Software & Information Industry Association ● Financial Information Services Division

Data Usage and Redistribution Notes

November 2003

The sale of market data provides a critical revenue stream for information providers and represents a significant expense item to information consumers.  As such, the contractual obligations, compliance procedures, reporting requirements and business strategy options associated with the external redistribution of market data are of essential interest to all FISD members.  Consider the following:

 ·         Vendors (including financial institutions) don’t always have control of the downstream redistribution of market data in open access datafeed environments. 

·         Market data policies are complex and there are inconsistencies in the policies of information providers on a global basis. 

·         Market data compliance is governed by contracts which are frequently open to interpretation, don’t always reflect the realities of existing technological environments and don’t always translate well to real-world applications.

·         Compliance involves many people (contracts, entitlements, IT/development, billing/reporting and sales) who don’t always understand (or care about) the practical side of market data rules.  

 The lack of control over data distribution (particularly given the existence of multiple sub-vendor relationships) can lead to under reporting, unauthorized redistribution and other forms of revenue leakage.  Inconsistent market data policy compliance and uneven enforcement can also lead to unfair competition from unlicensed data redistributors.  And addressing the perceived mismatch between commercial policies and marketplace realities could lead to greater compliance and additional revenue opportunities.

 ___________

 Our objective with this document is to define the primary categories of revenue leakage, identify and prioritize the full range of possible responses and implement a series of activities designed to both protect intellectual property rights and promote widespread (legitimate) data redistribution. 

 Redistribution Categories

 Uncontrolled Internal Redistribution (UIR) – this refers to the internal control and management of entitlements within in an entity authorized to receive and distribute market data delivered via a data feed 

• Many firms are likely to have a “market data manager” or “exchange relations manager” who understands contractual obligations – the entity intends to be in compliance
• UIR can occur when applications use market data without the knowledge or approval of the market data/exchange relations manager
• UIR can occur for applications in the “grey area” – not specifically covered in the policies of the information provider
• UIR can occur when there are differences in market data policies among information providers – i.e. an application might be authorized by some but not all of the IPs
• UIR can occur because the various parties do lousy reporting (or things get lost in the tri-party reporting relationship)
• UIR can occur because
  • The vendor/downstream redistributors don’t know the rules (or get inaccurate/misleading information from the upstream redistributors)
  • It’s not in the best interest of vendor sales forces to police IP policies (don’t ask, don’t tell)
  • Vendors or downstream redistributors don’t care about the compliance requirements
  • There are inadequate (technical, administrative) controls or poor compliance policies/procedures in place
  • The downstream redistributors disagree with the policies of the IPs
• What can be done?
  • Ensure that the policies exist, are well documented and clearly understood
  • Establish clear (technical and administrative) entitlement controls
  • Partner with the market data/sales manager to educate downstream applications on the compliance rules
  • Develop practical examples of redistribution as applied against contracts/policies
  • Create monetary penalties (or other forms of accountability) for non-compliance
  • Share information among IPs on the types of violations that occur
  • Develop more rational/consistent policies from IPs

Unreported External Redistribution (UER) – this refers to market data delivered to applications without proper contractual authorization or adequate entitlement controls in place.

• Outright data theft (pirates/bandits) looking to avoid market data fees (we need to define the methods of stealing data – from password sharing to server hacking to deliberate non-reporting)
• Existing price list of are perceived as not justifiable for the applications being fed by market data (too expensive)
• Uneven enforcement creating situations of unfair competition (rules not applied to all)
• Failure to communicate (or conscious decision to ignore) contractual obligations as data redistribution moves down the sub-vendor chain
• Lack of awareness/knowledge of rules governing external redistribution particularly by “new vendors” and ISVs
• What can be done?
  • The overall objective is the development of fair and enforceable Internet data licensing terms combined with effective action against unlicensed competitors
  • Identify the individual sites where UER is taking place (ferret out the data pirates) for compliance enforcement and get serious about cracking down on unlicensed web-hosts (perhaps jointly through FISD)
  • Identify the source(s) feeding the data to the UERs (shut off at source) – many are legitimate redistributors
  • Create a common database of “domains of interest” – all IPs and legitimate redistributors are interested in ensuring a level playing field
  • Work with software developers on the development of functional extraction parameters to better identify “domains of interest” (what is the definition of an interesting domain/website for data licensing purposes)
  • Update contracts and other compliance documents to fills gaps/holes as it applies to internet redistribution of market data
  • Examine commercial policies (fee schedules) to address new application areas not currently envisioned or covered by existing fee schedules
  • Simplify the contractual and administrative challenges associated with conventional market data agreements (as applied to Internet redistribution) – make online redistribution compliance easier and more practical

