For some time now, the Securities and Exchange Commission has longed for a system that would enable market regulators to follow the progress of every securities order from cradle to grave. Unlike most other Commission initiatives, this goal cannot be accomplished by a rule applying to broker-dealers. Instead, the cooperation of self-regulatory organizations is required. For those of us who tend to regard self-regulatory organizations as SEC stooges, it will come as a surprise to learn that SROs have been almost uniformly hostile to the concept.
In 1996, the SEC reacted to the Nasdaq market maker scandals by suing the NASD for failure to enforce and investigate potential misconduct by its members. The resulting order required the NASD to design and implement a system that would enable it to follow the path of an order from the time first received by a broker-dealer through to its ultimate execution or cancellation. In 1999, the SEC reacted to the specialists "pennying" scandal by suing the NYSE for failure to detect violations of federal securities laws and the NYSE’s own rules, failure to enforce performance-based compensation arrangements and failing to provide adequate surveillance. The resulting order required the NYSE to design and implement a system that would enable it to track orders into the market. In 2000, the SEC sued AMEX, CBOE, the Pacific Exchange and the Philadelphia Stock Exchange for failing to enforce compliance by its options market makers with exchange rules and the federal securities laws. Once again, the eventual order against these options exchanges required them to design and implement a system that would track the path of an options order through execution and cancellation.
Litigation is always messy business. In this case, the markets are cursed with three different and inconsistent systems. The NASD created the Order Audit Trail System, or "OATS," which tracks equity orders for over-the-counter and Nasdaq stocks. OATS is now operated by the NASD’s successor, FINRA. The NYSE instituted the Order Tracking System, or "OTS." And, the options exchanges created the Consolidated Options Audit Trail System, or "COATS." None of these systems is compatible with each other.
Existing order audit trail systems have been fabulously expensive to design and implement. Broker-dealers have paid millions in fines for failure to comply with rules requiring accurate order input into and error checking output from such systems. However, there is no system that tracks all orders in NMS stocks, let alone all equity securities. Since the market for NMS stocks is fragmented into some 40 venues, existing order audit trail systems, instituted at great expense and considerable effort, are next to useless as a regulatory tool. As a result, existing order audit trail systems failed to identify the causes of the May 6 "flash crash." In fact, it is difficult to point to a single enforcement action where proof of wrongdoing was elicited by examining data provided through OATS, OTS or COATS.
Differences in regulatory requirements tend to favor some competitors over others. It soon became apparent that FINRA’s OATS system was much more intrusive and expensive to comply with than the NYSE’s OTS. As a result, a broker-dealer with a choice of execution venues will prefer the NYSE to Nasdaq because OATS renders a Nasdaq execution more costly than compliance with OTS. Moreover, many of Nasdaq’s competitors, such as BATS and NYSE ARCA, have no order audit trail rules for equity executions, rendering execution on these venues less expensive than either Nasdaq or the NYSE, all other things being equal.
Nasdaq reacted to getting the short end of the regulatory stick by complaining. Nasdaq contended in a 1994 petition for public rule-making that all equity exchanges should be subject to the same order audit trail requirements.
Nasdaq’s petition was a welcome development for the SEC, since it amounted to support for an SEC initiative to strengthen order audit trails, rather than the withering antagonism encountered in all prior attempts to implement such systems. Though Nasdaq’s motives may not have been pure, the SEC apparently took heart from Nasdaq’s desire to expand its competitors’ audit trail requirements, and came up with more rules. The current proposal is much more aggressive than anything the SEC has tried to accomplish in the past.
The SEC’s proposal is aggressive for two reasons. First, it would require the information to be available in real time. FINRA’s OATS system generally requires end-of-day submissions and allows a two-day error correction process. And OTS only requires production of information on request, so the SEC’s current proposal would represent a significant change in the way order information is currently gathered and reported. Second, the proposal would require clients to be identified in the audit trail, an item that is not required under any existing order audit trail requirement. Each of these elements would increase the burden and cost of compliance significantly, as compared to existing systems.
Most trading in NMS stocks occurs through highly automated systems. In general, requiring that real-time data be generated by such systems involves some programming, which can be expensive and time-consuming, but is within the limits of practicality. However, many orders, particularly institutional block orders, are traded over the phone with manual system inputs. Some smaller broker-dealers (there still are small broker-dealers trading NMS stocks) do not use automated systems, but key trades into OATS by hand. These broker-dealers are required to use synchronized clocks to register the timing of orders and trades, but they will never be manually input into OATS in anything approaching real time. At a minimum, achieving some semblance of real time would require that an employee be devoted to inputting orders and trades as they occur, rather than mopping everything up at the end of the day, which is the current practice for OATS compliance. It also would be necessary for someone to review trade inputs throughout the day to catch errors, particularly when orders are input manually by error-prone humans performing boring tasks. The expense of review increases exponentially as the timing of review approaches real time.
In the markets for OTC equity securities, most orders and trades are input manually. Although automation is making some inroads into these markets, real-time order audit trails are impractical at the present time, unless the SEC requires the broker-dealers operating in these markets to invest in expensive systems. The result of such a requirement would be many fewer small broker-dealers and increased concentration within the industry.
On the other hand, a client identification requirement would have the biggest impact on the largest broker-dealers. Most institutional investors, and a sizable number of retail investors, use several broker-dealers. Client identification would require that each client have a unique identifier–something like a social security number or taxpayer ID number–that could be used by multiple broker-dealers. For competitive reasons, clients generally wish to avoid having their broker know that they use other broker-dealers, and they certainly would resist having their broker know the names of the competitors for their business. Broker-dealers also do not wish to have their clients known; client lists are a competitive tool, jealously guarded by most broker-dealers, even though it is likely that everyone is calling on the same accounts. A common client identifier will raise the probability of information leakage, resulting in occasional scandals and loss of confidence in the systems for protecting valuable proprietary information.
FINRA has reacted to the SEC’s plans by proposing to extend OATS to all orders in NMS stocks, which would at least sweep orders entered into BATS and NYSE ARCA into its grasp. The extension of OATS should have the virtue of making the information gathered useful. That said, anecdotally we have heard that it is very difficult to search for information in OATS.
This brings us to the most difficult aspect of the SEC’s proposed order audit trail. The securities industry generates an awesome amount of information. Gathering the information is one thing. Sifting through the information haystack to find the useful needles buried within is another matter, and that requires extremely robust computing power. I suspect that this cannot be accomplished without using systems that rival the type of computer technology used by scientists to study global warming or astronomy. Assuming for the sake of argument that such systems exist within the limits of current technology, my guess is that they are very expensive to own and operate. Will the SEC mandate that such systems be maintained by the industry, or is it planning to ramp up its own data-mining capabilities in-house?
There is no doubt that a fully integrated, real-time order audit trail system for all securities, tracking the identity of every client and every event in the lifespan of every order would be a useful regulatory tool. But, such a system would be extremely expensive, extremely problematic for confidentiality and, worse yet, may not work.
Stephen J. Nelson is a principal of The Nelson Law Firm in White Plains, N.Y. Nelson is a weekly contributor and columnist to Traders Magazine’s online edition. He can be reached at sjnelson@nelsonlf.com
The views represented in this commentary are those of its author and do not reflect the opinion of Traders Magazine or its staff. Traders Magazine welcomes reader feedback on this column and on all issues relevant to the institutional trading community. Please send your comments to Traderseditorial@sourcemedia.com
