March 17, 2004
In the current lively debate about market structure one of the most hotly debated topics is the continued viability of the trade-through rule, which requires that any market participant trading through a better-priced order must satisfy that order if requested to do so, resulting in listed stock orders by and large being sent to the marketplace with the best price. Proposals have ranged from outright repeal to relaxing the rule to permit de minimis exceptions and/or to create a bifurcated market in which so-called "fast markets" can trade-through "slow markets", but cannot trade-through other "fast markets". The most common complaint put forth by the advocates of weakening of the rule is that it is anti-competitive and anachronistic, and compliance with it in the context of the manual nature of the NYSE prevents traders from executing transactions with the speed and certainty they desire. Advocates of maintaining the rule, on the other hand, insist that "best price" and investor protection necessitate its preservation. Unfortunately, the basic issues surrounding the trade-through rule have become obscured in contrived complexity, in part by ECNs, certain large broker-dealers and others who stand to benefit from the repeal or significant weakening of the trade-through rule.
Before we begin our analysis of the trade-through rule we thought it would be useful to describe a few important facts about our background that we believe distinguish us from many others weighing in on the advisability of the trade-through rule. First, we are not Luddites, as some might like to characterize all floor brokers. We firmly believe in automation. In fact, we were the co-creators of DOT for non-member firms in the early 1990s and currently service numerous quant and portfolio trading customers whose execution needs demand sophisticated technological and algorithmic tools. Second, we are pure agents that represent the public in both the listed and over-the-counter (OTC) markets from the floor and an upstairs desk, respectively. Importantly, we never act as principal, never shop or internalize order flow and never accept payment for order flow. Simply put, we execute stock transactions in a manner that deliberately avoids conflicts of interest. Third, while we have been members of the NYSE since 1979, we are not NYSE seat owners, and have been members of the ArcaExchange since its inception. With our sole business being execution, we go to the markets with the best prices for our customers. In the case of OTCs and exchange-traded funds (ETFs), that means we are heavy users of ECNs. In the case of listed stocks, today that means we are heavy users of the NYSE.
Since our only source of income is the commission revenue we receive from our clients, our loyalties are clear. We have for years and shall continue to deliver our order flow and offer our assistance to those markets providing the greatest liquidity and best prices available. For this reason we are strong supporters of the auction at the NYSE, while acknowledging its need to aggressively deal with its structural problems. We also continue our commitment to all alternate trading platforms that promise the potential of better prices. Our goal is to ensure that there is a system in place that protects our customers just as we try to with our own efforts.
With that in mind, let's examine the facts of the way the trade-through rule operates as well as the empirical evidence to dispel the myths surrounding the rationale for the repeal or relaxation of the trade-through rule.
One of the most interesting things about the debate on the trade-through rule is the way in which the debate has been framed. Almost the entire focus of the debate has been whether the NYSE is fast enough to accommodate the demands of traders wanting to hit a bid or take an offer, while the fate of the limit order on the book has been virtually ignored. While certain investors might be willing to sacrifice "best-price" in favor of other execution variables, such as speed and/or certainty of execution, it is important to recognize that the trade-through rule is not really about the order being executed but the order that is not executed, i.e., traded-through. The advocates of trade-through repeal are prioritizing the interests of the order that wants to trade-though over the order that would be traded-through. In some sense the approach is one that looks at each order in isolation from other orders, not in the context of overall order interaction of all buyers and sellers with each seller competing for the highest price and each buyer for the lowest. Requiring the interaction of all buying and selling interest present in the market at a particular point in time is important for the integrity of our markets and the perception by investors that they are fair. One of the primary goals of Congress' creation of the national market system in 1975 was to make sure that all investors, irrespective of their connections, affiliations and locations, were provided with the same equal and fair access to market information and pricing. We cannot simply accommodate the interest of one market participant to the detriment of all others.
We are believers in free markets and competing marketplaces, but just as advocates of free speech must acknowledge that a person falsely shouting, "Fire!" in a crowded theatre should be liable to a person injured in the ensuing chaos, we believe certain investor protection principles must be held above free market principles. In fact, investor protection should be uniform across all markets and should not be limited to listed markets in the case of the trade-through rule. We welcome recent speculation that the SEC proposal expected in late February/early March on trade-through may actually suggest a universal trade-through rule, expanding the applicability of the rule to include Nasdaq stocks as well.
While we believe the debate has been improperly framed to a great extent, we nevertheless will examine the arguments put forth by advocates of trade-through repeal/weakening in their own terms and dispel the additional myths that are central to their line of reasoning.
The most common analogy being proffered by detractors of the trade-through rule is the difference between a local convenience store and a supermarket. As consumers, we often opt to buy milk or other items at a local store which might have higher prices because of the convenience arising from its proximity to our home rather than traveling an extra mile or two to the grocery store. Similarly, advocates of the repeal of the trade-through rule argue that investors should be permitted to trade-through a market with superior prices advertised (the supermarket) because of the speed and certainty of execution in another market (the convenience store). More directly stated, investors are "big boys" and should be free to choose their own definition of "best execution"-- if speed is what they choose, so be it and get out of their way. While this line of reasoning is compelling on its face and "best execution" indeed can be multi-faceted and not limited to simply best price, the analogy is misleading.
