Prepared Remarks of Paul S. Atkins
Thank you for the invitation to participate on this panel today. It is a pleasure to be here. I believe that this is the first time that I have appeared at a formal SEC-related event since I left the Commission in 2008. I am sorry that I cannot see you all in person at 100 F Street. This Committee has an important mission to give considered input to the Commission, so as a taxpayer I thank you all for your service, and especially Jennifer Marietta-Westberg for her service as chair of this committee, and those of you for whom today is your last meeting. I should add that it is also wonderful to see the various members of the Commission here today.
Since I first came to the SEC in 1990 to work for Chairman Breeden and then Chairman Levitt, my perspective on debates around equity market structure is that the more things change, the more they stay the same. The basic themes of the debate are pretty much the same now as 30 years ago. In 1975 (when, I am happy to say, I was still a kid), Congress identified five key components of a properly functioning “national market system”: efficiency, competition, price transparency, best execution, and order interaction. I would argue that today’s U.S. equity markets rate very well in all areas, particularly considering treatment of trades executed for individual investors.
Technology has achieved incredible efficiencies measured by speed and accuracy. There certainly is robust competition in the marketplace (although among a diminishing number of participants). There has never been such a level of transparency as we have now through disclosure of fees and costs. Interaction of orders as measured by tight spreads, interconnectedness among trading venues, and execution speed is beyond the wildest dreams of 30 years ago. Finally, with respect to best execution, individual investors today obtain overall sophisticated execution quality at incredibly low cost, with prices typically better than the best displayed prices.
That is not to say that things cannot be improved. Many of the problems that we have today in the markets stem directly from the unintended consequences of regulatory policies. This sort of situation usually results when policymakers have seen a fast-evolving market and then attempt to impose new rules rather than take a considered approach to the significant, long-term downside of such rules.
In 2005 this was the crux of my and fellow Commissioner Cynthia Glassman’s dissent on the adoption of Regulation National Market System. Rather than defining the fundamental obligations of market participants, such as clarifying brokers’ duty of best execution, lowering barriers to competition, and improving access to quotations, Reg NMS sought to dictate very specific processes to address perceived ills on still-emerging and still-evolving technologies and tactics by which equity markets executed trades.
Upon Reg NMS’s approval and its implementation in 2007, its supporters claimed that Rule 611, or the Trade-Through Rule, would create greater market efficiency and transparency, while ensuring that investors received the best price upon order execution. Many positive aspects of the market today, however, have occurred despite Reg NMS, not because of it.
Reg NMS is unnecessarily complex, invites gamesmanship, induces widespread market fragmentation, disperses liquidity, and diminishes transparency. Indeed, we have seen Reg NMS drive order flow away from exchanges, fuel unhealthy trading volatility, and expose the markets to flash crashes because its rules have led to atomization of orders and liquidity among a dizzying array of trading venues. None of these outcomes is a hallmark of the worthy goals of a fair, orderly and efficient market.
I would highlight several areas of current concern.
First, Reg NMS and its Rule 611 have not served investors or broker dealers particularly well. Orders today are atomized not necessarily because large traders do not want to show their hand, since they work their large orders mostly on ATSs and elsewhere (just as they did in the old days in the upstairs market), but because of the gamesmanship by those that seek to take advantage of the Reg NMS structure and the arcane rules that have been adopted by the various trading venues. This has also resulted in an increase in the number of trading venues to almost 300, including internalising brokers, as well as the now 150 or so order types that can trap the unwary.
Second, NMS has fostered a near-exclusive focus on price as the measure of execution quality, following the old adage of “You get what you measure” under Rules 605 and 606. Thus, I fear that broker-dealers prioritize price above all else in their execution decisions, regardless of the needs or preferences of their customers. This leads not to order protection, but to order discrimination. Thus, the so-called “Order Protection Rule” is a true misnomer.
Worse, the basic measure of pricing and best execution, the National Best Bid and Offer, is a flawed standard. Prices are like snowflakes – they are quintessentially ephemeral, and certainly are not ever exactly alike. The concept of “best execution” depends on many factors, such as certainty, speed, but most importantly SIZE. Just because the top-of-book price at one exchange for 200 shares is the NBBO, does not mean that the same price is good for 500, 1000, much less 100,000 shares to be traded on another exchange. So, given the current market structure, is it not odd that the NBBO, based on the SIP spread, is still used as the benchmark for market position and best execution? In fact, does it even make sense to apply it to retail orders, which these days almost never go to an exchange?
Third, Reg NMS contributes to the concentration of market participants. What effect does this have on overall competition and innovation? We may never know for sure, but today there are half the number of broker-dealers and far fewer market makers as 30 years ago. That is directly attributable to costs of doing business, including the costs of meeting regulatory obligations.
Firms are locked in a high-tech arms race for speed, directly attributable to provisions of NMS as well as customer demand. When regulation is inefficient and causes distortions, that necessarily raises costs and drives competitors out of the market. A market regulator like the SEC must question what increased concentration means for the stability of the markets and the financial system overall. Has it pushed trading facilities towards increased homogenization and driven exchanges, in particular, to try to differentiate themselves in ways that do not necessarily enhance issuer and investor choice or serve their interests?
Last, the SEC is now the gatekeeper for approving data fees, which are lightning rods for increased costs to market participants and significant litigation. The monopolistic market data regime that the SEC has allowed to develop, including through the NMS regime, is yet another driver towards market distortions, increasing costs, and the resultant concentration of market participation. For best execution and to meet customer demand, a broker-dealer must pay prices that are not set by market forces.
Given some of the volatility in the capital markets that we have seen recently, as well as emerging retail investor trading platforms, there is an opening for the SEC to do what it should have done 15 years ago and take a comprehensive view of equity market structure.
This review should be part of a formal rule-making process and include a hard look at Reg NMS and its effects the prescriptive approach it had had on the markets, including outright removal of outmoded Rule 611. The last such detailed study was the Market 2000 Report that began in 1992 with a Market Regulation Division concept release under Chairman Breeden and concluded with the report that we released under Chairman Levitt in 1994.
In taking such an assessment, the SEC will fulfill the worthy goals set forth in its mission to oversee fair, orderly and efficient markets.