Don't Let Vested Interests Block E-Commerce Progress

The magnetic stripe card for financial transactions was carefully designed, in the late 1960s, to sell large computer systems and dedicated communications networks.

The information content of the magnetic stripe was, on purpose, held to a bare minimum. It contains the identification of the card issuer and the cardholder's account number.

All other data, controls, account options, identification, and even available balances resided in a big data base, in a big computer system, at the end of a big dedicated network.

The concept has worked well. For the cardholder it has provided a new world of function, access, and convenience. It put card associations in business and generated significant on-line authorization revenues.

It also helped sell a lot of large computer systems and networks. It keeps a lot of card suppliers gainfully employed. It has allowed products and functions to spread worldwide to millions of terminals, cash dispensers, and cardholders-all at the end of large dedicated and proprietary communications networks.

Is anything wrong with this picture?

Yes, and fortunately, it is called technological progress.

The central mainframe computers of the 1970s are being overtaken by distributed processing with personal computers, servers, and local area networks. And that progress will be overtaken by a new communications revolution over the next two decades-open communications systems replacing dedicated networks.

Moving value information via telephones, cellular phones, and the Internet will become routine. The merger of telephone and cable distribution systems (typified by the recently announced acquisition by AT&T Corp. of Tele-Communications Inc.) will further fuel the explosion of broadband communications. Direct broadcast satellites also offer new paths for value-related information and transaction implementation.

Behind the scenes, the people and organizations with vested interests in systems based on the magnetic stripe card are hunkering down to preserve their roles.

The credit card associations' Secure Electronic Transactions proposal for the Internet is a good example of constraining technological progress to perpetuate a 1970s solution. It uses the Internet as a front end to the existing card networks, with a complex 26-step process designed to assure security.

Contrast SET with the Mondex solution of authorizing a transaction between the smart card and the local terminal. With Mondex the transaction is captured and batch-transmitted overnight without dependence on costly on-line authorizations on the dedicated networks.

Also, contrast the MasterCard-Visa solution with the French banks' smart credit card. Nine out of 10 smart card transactions are authorized between the card's chip and the point of sale terminal that performs batch data capture. The remaining 10% are authorized on-line with a dial-up connection. This has reduced losses in the French bank card system by 10%, compounded annually.

The smart card with its built-in computer, program logic, authorization rules, and dynamic updating capability does most of the job of controlling card use. The solution is faster and less expensive and opens up major new market opportunities.

The added intelligence in the card allows credit to be extended to many more customers on a more controlled basis. It also allows the use of the bank card in locations away from dedicated networks-at home, in a taxicab, in a television set-top box, in an open-air market, at point of sale devices not connected to dedicated networks, in airplanes, on trains or buses, at schools, on picnics, etc.

Various technical approaches to smart card systems are available, but each vendor's solution is different. Working with Sun Microsystems Inc., creator of the Java programming language, several card suppliers have adopted a Java-based approach that comes very close to providing a "level playing field" for smart cards in the new communications environments.

Initially, Java will require either fewer applications on a card or a larger-capacity chip card to accommodate Java implementation needs. At the current rate of price-performance improvement, this added-cost factor will soon diminish and there will be vastly improved cross-vendor options.

A more problematic alternative is that of the Open Card Framework, or OCF. It is a throwback to magnetic stripe architecture in that it would take everything required for cross-vendor interoperability and put it on network servers.

OCF proponents include IBM and Visa. Is IBM trying to sell future systems? Is Visa trying to extend its proprietary network role?

This OCF solution would run counter to the 30 million-card French smart card transaction system, in which 90% of the transactions are authorized by point of sale units, without authorization via an on-line network. An OCF solution would be inimical to the Mondex program that relies for authorization entirely on a card-terminal connection.

OCF would also block any number of potentially important smart card applications where quick and economical authorization is desired between the smart card and its acceptor, whether in homes, airplanes, taxis, emergency medical situations, or cellular phones.

The industry needs leadership that puts future marketplace needs ahead of any single entity's proprietary network interests. Thirty years of banking and technology evolution have passed the magnetic stripe by.

Now is the time to aim for a completely unconstrained, open- communications future for financial institutions. Java seems to have a strong edge when it comes to smart card applications, though Mondex with its Multos operating system may have a short-term advantage for deploying multi-application cards.

For reprint and licensing requests for this article, click here.
MORE FROM AMERICAN BANKER