Can One Size Fit All?

Can One Size Fit All?

The industry is clearly in need of a multi-purpose classification standard, but the diversity of most financial products makes the task of creating one a daunting prospect, says Martin Sexton

As instrument classification again comes under the spotlight, with the imminent release of the final draft of ISO 10962-the Classification of Financial Instruments (the ISO CFI) and the publication of the ISITC's Settlements and Reconciliation classification, the question being asked is, "Is it realistic to expect a ‘one size fits all' instrument classification"?
In the distant past, instrument classifications were based around trading desk operations, hence the creation of broad categories such as equities, debt and derivatives, and to this day these tend to form the basis of most classifications. Under these categories, one normally finds less than a handful of attributes to support specific instrument variants. As investment strategies became more complex and a better understanding of advanced product structuring has developed, there is a realization that to be able to adequately represent the characteristics of a financial instrument, more of its properties need to be captured within a taxonomy.
When one examines financial products in detail, it becomes apparent that rather than dealing with a relatively simple species taxonomy, it is more akin to understanding the human genome. For example, hybrid and structured instruments can contain equity, debt and derivatives components, making it extremely difficult to identify the appropriate classification code to be assigned, especially when only a limited number of the attributes are available for use. Therefore, simple solutions have evolved based on how the instrument is traded in the market.
Consequently, if the market treats the product as an equity, then it gets classified as an equity, otherwise if the market treats the instrument as a debt product, it gets classified accordingly. This scenario works well if one keeps the top-level categories and the hierarchical depth to a minimum, however the more complex a classification becomes, the more difficult the code allocation becomes and the more chance of ambiguities being introduced.
In addition to the limited number of characteristics that can be represented, there are other potential areas of concern. These include, an inconsistent approach to the hierarchical tree structure (ie, it may contain branches of mixed semantic meaning), or instances where more than one piece of information is represented within an attribute or jurisdiction, or market-specific characteristics are embedded within the classification.
Given the detailed analysis required to construct an unbreakable classification, it is not surprising to see users opting for a quick-win alternative and designing flat instrument type code lists, rather than taking on this onerous exercise.
Whether a classification is used as part of the investment decision-making process, to support corporate governance or to adhere to regulatory requirements, there are a mind-boggling number of classifications an organization needs to consider (see table).

Overlap
There is obviously a certain amount of overlap between the various classifications, and given that the ISO CFI has not achieved the market traction one would have expected, this highlights the need for an industry-wide review.
A semantically correct and multi-purpose CFI should be within the grasp of the industry. But identification of the resources needed to undertake an assessment is likely to be an issue, at this time of cutbacks and reduced resourcing commitments.
Identifying an organization willing to take on the role of a registration authority of an enhanced classification could be a major issue too, especially if the standard goes beyond the needs of the registration authority itself, thus resulting in an onerous amount of increased maintenance with little benefits to show in return for the registrar. In the past, vendors have shown interest in taking on this role, since they have a reasonable business case for doing so, that being to "lock in" customers to their own service offerings.
Another solution to the maintenance question could be for a financial instrument classification to be managed and maintained as part of the ISO 20022 repository. At the end of the day, a classification is merely a code list, albeit with structure, and as such it lends itself for incorporation into the ISO 20022 repository, which supports hundreds of code lists used by the industry. Alternatively, a mechanism that takes advantage of the latest online community-based offerings could also be considered.
Given the current commitments to the existing standards, it is unlikely we will see a complete overhaul of a standard, like the ISO CFI. Nonetheless, the industry does require an instrument classification scheme that can be used to support the investment decision-making process and for identifying the appropriate risk and fair price calculations to be applied to an instrument. To achieve this, one requires a clear definition of the business contexts the classification is designed for (or "classification facets"). Identification of the key attributes (or characteristics) can then follow, without any preconceived hierarchical structure. Having the attributes independent of the taxonomy has the advantage of providing the capability, if so required, of applying separate views (taxonomic hierarchies) depending on the business context, therefore ensuring the industry is provided with a multi-purpose classification standard.