Most customers could care less whether they enter their 4-digit PIN or sign a sales draft when using their Visa or MasterCard logoed debit card (“check card”) to make a purchase. The price they pay for goods or services is the same either way.
But the customer’s mode of choice has had a material impact on the cost of the transactions to merchants – at least until recently. For those merchants that have “interchange-plus” agreements with their banks, allowing for pass-through debit network costs, PIN-debit transactions historically have cost far less than if the same transaction was processed as Signature-debit – i.e. with no PIN validation. As recently as two years ago, the cost of PIN-debit was as little as half that of a Signature-debit transaction, depending upon the sale amount, merchant type and other factors.
The PIN advantage always was particularly pronounced for merchants selling higher-than-average priced goods and services, i.e. items priced above $40-$50 or so. That was because the debit networks charged more in transaction fees than, say, credit card transactions (about double), but at least they always “capped” the percentage assessed on the sale amount for most transactions. There was a maximum fee for the most common types of transactions no matter what the ticket size. So, an average retail PIN-debit transaction often would cost no more than 60 or 70 cents all-in.
This was not so for a Signature-debit transaction, however. Signature-debit was priced like credit card transactions, with fees more or less proportion to the price of the merchandise or services.
The gradual tectonic shift in debit card pricing strategy over the years turned into a tremor felt throughout the payments industry in 2009 when Discover’s PULSE network (the third largest) removed its fee cap on most types of PIN-debit transactions. The Pulse percentage fee now floats with the ticket size for standard transaction types. Since then, the other major EFT networks – Star (First Data Corp.), NYCE (Metavante), Accel/Exchange (FiServ) and CU-24 – have followed suit and have eliminated or moved-up the caps on the most common PIN-debit transactions.
The final and most influential change just occurred in April 2010. As part of Visa’s semi-annual interchange modifications, Visa raised the Interlink interchange rate for PIN-debit pricing from 75 basis points and $.17 per transaction to 95 basis points and $0.20 for standard transactions. Visa PIN and Signature pricing are now the same. Considering that Visa’s Interlink debit network has a 40%+ share of the debit card market and that it processes an estimated 73% share of all Signature-debit transactions, the practical impact of Visa’s change to merchants is likely to be profound.
Rate strategies vary and the devil is in the comparative pricing details between networks and transaction or industry categories. So, there still can be found certain price advantages for PIN-debit for some industries and in some ticket-size scenarios. However, there no longer is a noticeable safe haven price gap between PIN and Signature to be exploited. That is particularly counter-intuitive considering processing rates traditionally have had some relationship to transaction financial risk; PIN-validated transactions have been viewed as inherently less prone to repudiation, thus less risky to the account holders and banks involved than transactions with signature verification.
Many merchants are resigning themselves to the new reality in which Signature-debit reigns supreme. With little price advantage remaining between PIN and Signature, major industries like pay-at-the-pump fuel operations are seriously are debating whether to make the estimated $20,000 per location investment required to convert current PIN acceptance devices to a new generation of more secure PIN data devices as mandated by the PCI Security Standards Council Data Security Standards . It even remains to be seen whether Cosco, Home Depot, Wal-Mart and other mega-merchants that, due to the price differential of Signature-debit, have not even offered Signature-debit as an option to POS debit customers, will soon do away with PIN-pads.
Major check card issuing banks have been actively promoting Signature-debit in promotional mailings to their cardholders. They have offered incentives on purchases that only apply when a Signature-debit transaction is made. JPMorganChase, according to news reports, has made the inaccurate inference to its cardholders that Signature-validated transactions are somehow more secure than PIN transactions. Presumably they are promoting Signature over PIN debit because they make back more money, as card issuers, from the resulting Interchange fees being paid when a Signature-debit transaction is processed via the credit card networks, rather than when a PIN-debit transaction is processed by one of the Electronic Funds Networks (“EFT”) networks. Now, that difference has been largely negated so there is little need to promote Signature-debit per se.
