud on signature-authenticated transactions is ~6 cents/$1000 while the equivalent figure for PIN-authenticated transactions is 1 cent; this stands to reason since, if you lose your debit card, it is easier for a criminal to use if fraudulently by forging a signature than by guessing the PIN.
It is remarkable, then, that alone among developed card markets the US, has persisted with signature authentication in both credit (which is always signature-authenticated except when a cardholders obtains a cash advance from an ATM) and debit (where two-thirds of transactions are signature-authenticated versus one-third being PIN-authenticated); indeed, only 25% of merchants in the US have the point-of-sale (POS) equipped with the PIN pads necessary to accept PIN-authenticated transactions. Equivalent figures overseas are over 90% in developed Europe, over 70% in Canada and Latin America (combined) and in Africa, and over 50% in Australia. These figures are based on adoption of chip-based cards (under the “EMV” standards) which, outside the US, are almost exclusively chip-and-PIN (see Exhibit 1).
Exhibit 1: EMV Adoption Rates by Region
Source: EMVCo, 2011Q4
Debit, in particular, provides a revealing case study on the determining impact of interchange on industry structure. Debit cards first appeared in the US in the early 1990s and they were exclusively PIN-authenticated. They were an innovation introduced by the Electronic Funds Transfer (EFT) networks (such as STAR now owned by FDC, NYCE now owned by FIS, and PULSE now owned by Discover) who realized that ATM cards could provide DDA-access at point-of-sale (and so provide an electronic alternative to writing a check) provided merchants installed PIN pads at checkout terminals. Banks embraced the idea because processing an ATM transaction was cheaper than processing a check transaction and created an incentive for merchants to install PIN pads with reverse interchange: on each debit transaction, the bank paid the merchant (not the other way around as it is today).
It is highly likely that debit would have evolved to be PIN only in the US (as it is today in Canada, for example) but for the use of market power by V and MA in credit to develop debit franchises. Specifically, in the mid-1990s, V and MA introduced debit cards that were processed just as credit cards are today: customers signed for transactions (rather than authenticated with a PIN as with the EFT-based “PIN” debit transactions) which were processed over the same networks V and MA used for processing credit card transactions. Indeed, the only difference between these “signature” debit transactions and credit transactions was the issuer bank settled the former against the cardholder’s DDA account and the latter against a pre-approved line of credit.
Crucially, to create an incentive for issuers to distribute their signature-debit cards in preference to the PIN-debit cards enabled by EFT networks, V and MA insisted that merchants pay interchange to issuer banks at nearly the same rate as on credit cards. Merchants, of course, did not need to make any changes at checkout since the signature-debit cards used the same point-of-sale technology as credit cards, and V and MA were able to guarantee merchant acceptance of their debit cards through the honor-all-cards rules: these “all-or-none” rules, in addition to an “honor-all-issuers” component insisting that merchants could not pick and choose between which issuers of Visa-branded cards they accepted, were extended to include an “honor-all-card-products” component insisting that if merchants accepted Visa-branded credit cards they also accepted Visa-branded debit cards.
In 1996, the merchants litigated arguing, among other things, that honor-all-card-products represented a violation of the anti-tying provisions of the Sherman Act (with Visa using market power in the credit market to extract premium economics in the debit market) and the case was settled in 2003 with V and MA paying multi-billion dollar fines and allowing merchants to accept credit cards without accepting debit cards from the same network. By then, however, signature-debit cards commanded over 60% share of the debit market and, in order to compete (i.e. in order to encourage issuers to enable debit cards for PIN-based debit), the EFT networks introduced interchange payments from merchant to issuer, and raised them to near the level of signature-debit cards (see Exhibit 2).
Exhibit 2: Average Fee Charged to Small Retailer on a $100 Visa Transaction
Source: NY Times. See Appendix for Interchange after 2009 and Durbin Impact
In short, by leveraging market power in credit, V and MA were able to use signature debit to crowd out competitors even though PIN debit is a more secure product. Given this history it is surprising that the Fed argued against increasing competition in the signature debit market (in recent litigation on implementation of the network exclusivity provisions of the Durbin Amendment) on the grounds that to do so “could potentially limit the development and introduction of new authentication methods” and that the Court accepted this argument adding that more competition in signature debit could mean issuers “would be unable to compel merchants to accept new authentication techniques”. The reason the US debit market is expensive, slow, and at high risk to fraudulent behavior is the super-profits historically generated by issuers and networks on signature debit.
Canada: A Different Model for Debit
Merchants have long argued that interchange, particularly when combined with honor-all-issuers rules, represents a form of horizontal price fixing among issuers, and stifles competition among issuers to have their cards accepted by merchants. Indeed, this is a key allegation in the case filed against Visa by WMT yesterday but US Courts have traditionally not agreed and ruled that interchange fees were reasonable and essential for a payment card system. The Fed has consistently recommended against the regulation of interchange on practical grounds even while acknowledging that market-set rates may not maximize economic efficiency. The GAO and others have expressed concerns about the impact of fee regulation on cardholder fees, on whether merchant savings would flow through to customers, and more generally on the balance of costs and benefits between cardholders, merchants, issuing and acquiring banks, and networks.
Fortunately, none of this is entirely hypothetical since zero-interchange payment networks operate in other countries, including Canada where debit runs over the not-for-profit Interac network. There is no interchange and network fees are less than 1 cent/transaction. It is a “good-funds” model in that authorization and payment occur in real time, and uses chip-and-PIN (under EMV standards) domestically generating fraud losses that are one-third of those incurred when Interac cards are used overseas with their mag-stripe which is retained for global interoperability.
Without a merchant subsidy via debit interchange, Canadian banks charge consumers more for checking accounts and debit card transactions than US banks. Consumers can pay over $10/month for a checking account with unlimited debit transactions or $4/month for a checking account with a limit of 10-15 free debit card transactions after which there is a cost of ~60 cents/transaction. Furthermore, Interac cards do not offer charge-backs (where a transaction is unwound based on a consumer complaint) and cannot widely be used overseas or at all outside North America.
Advocates of the US interchange-based system highlight the drawbacks of a zero-interchange system in the form of higher banking costs and lesser card benefits to consumers (because of the lack of merchant subsidy), and dismiss the notion that merchants pass on interchange savings to consumers in the form of lower prices. In other words, they argue that interchange generates a free-good to consumers: they pay no more at point-of-sale and get merchant-subsidized benefits out of their bank.
The economic reality is that, through the interchange mechanism, some consumers are getting a benefit which they are not paying for (because either merchants do not raise prices or, if they do, the higher prices are, in practice, borne by all consumers even those offering other tender). In the Canadian system, consumers pay for the debit services they use, and it is hard to see how this creates a less efficient economic outcome.
Appendix: Credit and Debit Interchange Tables
Exhibit A1: Current Credit Interchange Schedule for Visa
Exhibit A2: Debit Interchange by Network
Exhibit A3: Average Debit Interchange (cents/transaction) for Unregulated Issuers
- Senator Durbin accepted this argument when a network set the interchange rates but not when they were set by the issuing bank. In his April 2011 letter to Jamie Dimon, he writes: “It is important to make clear that if Chase wants to set and charge its own fees in a competitive market environment, the amendment does not regulate those fees. The only regulated fees are those that banks let card networks fix on their behalf”