From: George Rutherfurd
June 15, 2005
The NYSE letter of June 7, 2005 is seriously inadequate as it does not address in any meaningful way the broad policy objections to its proposal made in my previous correspondence, nor does it correctly address my specific technical objections to the manner in which the proposed amendment is drafted. In fact, one of the NYSE's examples as to how its market operates is so ludicrously (and comically) wrong that it is obvious that the NYSE simply doesn't understand what it is doing here.
And, to add insult to injury, the NYSE managed to misspell my name on no fewer than seven occasions!
The Policy Problem: Orders Not in the Market at the Time of a Trade Are Not Entitled to Participate in That Trade
The NYSE's proposal is quite simple, really. If the specialist has traded as principal, but before the trade can be reported the specialist receives a superdot order executable at the price of that trade, the specialist must "yield" to the superdot order (i.e., permit the superdot order to take the specialist's place in the trade).
This seemingly straight-forward proposal is, in fact, a conceptual mess. The NYSE is actually proposing that an order that is not even in the market when a trade takes place is nonetheless absolutely entitled to participate in that trade. As I noted in my earlier correspondence, such a notion is both radical and bizarre, and is completely unknown to any securities market anywhere. I indicated that the SEC would have to surmount a tremendous conceptual obstacle and carefully rationalise any new law it might create in this regard, in the highly unlikely event it were inclined to approve the NYSE's strange proposal.
In its rule submission, the NYSE offered virtually no justification for its proposal other than broad, vague (meaningless) references to agency law and the notion that public orders ought to trade with each other. As pointed out in my prior correspondence, this rationale (meagerly presented as it was) would hardly suffice. The specialist obviously has no agency relationship to an order he or she has not even received at the time of the specialist's trade. Concepts such as public orders interacting with each other, and not with the specialist, are relevant only with respect to orders actually in the marketplace prior to the consummation of a trade (not at all the situation posited in the NYSE's proposal).
One would have thought that the NYSE would have attempted, through carefully reasoned legal argument, to at least attempt to make some sort of rational case for its unprecedented proposal. But as is the case with all too many recent NYSE rule submissions, rational argumentation is nowhere to be found. (In my letter of February 18, 2005, I discussed SR-NYSE-2004-06, a particularly egregious example of the NYSE's inexcusable ignorance of its own rules and/or deliberate bad faith in dealing with the SEC staff. See also my comments on SR-NYSE-2004-05, a promising, general concept for enhanced electronic trading (the so-called "hybrid market") severely undermined to date by the inability of the NYSE staff to propose implementing rules that actually make sense).
Which brings us to the NYSE's June 7 letter. In response to my broad policy concerns, the NYSE has dropped its specious agency law/public order interaction rationale, and one should be grateful that this insult to the public's intelligence has been abandoned. Rather, the only justification now on offer for the proposed amendment is the classic "devil made us do it", the devil in this case being SEC auditors (Office of Compliance Inspections and Examinations, or OCIE) concerned about a possible appearance of impropriety in that NYSE surveillance systems apparently cannot distinguish between proper versus improper specialist principal trading. This is a typical auditor problem/response, can't see the forest for the trees type of recommendation. (I mean no insult to OCIE here. This is simply what auditors do). Surely, the legal absurdity of the proposed solution is far worse than what is presented in the NYSE's June 7 letter as, at most, a perceptual rather than actual, substantive problem. (Presumably, whatever may have been the case or not previously, the specialists have all gotten religion by now after being bludgeoned by the SEC settlement orders). Clearly, the solution to whatever the problem in fact may be is enhanced surveillance, not bizarre, radical new law.
The OCIE staff are perceived in the securities industry as highly competent auditors, but they are not charged with evaluating broad market policy and market structure issues. Surely, the sophisticated staff of the Market Regulation Division (for whom I have developed tremendous professional respect over the years) understands the dangerous legal precedent that would be set should the NYSE's rule submission be approved.
The NYSE's Use of the Term "Yielding" Can Only Be Described As Ignorant
The NYSE consistently uses the term "yield" to describe the process by which the specialist gives up his or her participation in a trade to a superdot order that arrived after such trade was consummated. In my prior correspondence, I indicated that the NYSE had misused the term "yield", which is commonly understood to refer to being required to give up an opportunity to effect an impending trade. The term commonly used when someone gives up his or her participation in a trade that has already taken place is "substitution of principals." This is an obvious point (and hardly my principal objection to the NYSE's proposal) and one would have thought the NYSE, if it insisted on pressing its proposed amendment, would have simply conformed the language of the proposal to standard securities industry usage and moved on.
The NYSE's response here, though, is entirely baffling, and can only be described as embarrassingly ignorant. Rather than acknowledging the need to refine its drafting (common enough, and unremarkable, in SRO submissions as well as the SEC's own proposals), the NYSE seems almost hysterically committed to its legally imprecise, historically inaccurate, and unprecedented usage. My objection here is not simply a matter of editorial comment or drafting preference. Word choices matter a great deal as a matter of law in rule proposals, particularly when words have consistent historical connotations that are well understood by seasoned securities industry professionals.