What is causing the major strategy shift by the EFT networks and Visa? It is all part of the overarching competition between networks to woo financial institutions to issue cards processed by their networks. The networks make money on the transaction processing (“switch”) fees assessed to every transaction that occurs (from about 3¢ to as much as 8¢ for PIN-debit, depending on the network). The majority of the fees charged to the merchant – per transaction and per sales dollar – are paid to the card issuing bank. So what we have been seeing is essentially a bidding war among the networks to get their brand “bugs” printed on the back of the most cards issued.
Having said that, it is not at all obvious yet just how much of a shift in real fee income will come from all this market re-positioning and effective price parity between PIN and Signature-debit. Debit cards have become more and more popular in recent years, especially during the recession. Debit now surpasses credit as the preferred means of POS payment. But debit cards, in whatever format, still are essentially plastic checks, with more or less the same consumer behavior patterns as cash. As such, debit transactions might be more common than credit card transactions, but the dollar volume of debit is far from matching total credit card purchase volumes. Understandably, debit purchases are limited to bank account balances. Higher value goods and services continue to be purchased on credit – although less credit is being extended during the recession. According to a 2009 PULSE study, the average pin debit transaction in 2008 was about $42 for PIN-debit compared to $37 for Signature-debit. PULSE estimated more a quarter of all debit transactions in 2008 were for purchases of less than $10. This compares to average retail credit card purchase amounts that are usually approximated to be about $70-$80.
Processing cost is the paramount issue to most merchants, yet there still are several soft reasons for reasons for a continue acceptance of PIN-debit transactions:
- Customers often are more comfortable with the security of PIN validation. After all, the card is accessing their bank account. Presumably, only the card owner knows and uses the PIN.
- There are many consumers without credit cards, or who have reached their limits. PIN and Signature-debit let the customer make a cash-less purchase with the funds they have in their bank account but happen not to have in their wallet.
- For the merchant, funds process through the EFT networks almost immediately. By contrast, a Signature-debit transaction, run through the credit card networks, might take 2-3 days to be debited from the consumer’s account. This can mean less chance of insufficient funds and sometimes faster credit of sale proceeds to merchants.
- PIN transactions typically can be completed faster in the check-out line and require less documentation and storage by the merchant. A PIN-debit transaction has considerably less chance of being repudiated successfully.
- A PIN-debit transaction allows for cash-back to the cardholder at the POS; this is not possible with a Signature-debit transaction.
Even so, much the appeal of a PIN transaction is proves to be somewhat illusory:
? While true that PIN-authentication is relatively fool-proof security for a specific transaction, the card remains vulnerable as long there is a Visa or MasterCard logo on the card and the card can be alternatively tendered for a Signature-debit transaction. Signatures or IDs, obviously, are not always verified by store clerks.
- Both PIN-debit and Signature-debit cards benefit from consumer protection laws and practices, although the protections come from different laws. If reported lost or stolen in a timely manner, the cards have a maximum of $50 liability to the account holder.
- PIN-debit transactions usually do not accrue rewards credits from the issuing banks, whereas Signature-debit cards do. This feature, alone, is accelerating the shift from PIN to Signature.
There is little merchants can do to shape the system to their benefit. Merchants are under contractual obligations by their acquiring bank sponsors to accept all cards with a Visa or MasterCard logo. They are prohibited from attempting to “steer” a customer card transaction from one mode of acceptance to another. Card company rules further require merchants with PIN-processing capability to proactively provide customers the choice of making a PIN-debit or Signature-debit transaction. At best, merchants can only make the threshold decision to either enable acceptance of PIN-debit by making a PIN-pad available at the POS or to accept only Signature-debit transactions at the POS.
Written By: Charles S. Crawford, Executive Vice President for Strategic Development
Electronic Payment Exchange
The Tokenization Bandwagon Is Music to Our Ears
Wednesday, June 23rd, 2010In May, EPX issued a press release entitled “Electronic Payment Exchange Enters its Tenth Year of Issuing Tokens for Securing Credit Card and ACH Transaction.” The fact that EPX pioneered such a novel and important technology for protecting merchants and cardholders from the risk of data compromise was not unusual. Our company was founded in 1979 as the first independent processor of electronic checks for merchants. Since then, we’ve been consistently bringing important innovations to market. But giving merchants and consumers the protections of credit card data “tokenization” in early 2001 was all-the-more impressive when seen in the context of the Times.
