U.S. Securities & Exchange Commission
SEC Seal
Home | Previous Page
U.S. Securities and Exchange Commission

Speech by SEC Commissioner:
Remarks before the Boston Securities Traders Association


Paul S. Atkins

U.S. Securities and Exchange Commission

Boston, MA
March 9, 2005

It is a great pleasure to be here in the City of Champions. My 11 and 9 year-old sons are big Red Sox fans. So, our household celebrated along with you. Perhaps the best aspect of getting out of Washington and coming up to Boston is that here when someone talks about "SOX", they are not talking about the Sarbanes-Oxley Act! Before I get too far, however, I must make my own compliance folks happy back in Washington and say that the views that I express here are my own and do not necessarily represent those of the Securities and Exchange Commission or my fellow commissioners.

This dynamic marketplace that the SEC supervises is in constant flux, as any market should be. One notable dynamic has been the growth of the clout of institutional investors. The growth in institutional investing over the last forty years has increased the market power of these players through the sheer volume of their demand for various services, such as financial products, research, execution, clearance and settlement, securities lending and other financial services. Institutional investors have become the most prominent and profitable clients on Wall Street. That contrasts with the banking industry, where the retail customer has lately become king again. In what might seem now like ancient history, relationships drove business transactions and fund money may have been allocated without negotiation. Today's competitive markets require all institutional representatives to spend their clients' money wisely, lower expenses and place their fiduciary responsibilities above all else.

Institutional investors have historically used this market power to promote market structure changes. It is in their interests, after all, to have fast, efficient, liquid, and transparent markets. Negotiated commissions and the evolution of the Nasdaq market represent two momentous market reforms that were brought about by institutions promoting competition. By the 1970s, institutional investors had diverted sufficient order flow away from the NYSE that they undermined the fixed commission schedule and created the third market. The SEC followed with regulatory action in 1975 only after huge holes had already been knocked into those anti-competitive practices.

Similarly, ever-growing institutional use of ECNs, which picked up following the adoption of the Limit Order Handling Rules in 1996, ignited the competitive pressures that forced Nasdaq to fully automate. Both feats served individual investors by not only lowering transaction costs but also increasing the transparency and accountability in the marketplace. This trend continues as commission rates drop and institutional investors participate directly in the markets. I hope that institutional investors will continue to represent their clients' interests zealously as the future market structure begins to unfold. Increasing market efficiency through market based solutions rather than government-structured intervention can further individual investors' long-run interests as well.

We have seen, especially in the area of market structure, that SEC rulemaking can take years to design and implement. By the time a regulation is adopted, market forces may already have solved the problem. Thus, it is almost a given that the regulator lags behind as the market continues to innovate and evolve. Regulated areas also sometimes can develop a syndrome where market players tend to rely on the regulator to solve difficult problems instead of seeking market based solutions among themselves. Call it passing the buck.

History shows that market forces and competition may achieve superior results and provide investors with more benefits than those gained through government intervention. Government can provide the basic framework of laws and obligations for fair dealing and then police for compliance with those regulations. But ultimately, market participants should leverage their market power to promote competition and innovation in the marketplace. They are the ones who know the lay of the land best.

About a year and half ago I met with a number of institutional traders to discuss market structure reform. Those traders implied during that meeting that SEC rulemaking should require floor-based exchanges to automate. I don't believe I satisfied your brethren with my response, but I will provide you with the same advice I offered them anyway: use your market power, and order flow, to create competitive change. Because so many barriers to competition are in fact created by our own SEC rules, the SEC should stick to trying to remove barriers to competition that create unfair advantages. That is our true strength, and we would have plenty to do just in addressing that.

With that philosophical approach, I would like to review a little bit of history by discussing several issues that served as kindling for the SEC's Regulation NMS, which is our four-part stab at market structure reform. Particularly, I want to talk about how this proposal has morphed to address issues not originally contemplated, as well as the status of the rulemaking.