 Derived Data – refers to the fine line between “market data” and the creation of “original creative works.”   The derived data discussion also covers questions about whether exchanges have the right to charge for derived data. 

• When (if ever) does derived market data cease to be market data (i.e. the exchange’s intellectual property and subject to exchange contractual obligations?
• Are exchange contracts that specify that market data is always market data regardless of how much it is processed, changed and incorporated into other IP creations enforceable?
• Exchanges clearly can’t afford to limit the definition of market data only to the information exactly as originally transmitted.  Many exchanges reasonably argue that there is value in using market data as a dynamic updating component of the vendor or user’s intellectual property creation (think indexes)
• FISD has created a definition of derived data (I think it went something like this)
  • If processing leaves market data in recognizable form or the original data can be re-engineered from the derived work, it remains market data and all contractual obligations and fees apply
  • If processing turns the market data into some other original created work (index, NAV) that can’t be re-engineered, the work is not market data and outside of data fees and contract terms
  • Exchanges may still charge a license fee for the right to use market data to create the derived data (i.e. index creation license fee).  The fees and contractual obligations associated with license fees are commercial issues between the exchange and the entity looking to create a derived work.
• What can be done?
  • The best strategy (hopefully) is to turn the issue into a discussion about “reasonable fees” for the user’s right to create original property from market data
  • The other approach is likely to become a snarl of issues related to “data ownership” and overly broad definitions of “market data.”

Screen Scraping – refers the general practice of extracting information from web sites, so that it can be used in other contexts.  Screen scraping can be a method of unreported/unauthorized onward redistribution. 

• There are a number of commercially available screen scraping products and screen scraping is a viable method of getting at web based data
• Most programmers describe screen-scraping as an ugly, ad-hoc, last-resort technique to get at data that is very likely to break on even minor changes to the format of the data being snooped.
• Many firms use spiders (a program that visits web sites, reads their pages and other information in order to retrieve information from unstructured sources and store in a structured format).
• The technologies used by screen scrapers depend heavily on static displays of content in order for the aggregator to build reliable scripts\methods of access to information sites.  
• Using frame environments, javascript, and dynamic presentation (each time the page is displayed) that change the location\layout of the site on an invisible level without making noticeable user interface changes will make the job nearly impossible for the aggregator to collect information.

Electronic Trading Engines – refers to exchange policies attached to a class of vendors (Independent software vendors - ISV) who create software for trading firms to conduct transactions on the various exchanges.  Since they are trading, they are clearly getting market data.  In order to encourage penetration of such systems, exchanges have generally waived market data fees for these vendors and users. 

• The electronic trading engine problem is mostly an exchange policy decision issue – is it necessary to waive market data fees?  Will reversing market data fees waivers cause trading migration?
• The money “left on the table” is very significant
• The problems cited by the exchanges are
  • How do fee waivers for electronic trading affect the exchanges commitment to maintain a level playing field?
  • How significant is the loss of revenue and is it necessary to continue to waive fees for electronic trading?
  • How many exchanges charge for data fed into electronic trading platforms?
  • What is the potential for additional data leakage from electronic trading platforms?
• What can be done?
  • FISD could survey the exchanges to define policy approaches
  • FISD could provide a “safe haven” venue for discussion of the issue