The purchase of a carton of milk or a loaf of bread is not surprisingly very different on many levels than the purchase of a stock. Two key differences stand out, however. First, a consumer buying food has the ultimate goal of sustenance for him/herself or his/her family. Price and ease of purchase considerations impact the decision as to where the purchase is made. Investors, on the other hand, have the end-goal of making money. Therefore, getting the best price is much more central to the entire undertaking. Second, in the case of buying stocks, the ultimate beneficial owner of the stock in the vast majority of cases is not the person actually making the purchase. An entire regulatory structure has been put in place to protect these beneficial owners from the intermediaries who represent them in these transactions. These intermediaries can be broadly grouped in to two categories: (i) broker-dealers representing institutions or retail investors and (ii) institutions representing their fund holders.
Sadly, broker-dealers all too often abuse orders they are given, benefiting themselves, not the client. Trading at the wrong price can be a profitable enterprise for broker-dealers when either trading as principal with clients (and therefore earning a greater profit if they match their inferior priced order with that of their customer) or when earning a commission from both the buyer and the seller (and therefore having an interest in bypassing the better-priced order so they won't lose the second commission). In each case, there is a disincentive to explore the markets for better-priced interest, and instead internalize the order flow. Abandoning the trade-through rule will likely contribute to the dangerous trend of increasing internalization by broker-dealers. Similarly, institutions have fiduciary responsibilities to their fund holders whether they are using brokers or directly accessing the markets. While institutions are given discretion by their end customers, this must be exercised responsibly and "best execution" guidelines followed.
The job of regulators is to construct and apply rules that help ensure compliance with these fiduciary duties. The system itself must be designed with integrity; otherwise the lowest common denominator will define the industry. We have witnessed this with respect to broker-dealers in the Nasdaq dealer scandals in the early 1990s and more recently in the research scandals and NYSE specialist investigations, and now we are seeing it with respect to the mutual fund industry in the market timing scandals. In each of these cases, the system itself was weak, and overly-dependent on the integrity of the individuals comprising the system, which is always a recipe for disaster.
The trade-through rule is one of the building blocks of the system, a guarantee that all customers will be treated fairly by the intermediaries serving them. We can think of no compromise in the trade-through rule that wouldn't result in a serious weakening of the integrity of our markets.
Returning to the supermarket analogy we began with, it just may be that requiring broker-dealers and institutions to walk that extra mile to get the best price is a sensible protection for the investing public.
The complaints about the trade-through rule being an impediment to both investor trading strategies in listed stocks and competition from other exchanges/ECNs reflects a fundamental misunderstanding of the way the rule is actually applied on the NYSE floor itself and the way competition has already developed. In addition, these complaints are unsupported by statistics on the NYSE and its competitors' fill rates.
Based on our firm's 25 years of trading on the floor, it's hard to see how the trade-through rule is in practice much of an impediment to effectuating trading strategies or competition by other market centers. If better-priced orders displayed through the Inter-market Trading System (ITS) are small, they are in fact simply traded through and then filled after the larger trade happens on the Exchange since ITS rules mandate that any market whose better-priced order was traded-through can demand an execution within five minutes of the trade-through. On the other hand, if the better-priced orders are large, the fiduciary responsibility of the floor broker will require him to trade with that better-priced order because the price difference would be meaningful. Thus, with respect to smaller orders the argument that the investor is merely being slowed down by the trade-through rule turns out to be specious since in reality the investor is not slowed down, while for larger orders the trade-through rule ensures that the floor broker complies with his fiduciary obligations to his clients.
The example of the trading in Lucent Technologies (LU) is instructive in this regard. Suffering from both the end of the telecom boom and its own company-specific travails, LU's stock has fallen to the single digits and into the "penny stock" category. However, it remains an ultra-liquid stock, trading approximately 50 million shares a day. This volume is spread across multiple marketplaces. In fact, the NYSE's market share in the stock has dipped below 50%, while in most stocks the NYSE's share is approximately 80%. The trade-through rule has not proven to be any impediment at all to the NYSE's electronic rivals in taking away LU market share. What seems to have done the trick has been competitive pricing, driven by a low-priced stock with high volatility. Day traders and retail investors in particular have been active in the name and are more likely to trade over ECNs either directly or through the online brokerages that use the ECNs and the third market.
There is only one class of investors who in our opinion have a strong case as to the trade-through rule negatively impacting their trading strategy-- arbitrageurs (and, in their most popular incarnation today, stat arbs). Often the prices at which their trades are filled are secondary to speed in the view of the arbitrageur. An arbitrageur may justifiably not care about the particular price of either leg of a trade (as long as it is within certain parameters that still make the combined trade profitable). But what about the interest of the better-priced order that would in fact be traded-through if the trade-through rule did not exist or were substantially weakened? Also, we would note that the NYSE has provided the vehicle of Direct+ for automatic executions for 1,099 shares or less, and the stat arbs seem to be prospering with this offering. Besides, they are only one class of investors (and some would argue not one the market and the NYSE should put on par with retail and institutional long-term investors) and we should only go so far in accommodating them.