In 2002, the initial call for market structure reform began when market participants questioned certain perceived regulatory inequalities and market inefficiencies. They focused on competitive issues associated with the integration of manual (floor based) markets and automated markets. At that time, the exchanges voiced concern over Island's ETF trading program, which fell outside of the national market system. Problems in the Nasdaq market emerged when a manual exchange began trading Nasdaq securities. Other issues on the Nasdaq side included ECN access fees, locked / crossed markets and subpennies. Nearly three years later, the market resolved these integration problems prior to, and mostly without, SEC action.

During this time, though, we ploughed ahead and proposed Regulation NMS, including a market-wide, SEC-imposed trade-through rule focusing on fast versus slow quotes (first with an order-by-order opt-out, now without one), outlawing subpenny quotations for must stocks, prohibiting locked and crossed markets, and changing the market data formula, but not its approach. Never mind that this Regulation NMS now addresses many issues that no longer exist and even some problems that never existed. After all, sometimes government bodies make regulations not because they are needed, but because they can. For instance, proponents of Regulation NMS suggest that a uniform trade-through rule should be adopted because the advancement in technology allows compliance with such a rule. Political pressures also lead to regulatory actions, and in this case recent events involving governance issues in New York provided the political fire to justify Regulation NMS's adoption.

The first major market structure issue to require SEC attention involved Island's trading of ETF securities. Island garnered tremendous market share from the Amex while operating outside of the national market system. While Island was innovative and competitive, some market participants argued that its lack of transparency and the difficulty in accessing Island's orders hurt investors. Island argued, however, that it could not operate under the cumbersome ITS operating regime. Island remained outside of the national market system even after SEC regulations required otherwise.

Island's increasing market share demonstrated investors preference for an automated market to trade liquid ETF products. Investors voted with their feet and shifted volume from the manual exchanges to automated markets. For the first time since the early 1970's, true competition for trading exchange-listed securities emerged. Island's refusal to enter the national market system due to the anti-competitive nature of the trade-through rule produced a firestorm of debate within the industry and created a policy dilemma for the SEC. The ECNs argued that the trade-through rule was anticompetitive and protected manual exchange volume. The exchanges argued that the trade-through rule offered important investor protection against executions at inferior prices. Finally, some investors wanted the freedom to determine how and where to execute their own transactions. The SEC split the baby by adopting in 2002 a temporary ITS de minimis trade-through exemption for certain ETF products. This de-minimis trade-through rule facilitated the integration of automated and manual markets in ETF products. Thus, the SEC brought ECNs into the national market system for ETFs without undue burden.

Like the exchanges, Nasdaq experienced growing pains during its transformation into a fully automated market. Inadequate intermarket linkage and high access fees created inefficiencies, such as crossed markets, that aggravated some market participants. Ultimately, the access limitations dissipated when the marketplace rejected floor-based trading of Nasdaq securities and the individual market centers, including SuperMontage, INET and Archipelago, modernized access by obtaining and utilizing outbound routing technology to interact with each market's orders.

Market competition and SRO rulemaking reduced access fees to new lows, drove most subpenny quotes from the market and even centralized Nasdaq's openings and closings. The efficiencies generated in the Nasdaq market over the last few years lead to a significant decrease in internalization while the extremely competitive Nasdaq market spurred consolidations which reduced market fragmentation.

On the listed side, underlying structural problems came to the fore in 2003 when the specialist scandal spurred industry demand for the automation of the NYSE. NYSE responded competitively by announcing the development of the hybrid market. The hybrid market automates a portion of the specialist's book and is a first step toward addressing investors' needs. These events, and others, provided the impetus to push forward with Regulation NMS. The contentious part of the rulemaking boils down to whether competition should force the automation of the floor-based exchanges or whether the exchanges should be guided to automation through regulatory changes, with all of the possibilities of people gaming the rules as we have seen over the years. Of course, this debate cannot be held without discussing the trade-through rule.