Back then, cardholder data security was not exactly the front-of the-forehead issue that it is today. There had not been a notable card data breach in the 35 years since revolving credit cards had been used. The first relatively large and publicized incident came just after the Y2K ball dropped in Times Square in January 2000. Online retailer CD Universe exposed 300,000 customer card records. (Of course nowadays a breach exposing a mere 300,000 records would be considered a lucky break.) Since that first major incident, ever more damaging breaches have occurred like clockwork. Two were of Guinness proportions: retailer TJX in 2007 and processor Heartland Payments in 2009, both of which reportedly exposed more than 90 million card numbers.
When EPX started tokenizing data, Visa had just begun to formulate the first generation of data security standards. At first, Visa’s compliance targeted only e-commerce payment gateway operators, not merchants. MasterCard did not initially see the need for standards, so offered data security consulting services. The launching of the Payment Card Industry Security Standards Council was still five years away.
So, understandably, EPX’s breakthrough came with no public fan-fare and unknowable future significance. Our engineers simply were looking for a way to make transaction lifecycles and follow-on transactions more efficient, and our merchant customers more secure. Being engineers, they didn’t call what they created ‘tokens.” They called the codes card data “GUIDs” and “Replacement Values.” (Surprising, isn’t it, such a sexy name didn’t catch on?) The generic catch-word for such codes, “tokens” did not come into vogue until 2005 when Las Vegas payment gateway operator and software licensor, Shift4, Inc. coined the term. (Shift4’s process of code generation within the merchant’s environment, and their data flow is significantly different from EPX’s off-premises approach, but more or less aims at the same purpose.)
As EPX gained practical experience, naturally we kept evolving and perfecting our technology to make it more effective, practical and efficient. As breaches kept hitting the headlines, we kept hearing loud and clear from merchants, particularly CTOs, that they would be delighted if they never had possession of the vulnerable cardholder data in the first place. And, they truly loathed having to spend so much time and IT budgets system major (non-ROI) system remediations to comply with new PCI Data Security standards. With this guidance from the market, along came another set of EPX breakthroughs in 2005. We invented a number of methods for at-risk card data to be securely captured and stored only by EPX and never routed to the merchant. Ever. We filed for a patent for the sophisticated processes that are now at the heart of EPX’s BuyerWall™ security suite.
As the number and scale of data breaches increased over the years and PCI compliance became mandatory and urgent, the IT Establishment naturally first turned to the familiar techniques they had been taught in schools and had been using for decades: encryption, firewalls and other data “hardening” techniques. Several front-end only gateway operators had been offering forms of tokenization. There also were several companies providing software-as-service (SaaS) outsourced tokenization and still others selling do-it-yourself in situ tokenization hardware and software to merchants. Yet, tokenization remained mostly marginalized as an emerging technology …and too-good-to-be-true… by Conventional Wisdom.
Then, a funny thing happened along the way to achieving cardholder security Nirvana: Heartland. Heartland Payments and others quickly became iconic in proving that Encryption Does Not Necessarily Equal Security. Since Heartland, hard-pressed CTOs and cash-strapped CFOs have been gradually opening their minds and wallets to alternative security approaches like tokenization.
Yet, oddly, EPX stood alone for all these years as the ONLY full authorization / capture /clearing / settlement processor providing tokenization. The giant end-to-end processors like Global Payments, TSYS, Elevon, Fifth Third, and First Data Corporation stayed on the sidelines, leaving it to their merchants to solve the card data security problem on their own. Finally, in 2009 Fifth Third Bank (which has its own in-house front and back-end processing) and then First Data (the world’s largest processor) respectively launched their versions tokenization. JPMorgan Chase’s Paymentech merchant acquiring company is not a self-contained end-to-end processor, but in the past few months has been sporadically promoting its Orbital gateway as having tokenization capabilities…although they appear to be using bolted-on functionality provided by a third-party vendor.