One extension of the supermarket analogy put forth by some of the NYSE's competitors is that not only are investors forced to trade on the more inconvenient (i.e. slower) exchange in the pursuit of "best price", but often when they arrive there the advertised price is actually gone and there is no longer a way to satisfy their order. This would be a very troubling eventuality if it were true, but, unfortunately for the critics, it simply isn't. Fill rate statistics belie the argument. Certainty of execution is actually more likely at the NYSE than at its peers. According to publicly available Rule 11Ac1-5 data, the NYSE actually has the highest fill rates in listed stocks of all market centers, with 82% of all marketable orders filled during 2003 compared to 43% at Arca, 36% at BTrade, 32% at Instinet, 26% at Brut and 18% at Island. Interestingly, these ECN fill rate percentages are actually reduced dramatically to the teens and low single digits when the routing they each do to other market centers to fill their orders is excluded (other than Island which does not route away), which ironically in many cases means their fill rates include routing to the NYSE. The speed of ECNs, with quotes updating so fast in a fragmented, decimalized world, may ironically make certainty of execution even less likely in their systems.
Speed and certainty of execution are quite distinct. However, even the speed complaint rings a bit hollow at the end of the day. The average DOT market order of 2,099 shares or less is actually filled in 13.7 seconds, while still offering the opportunity for price improvement from the crowd. To the extent an investor is unwilling to wait even that short amount of time, the NYSE has been offering an auto-execution facility called Direct+ since late 2000 that has an average time to execution of .8 seconds for a guaranteed execution. Direct + actually accounts for seven percent of consolidated volume in NYSE-listed equities today, more than that executed by all ECNs combined. If the SEC approves the recent NYSE proposal to eliminate the 1,099-share size limit for Direct + orders, the 30-second limitation for consecutive orders and the restriction that all Direct + orders be limit orders, the self-interested nature of the ECN claims about the necessity of repealing the trade-through rule will become even more obvious. In fact, if a "fast market" vs. "slow market" proposal is adopted and the Direct + proposals approved by the SEC, the NYSE should actually be considered a "fast market" in light of the broad availability of guaranteed, speedy executions on the NYSE through Direct +.
It is worth taking a moment to examine the SEC's three cent de minimis exception to the trade-through rule that applies to ETFs since it is cited alternatively to support the notion that ECNs will be successful in trading listed stocks so long as they are relieved of the burden of the trade-through rule and as an interim step that should be adopted on the road to repeal of the trade-through rule. While it is undeniable that ECNs have done a great job in grabbing ETF market share from the exchanges, the fact is ETFs are a very different animal than single stocks. Accordingly, the ETF experience with the trade-through rule should not be considered at all analogous.
There are important distinctions between traditional stock listings on the NYSE and the Big Three ETFs (QQQ, SPY and DIA) that are listed on the American Stock Exchange, but also traded elsewhere through the unlisted trading privileges (UTP) rules. First and perhaps foremost, an ETF by its nature is a derivative product, not a stock in the traditional sense. Accordingly, an ETF's value precisely tracks its related index and underlying component stocks. The price of an ETF is even quoted in terms of a net asset value calculated by the listing exchange. In short, the prices of ETFs are quite transparent in comparison to that of an individual stock where a wide-range of investment opinions about a company's competitive position, earnings and growth potential, and appropriate valuation makes determining the appropriate trading price a function of the myriad opinions in the market.
Second and related to their derivative nature, ETFs trade very differently than other listed stocks. While true to their marketing sales pitch they do "trade like stocks" in the sense that they can be bought and sold intraday unlike an index mutual fund, they often are traded out of the derivatives groups of the large broker-dealers rather than their cash equities units. Because of the transparency of their pricing, the caution with which dealers typically commit capital on equity trading desks resulting in wide markets seems cast aside and the desks typically offer narrow-spread, deep two-sided markets in the Big Three ETFs. Even in a post-decimalization world, one does not hear the complaints about ETF liquidity that one routinely hears about the rest of the universe of listed and OTC stocks. The Big Three ETFs have unparalleled liquidity.
The argument that the success of the de minimis exception experiment for ETFs demonstrates that it should apply to all listed stocks is a red herring-price discovery still counts for listed securities.
While ECNs and the innovations they have brought have cleaned up the Nasdaq market and benefited investors immensely that does not mean that basic investor protection rules should be abandoned in the name of technological progress or speed. Automation is a powerful tool that should be harnessed to enhance the value proposition to investors, not pursued as an end in itself. To the extent NYSE technology and rules are felt inadequate to deal with the speed required by investors today (even after the latest NYSE proposals on Direct + are implemented) or somehow give the NYSE an unfair advantage, the SEC should require that technology and rules must be adopted to deal with that reality. Similarly, to the extent the ITS system needs to be re-visited to reflect current market realities, let's explore that. But let's take these arguments from self-interested parties about the necessity of the repeal of the trade-through rule along with their rhetoric and clever analogies with a grain of salt and examine the facts first.