The old trade-through rule certainly appears to have created barriers to competition in exchange-listed securities. Ample evidence is provided by comparing the current state of the Nasdaq market with the floor-based markets. Although the exchange market structure has become automated to some degree over the past 30 years, it essentially has not evolved since 1975. Competition among the ECNs within the Nasdaq market, by contrast, corrected market inefficiencies prior to the implementation of government regulation. However, some market participants claim that competitive forces cannot and could not create similar change on the listed side. This claim is based on the different nature of the quotes - the exchanges use mostly manual quotes versus the other markets' automatic quotes. The application of the trade-through rule, which is based solely on price, prevents effective competition between markets when quotes are not comparable.

The trade-through rule debate extends beyond the anticompetitive nature of the rule. It brings forth the question of whether, and if so when, government regulation should supplant investor judgment and independent pricing mechanisms.

But, we must remember that government regulation might not be the panacea that some expect it to be. For example, proposed Regulation NMS has only very general parameters as to how one characterizes access to a market's quotations - that is, what is a "fast" or "slow" quote. Determination of these access standards will be a vital component to the effectiveness of the new trade-through rule. As with all rules, implementation is key.

The proposed Regulation NMS provides the basic framework for access standards, but the SEC Market Regulation staff will be responsible for filling in the blanks following rule adoption. How will they go about this? What sort of lobbying behind the scenes will take place? Who will be accountable for the decisions, especially if they are determined without action by the 5 commissioners, and just by the staff in what we call "informal" rulemaking? This sort of standard setting may prove to be inefficient and slow compared to calibration through a competitive marketplace. Adoption of the trade-through rule would allow the SEC staff to control the pace of automation and competition through their control of the efficiency and effectiveness of access to markets.

We have seen through history how ineffectual and counterproductive government control of a market's pricing mechanism can be. This goes for broad-based controls (like President Nixon's wage and price controls) to more limited controls over trucking rates by the now-defunct Interstate Commerce Commission. The securities markets are currently efficient and robust. Competition narrows spreads, reduces internalization and increases the execution quality of all investors' orders. When government tries to influence price-setting, buyers and sellers lose the ability to get critical information for their transactions. A market economy functions best when buyers and sellers can make their preferences clear. The government is not well-suited to set prices, since it can only guess as to buyers' preferences.

Yet, the SEC continues to push forward in its efforts to become the market's price discovery mechanism. Just today, Chairman Donaldson outlined the SEC's reasons for adopting a uniform trade-through rule in his testimony before the Senate Committee on Banking, Housing, and Urban Affairs. The Chairman emphasized three important policy "goals" that would be furthered by a trade-through rule: 1) it would serve as an effective backstop, on an order-by-order basis, to a broker's duty of obtaining best execution for market orders; 2) it would promote investor confidence in the markets because trade-throughs may provide retail investors with an impression of unfairness; and 3) it would encourage the use of limit orders and thereby contribute to greater market depth and liquidity.

All three policy goals are commendable. I cannot find fault with these stated objectives. But I object to these goals for the same reasons I voted against publishing the reproposed version of Regulation NMS - they just won't be achieved. These objectives theoretically can be achieved by developing a central limit order book but that alternative is not seriously before the Commission, since it seems that the depth-of-book alternative in our reproposal does not have majority support among the commissioners.

The same policy arguments that I made against the voluntary depth-of-book proposal apply equally to a top-of-book trade through rule. The difference between the two alternatives is that the concrete results under the top-of-book approach are even more negligible than those the SEC would obtain with the depth-of-book alternative. This appears to be the first incremental step toward the implementation of a central limit order book.

The SEC waves the customer protection flag to support the rule's adoption instead of providing empirical evidence showing that it is needed. Retail investors may, in the end, be harmed by market inefficiencies the regulation may introduce. These inefficiencies could include increased transaction costs, increased spreads, decreased price transparency and an impression of infallible safety.

The first policy goal that my colleagues seek through a trade-through rule is an effective backstop, on an order-by-order basis, to a broker's duty of obtaining best execution for market orders. Currently, best execution standards protect retail and institutional investors in the Nasdaq market. Existing SEC and SRO examination programs closely monitor broker-dealer compliance with best execution principles. So, I would suggest that mandated pricing on an order by order basis appears to be redundant. Neither Chairman Donaldson's testimony nor Reproposed Regulation NMS provides evidence that the current harm to investors outweighs the cost and unnecessary inefficiencies that will be introduced to the market through regulation.

