The Successes

The biggest lesson to learn is every successful standards initiative has delivered value to all parties involved in the effort. If you do not have a compelling business case for everybody moving [to industry standards] from what already is being used, no matter how cute the standards are, theyre not going to gain traction, says Mary Knox, research director in financial services at Gartner.

In banking, one of the greatest win-wins has been in payment processing. Daniel Schutzer, vice president of strategic technology at Citigroup, illustrates the impact with examples of check or credit-card payments. From the consumers point of view, they can present any credit card or check, and from the merchants point of view, payments get done even though they dont have the same banking relationship [as their customers]. Its an example of how standards have both responded to a demand and created an industry.

That industry today con-sists of various industry-owned cooperative networks for interbank payments, including credit-card networks, the Electronic Payment Network (EPN), and the Society for the Worldwide Interbank Financial Telecommunication (SWIFT), the worlds largest financial network. SWIFTs core bank-to-bank messaging service is called FIN, and the protocol used to communicate on FIN has come to dominate the messaging used by banks and investment firms for payments even outside the FIN network. [Today,] it would be unthinkable to process a payment without using SWIFT, says Pierre Pureur, director in the financial services technology group at BearingPoint.

And what was the win for banks? Although the SWIFT [protocol] standardized data [in the payment process], the SWIFT network was a success not because of that standardization but because it delivered secure payments to the financial market, Pureur says.

In investments, the biggest win-win of standards has been improving communication and speeding the trade process between the buy side of the business (investment management firms) and the sell side (brokers). Dominant in that effort has been the FIX (Financial Information eXchange) Protocol.

FIX defines not only message fields and values but also a transport layer and therefore how to make a connection to a sell side [broker]. This makes it easy to add and switch brokers, because all parties to the transaction know how [the FIX message] is going to behave, says Scott Atwell, manager of FIX trading and connectivity at American Century Investments. Atwell also is co-chair of the Global Steering Committee and Global Technology Committee of FIX Protocol Limited (FPL), the industry association for FIX standards development.

American Century has been in-volved with FIX since the standards first development in 1995 and has been adopting FIX since 1996. However, it has been in recent years that American Century has realized the greatest impact of those standards. Today, it requires standards compliance from all broker partners and effectively handles 100 percent of external tradesabout $100 billion annuallyelectronically via FIX.

In the early days of implementation, we measured the impact [of FIX] by our ability to handle increased trading volume without adding traders and being able to add and switch from one broker to another easily, Atwell says. But now we realize the number-one benefit isnt efficiency, its risk reduction.

That is, by enabling straight-through processing from sale to settlement, FIX helps American Century ensure customer orders are processed accurately.

Similar but Different

Unlike payments and trades of listed securities, other financial transactions have a process complexity that traditionally has impeded efforts to develop and adopt industrywide data standards. In insurance, for example, except for highly commoditized lines of coverage, contracts may be unique and underwriting complex, requiring consumption of data from third parties with which no EDI relationships have been established. Part of the problem also is insurance has been regulated by the states, says Carol Chapman, financial reporting compliance specialist for American National Insurance. States just now are getting together and coming to agreements on common [data] formats.

The mortgage banking industry likewise contends with lengthy, complex processes and connections with multiple third parties to complete a single transaction. And in some ways, its EDI challenge is even greater than insurers 50-state problem. Adam Hall, vice president of process technology at IndyMac Bank in Pasadena, Calif., explains of the more than 3,000 counties in the nation, just 60 currently accept electronic mortgage document filing using standards. It will be a while before automation [of that process] occurs, he says.

Insurance, however, does have its strong points. In other industries, there have been failures of vendor-backed technology standards, says Matthew Josefowicz, manager of the insurance group at Celent. The nice thing about ACORD [standards] is they are industry backed and industry designed as opposed to being promulgated by one or two vendors.

Also, ACORD is different from organizations in other financial services in that it covers the breadth of products the insurance industry offers, albeit by different standard sets. In contrast, other sectors tend to have multiple organizationssuch as IFX (Interactive Financial eXchange) within B2B banking, OFX (Open Financial eXchange) in consumer banking, and MISMO (Mortgage Industry Standards Mainten-ance Organization) in mortgage banking.

However, according to Schut- zer, having several organizations hasnt hindered efficiency in the banking industry. These various organizations are each focused on different payments products and businesses with little interaction, such as business to business or merchant to consumer, says Schutzer, who has served on advisory boards for both IFX and OFX.

He also notes banking doesnt have the third-party interaction that insurance does. You come to Citibank, you open an account, and its done by Citibank employees. In insurance you have a network [including third parties such as] agents and reinsurers. Its more important [in insurance] to rely on standards to keep those entities working together.

For its part, ACORD is striving to combine its multiple standards, beginning with the creation of a common data dictionary for all insurance. It does not, however, set a time frame for the ultimate creation of a single XML standard, maintaining that existing investments by carriers prevent the organization from getting there directly and immediately. As additional transactions and processes are identified that impact all lines of business, we will seek opportunities to create uniformity across multiple standards, says Ron Dudley, vice president of standards at ACORD.

Shifting Focus of Standards

The initial interest of carriers in ACORDs XML standards was, similar to the world of investments, company-broker communication. However, the recent trend has been to apply ACORD XML standards within the enterprise.

If you go back two years ago, people werent sure what the benefits of standards would be, Josefowicz says. What has changed this year is people are understanding data standards are a real potential boon to the industry. As Web services and service-oriented architecture have taken hold in insurance, people are embracing the XML standards as a common language between various systems that dont integrate well or that have different data models.

In addition, in the past two years, vendors have been increasing their support of ACORD standards as an integration mechanism. Anything that would reduce the cost
of adding a component to
the infrastructure is valuable, Josefowicz notes.

