Smart Financial Reform Should Provide More Information, Not More Regulation
June 5, 2009
Many are advocating the virtue of more transparency but may not understand the mechanics involved.
The idea of gathering information to give to regulators is simple to understand, but very complex in reality. There are no certificates, no physical documents of any kind, just electronic data identifying products, counterparties and other attributes arrayed in automated data bases. When the attributes of a financial transaction are uniquely identified as in conducting a particular trade in a product such as a stock, bond, futures, option, derivative, mutual fund, etc.; and the identities of counterparties conducting the trade are equally uniquely identified; then the transaction can be accessed by automated means, aggregated and reported on to management in performance and risk reports, and to regulators.
The problem arises first around the uniqueness of the data attributes themselves – IBM for example has over 25 different identifiers used in exchange trading systems, and payment and settlement systems across the globe. Berkshire Hathaway as a business entity has at least seven different identifiers in critical information data services such as the Edgar filing system, S&P’s credit rating services, and in Markit’s derivatives price reporting service. Lehman Brothers had 9,076 security issues under 204 unique issuing identities, each with different numbered identities in both internal data systems and at external clearing and depository facilities. Lehman did, as many other systemically important financial firms do, operate globally making transparency a global aggregation issue for each sovereign regulator.
The second problem relates to the ubiquity of the information. If, for example, a financial institution had been doing business with Lehman across the world and they, or their regulator wanted to understand their potential loss exposure as Lehman deteriorated, they would, and did, have quite a challenge. They would have had to both discover and then collect all of the information across all the disparate internal systems and business applications storing information on Lehman, not all identifying Lehman the same way, then aggregate it, value it against closing prices, and then compare it against a credit limit that may have been extended to Lehman corporate at the top of the pyramid, or to any of its subsidiaries, or that was set to limit trading exposure to Lehman in a particular product or market. If the institution owned any of Lehman’s issued securities, or derivatives; or held them on behalf of clients or in managed accounts; or had any of a multitude of relationships in which it could be doing business with Lehman; then it would have again needed to pull all of this together to further determine its, or its clients exposure.
Given the urgency of any moment when financial transactions are entered into in real time and have the potential of risk exposures far exceeding the notional value of the transaction, aggregating this information would be most meaningful if it could be brought together in near real time, but certainly daily, so that management, and regulators could make informed decisions. This does not happen in today’s financial industry because the underlying identifying data and valuation information are sourced independently, with each financial institution performing duplicative functions in an attempt to represent each unique product, business entity and valuation price identically, but failing to do so. The consequence is that proprietary and conflicting valuation prices and identification codes exist across the entire range of identifying data, including such fundamental identifiers as symbols for corporate issuers, symbols used in contract markets, numbering conventions for securities, supply chain business entity identifiers, and counterparty identifiers. This causes built-in delays in mapping and reconciling data. In the U.S. securities markets it takes three days for this process to run its course, with significant risk and cost consequences in the interim, notwithstanding the difficulty in getting accurate data to regulators in any meaningful time frame.