In fact, the trade-through study prepared by our SEC economists cited very minimal trade-through rates. Even if these numbers are accurate are we really risking unknown market impact by adopting a rule that by our admission affects only about 2% of trades? One can infer from these statistics that brokers are indeed fulfilling their best execution obligations. The total cost of this regulation is difficult to quantify, but it certainly will include the cost of implementation, compliance, and other costs associated with unintended consequences and market inefficiencies. Retail investors will bear these costs directly or indirectly perhaps through wider spreads and higher brokerage commissions.

The second policy goal indicates that retail investors "may feel unfairly treated" if the market trades through their limit orders. One must assume that if retail investors do feel unfairly treated, that this impression will exist unless all orders are executed according to strict inter-market price-time priority. The fact remains that the SEC has not presented evidence that proves that investors do feel unfairly treated by trade-throughs. In fact, comment letters submitted by large retail firms such as Schwab, Ameritrade, and E-Trade suggest that the trade-through rule should be repealed. These brokers believe that competition protects investors far better than a trade-through rule. I question whether it is in the public's interest, or good public policy, to adopt regulation based upon anecdotal evidence.

Finally, the last policy goal provides that the trade-through rule "is designed" to encourage the use of limit orders and thereby contribute to greater market depth and liquidity. The SEC reasons that the additional depth and liquidity will help institutional investors execute their large orders. This argument is tenuous at best because the source of this newfound liquidity will be the institutional investors. I question whether institutions will increase displayed liquidity when free-riding still exists.

Regulation NMS does little to address this concern and it is hard to see how top-of-book protection will encourage institutional investors to post large limit orders. Furthermore, the SEC itself acknowledges that free-riding is often the reason why investors do not display limit orders in the market. Reproposed Regulation NMS itself puts a further shadow on this policy goal by noting that investors used more limit orders on Nasdaq than on the New York Stock Exchange. This fact is not surprising since institutional investors pioneered the use of ECN's to lower transaction costs in the Nasdaq market.

I am surprised, however, that institutional investors have not stormed the SEC to demand repeal of the trade-through rule as a way to replicate Nasdaq's competitive market structure. We are looking to you as controllers of liquidity for answers on the appropriate market design but we don't seem to be taking into account the competitive choices that you have made over the last few years. You must rely on yourself with order flow in hand to obtain the results your clients deserve.

The SEC should not adopt a regulation simply because its stated goals are noble. History is riddled with bad laws and rules that were "designed" to do something noble. Identifying a goal does not guarantee that the steps that you are taking will achieve that goal. I can put myself on an ice cream only diet with the goal of losing weight. It is unlikely that my goal will become reality because my diet methodology is flawed. Dr. Atkins would back me up here.

Regulation NMS can be contrasted with disclosure regulation, where the outcome of the regulation is clear- an increase in the dissemination of information. By adopting the trade-through rule, the government requires market participants to act in a certain fashion. Why should we adopt merit regulation where the benefits are dubious and the unintended consequences are unknown? The concern about unintended consequences is justified given that there has never been a trade-through rule on Nasdaq and the trade-through rule has never been enforced in the exchange-listed markets. Are we not putting at risk the world's best securities markets in this process?

The SEC should give competition a chance to design the markets by allowing NYSE's hybrid market to go forward and operate without the trade-through rule. Regulatory intervention is inappropriate at this time. It may stall the advancement of our floor based markets and prevent competition from developing the most efficient markets possible. Our capital markets are too important to allow the implementation of regulation when market participants, academics and staff members cannot provide a consistent answer as to the impact of such a rule on transparency, market behavior, or the independent pricing mechanism.

I would like to thank all of you who took time to write comment letters on Regulation NMS, the Supplemental Release and Reproposed Regulation NMS. Your letters can make a difference in shaping this rulemaking and future Commission initiatives and can be utilized for beneficial purposes in the years to come.


Modified: 03/11/2005