American National is an example of a carrier using XML internally to connect its various legacy systems and achieve both processing efficiency and development cost reduction. We were dealing with instances where literally hundreds of different hard-coded interfaces were required to move one piece of information through the system in its annuities business, explains Chapman.

XML has proven to be the technological Rosetta stone for American Nationals different data languages and has enabled the insurer to develop a unified front end used by both internal and external constituencies, regardless of the number of legacy applications involved in a particular transaction. The systems that store data are not 2005 systems; they are mainframes, Chapman says. To develop a presentation layer that gives speed and response [without using an XML layer] would have taken hundreds of times more money and development time than it did.

In insurance, carriers also are looking to ACORD as the basis for internal data modeling. Were seeing a lot of interest in enterprise data models this year, Josefowicz points out. And internal adoption of industry standards is a trend taking place in other financial services sectors, as well. American Century, for instance, is using FIXMLan XML version of the FIX Protocolin the gradual rearchitecting of its order management systems and in the development of new systems.

As we are redeveloping parts of the order management system, componentizing pieces, and rewriting them, we need to integrate one part to another and to represent data. It doesnt make sense to create our own format when FIXML is available, Atwell states.

The XML Question

XML-based industry standards still have a relatively low rate of adoption in the investment sector, according to Gartner. The global SWIFT network primarily uses its non-XML syntax, as well. Contrast that with insurance, where standards efforts are predominantly XML based.

Part of the reason is the earlier adoption and success of non-XML standards in other financial services sectors make it difficult to establish a business case for change. There isnt a business driver to move [existing FIX transactions to FIXML], and theres risk and cost in making a move, Atwell says. Its a tough sell.
Atwell also claims XML may not be well suited to some areas of financial services, particularly where the purported overhead of XMLbased on its self-describing natureis an impediment to high-volume transactions. He reports initial efforts by FPL to represent FIX data in XML resulted in messages that were more than six times larger. That presented real challenges, he says, and resulted in FPL transport optimizing XML in its FIXML protocol to reduce the overall message size and address transaction volume concerns.

Gartners Knox, however, asserts the main issue behind the lack of XML adoption in other financial servicesand investments, in particularis simply those industries havent been working with XML as long as insurance has been. If you talk about a very high volume of exceedingly time-sensitive transactions, [overhead of XML] may be an issue. [But] most of the issue around lack of XML adoption has to do with the maturity of the standards themselves at the semantic layer.

We havent had any issues [with XML], says Hall. And honestly, if you spend some time listening to a few developers discussing how to make a change in an interaction with EDI standards, needing to spend a half hour just discussing definitions, the benefits of human readable XML standards are evident.

For their part, most standards bodies are working on XML standards either on a going-forward basis or in a dual-use scenario alongside legacy syntaxes. Additionally, the industries are making the movecarefully and slowlytoward convergence of standards. For instance, FPLin addition to its ongoing work on FIXMLis working with the International Swaps and Derivatives Association (ISDA) on FpML, an XML-based language widely used in trading over-the-counter and complex derivative instruments.

Additionally, external influences, such as government regulation, will drive XML adop-tion. Consider XBRL (eXtensible Business Reporting Language), a dialect of XML whose first specification was published in 2000. There is a lot of regulatory interest in [XBRL] now, Knox says.

The banking industry, for example, is in the process of updating its call reporting, which will replace the submission of various call reports to several federal agencies with one XBRL-based report to the Federal Financial Institutions Exami-nation Council clearinghouse, with the goal of reducing the time of call reporting from two months to two days. Elsewhere in financial services, both the SEC and NAIC are working toward the ultimate receipt of XBRL-based reports. And ACORD has committed to working with XBRL to help insurers bridge from transactional to reporting data.

XBRL transforms financial information from text into intelligent data by applying tags based on standard accounting practices. [Regulators] are interested not only in receiving financial reports electronically but in having the ability to click on a particular number and having that number exploded with an explanation of how it was derived as well as the ability to roll up numbers easily [into new analyses], Chapman says.

Maturing standards, continued adoption, and regulatory requirements will make XBRL increasingly important not just to insurers own financial filings or as a tool to comply with financial transparency regulations such as Sarbanes-Oxley but also in their ability to consume information more easily as underwriters and to manage their own assets.

XBRL allows you, in a much more automated fashion, to pull together and digest information that is typically manually extracted and entered into spreadsheets, says Knox. One of the big promises [for XBRL standards], and one thats still far away from being realized, is internal reporting and consolidation of performance in-formation within the [insurance] enterprise.

But as the history of other efforts has shown, what will most influence the development and adoption of any XML standard is its capability to deliver value. Knox asserts scenarios that present a green-field situation for systems development will provide natural opportunities to do this. You have new instruments being introduced in the marketplace with new requirements for information, so as a result, new [XML] standards are being created, she says.

Future Focus

The progress of data standards adoption in insurance will continue to move faster in internal systems and integration vs. external interfaces with agencies, reinsurers, or other third parties, Josefowicz predicts. Thats not to say there arent successful [external] projects, but [external adoption] suffers because insurers arent looking at opening up their business practices; theyre just looking to do business easier and cheaper, he says.

In the ultimate comparison of insurance to other financial services, its worth remembering even in sectors where data standards are adopted universally today, it wasnt always the case. In the early days of the ATM, nobody was collaborating, Schutzer says. We were hoping if we put more machines on more street corners, we would get more business [than other banks].

It took years to move to an industrywide ATM standard and for ATM banking no longer to be marketed as a competitive advantage, and that was involving transactions everyone easily could agree upon. So perhaps insurancewith its complex contracts and transactions and divergent processescompares better than weve been led to believe after all.