Tokenization is not only a solution for credit cards, but also for other forms of payment. A few weeks ago, ProPay, a well-respected Salt Lake City-based payment ecommerce gateway company, made a nationwide announcement that it can now can encrypt and then tokenize electronic check routing and account holder data that is used in ACH transactions. Likewise, on May 19th, Sarasota, NY-based ACH Payments, Inc. said it now will tokenize checking account numbers used in its ACH processing. ProPay’s COO was quoted as saying: “ProPay is leading the industry and applying the same technology for protecting payment card information to the protection of ACH data…” We at EPX appreciate the executive’s exuberance; however, the “leading the industry” part was a bit over-stated considering that EPX started tokenizing ACH transaction data as well – more than nine years ago.
EPX always knew that what we innovated in 2001 would not suffice as the complete answer to data protection. Tokenization, for sure, is elegant and powerful…and is especially cost-effective for complex enterprises with lots of locations and transactions. It mitigates the vast majority of cardholder data risk – substituting codes for card numbers stored or “in motion.” In the case e-commerce transactions, the vulnerable data can be directly captured, encrypted and tokenized from the moment a customer or clerk keys in the data.
However, things are a little more complicated for retail POS “swipe” transactions. The card data can be potentially vulnerable for what I call the “first inch” – i.e., the momentary transit between the magnetic stripe to the point the data reaches the POS terminal or the payment module within a POS retail management systems’ software. Although only briefly exposed, the can be skimmed or otherwise criminally compromised. Also, such exposed card data at the front end-point of a transaction remains ‘in scope’ and subject to more cumbersome PCI Data Security Standards reporting.
We at EPX knew this was a problem to be eventually solved. We watched with particular interest last October as First Data Corporation and RSA (the security division of EMC Corp.) announced their solution: instant encryption as the data is captured by a POS terminal, then tokenization of the data once it is captured by First Data’s processing platform. They call their product (still being field tested) “TransArmor.”
We applaud First Data’s adoption of encryption+tokenization and expect the technology to be a game-changer in the industry due to the huge number of merchants FDC supports. And we welcomed the recent announcement by TransFirst’s Payment Processing International subsidiary (an ISO with a gateway) of offering encryption+tokenization capability. However, true-to-form, all this big news is déjà vu for EPX. In July 2009, EPX already had become the first processor in the world to introduce just such a solution –encryption of data all the way from the mag stripe to EPX’s firewall, then tokenization of the data once it entered our processing environment. EPX’s encryption + tokenization is functionally consistent with what First Data/RSA and PPI later announced. EPX uses an encrypting swipe device to capture the vulnerable data. We hold the only decryption key to the swiped data in our secure processing environment (i.e., neither the merchant nor any other party ever has access to the decryption key). We and our merchants use EPX BRICs (tokens) exclusively as transaction reference codes for all operational reference purchases thereafter.
These days there are an increasing number of companies offering what might be broadly called “tokenization.” The differences between approaches can be hard to discern. The most important differentiator, however, is in the fundamental integrity of the token creation protocols. From the beginning, EPX engineers had the foresight to not take the obvious short-cut of simply creating the token algorithms from credit card numbers and banking account numbers. EPX codes, instead, are based upon the unique and very specific characteristics of each specific transaction and its place in time, among other characteristics. In retrospect, as criminal rings have become so much more skilled at reverse-engineering financial account numbers, we now know how much more secure is the EPX approach than others. If the card number, or checking account number is not in the merchants systems – or the source of the tokens – the data cannot be stolen and deciphered. In other words, it has no “street value.”
In the 31 years EPX has been in payments business we have made many breakthroughs by simply pursuing what is most effective, what is most efficient and what serves our merchants best. We never have waited for others to lead the way, nor will we in the future.
Posted by Charles S. Crawford
Executive Vice President
Strategic DevelopmentElectronic Payment Exchange
Tags: BuyerWall, charles crawford, credit card processing, credit card processor, data breach, debit card processing, debit card processor, Electronic Payment Exchange, encryption, end-to-end encryption, EPX, payment processing, payment processor, PCI compliance, tokenization
Posted in EPX Commentary | 2 Comments »