As if in total ignorance of the manner in which U.S. markets (to say nothing of a number of world markets) have traditionally used the term "yield", the NYSE defaults to a broad, general dictionary definition. According to the NYSE, the term "yield" means to "give up possession of; relinquish" and that one must possess something before one can be made to relinquish it (all of which is fair enough, as far as it goes, but obviously devoid of the specific securities industry context as to what it is that is actually being "relinquished"). The NYSE goes on to maintain that this "general meaning" in fact supports its position that the term "yielding" can describe the process of giving up participation in a trade after it has been consummated. The NYSE further (and more to the point, at least in terms of statutory reference and appropriate legal authority) maintains that Section 11(a)(1) of the Securities Exchange Act uses the term "yield" in a manner consistent with the NYSE's proposed rule.
Quite simply, the NYSE's position here is laughable. Neither the SEC nor any SRO (including the NYSE, until the rule submission under discussion) has ever used the term "yield" in the manner contemplated by the NYSE's submission. The term "yield" has always been understood to mean relinquishing the right one would otherwise be entitled to assert to effect a trade that has yet to be consummated, not to give up participation in a trade that has already been consummated. This concept is so well-understood that I have dificulty believing it can be seriously in dispute, but the confusion is manifest in the NYSE's own drafting of the June 7 letter. On page 2 of the letter, the NYSE states, "Under proposed Rule 104.10(11), the giving up of the right to trade - that is the yielding - simply occurs after oral consummation of the trade rather than before." The NYSE thus misses the essential point: under its proposal, the specialist does not give up the right to trade (classic yielding), as the trade has, in fact, already been consummated. Rather, what the specialist is being asked to give up is his or her continued status as the contra party to the contract created by the consummated trade. This is obviously a very different issue, as a matter of law, from the right to have consummated the transaction in the first place, which is what yielding is addressed to. Given the disparate legal principles involved, this is hardly, as the NYSE would have it, a "distinction without a difference."
Anyone in doubt here need only look to the plain language of Section 11(a)(1)(G) of the Securities Exchange Act, the most prominent (and relevant) instance of the use of the term "yield" in the U.S. securities regulatory framework. (The NYSE letter acknowledges the terminolgy used in the statute, but does not appear to grasp its meaning). Section 11(a)(1)(G) refers to the yielding of priority, parity, and precedence, traditional auction market concepts governing who is entitled to participate in an impending trade, and who must take a back seat, when there are competing bidders or offerors that can trade against incoming contra side interest. The terms "priority, parity, and precedence" have never had any relevance whatsoever to a trade that has already been consummated. Yes, the NYSE is correct that Section 11(a) does not define the term "yield." But the reason is obvious: the context in which the term is used is so crystal clear that any definition would be superfluous.
While the NYSE's position here is clearly ignorant, the broader question, I suppose, is who cares? Why am I spending this much time on a subsidiary point? What does it matter what terminolgy is used? The answer, I submit, is that legally precise terminology consistent with historical connotations is of the essence in any readily comprehensible regulatory framework.
Seasoned securities professionals will understand the term "yield" to refer to not being able to participate in an impending trade. Similarly, seasoned securities profesionals will understand "substitution of principals" to mean giving up participation in a trade that has already occurred. And this term captures the precise legal significance of what the NYSE is proposing far more accurately than "yielding", which is misleading in this context. When a trade is consummated, a contract is entered into. The specialist, upon consummation of the trade, becomes a principal to that contract. What is significant to remember about the NYSE's proposal is that the consummated trade is not "broken" just because of a subsequently arriving superdot order. Rather, the contract stands. (As it should, in fairness to the contra party). Under the NYSE's proposal, the subsequent superdot order would become the principal to the contract, replacing the specialist. Had the trade been "broken", with a new trade then to be effected in the maket on behalf of the superdot order, the term "yield" might be relevant. But since the original contract stands, the only conceivable, legally acceptable usage here is "substitution of principals."
It is clearly in the public interest that precise terminology, consistent with historical usage and traditional understandings, be employed. In the unlikely event that the SEC can overcome the huge conceptual hurdle that the underlying legal issue with respect to the entire proposal is absurd, the SEC must stand firm on the issue of appropriate rule language.
The NYSE's Discussion of Rules 76 and 91 Betrays A Fundamental Misunderstanding of How Those Rules Operate and What They Are Intended to Accomplish
Rather hubristically, the NYSE indicates that I have not understood its market, and that it was responding to my comments simply to provide "useful information" to the Commission. I will review the NYSE's discusion of its market; the SEC staff should then have little difficulty determining who really understands what.
In my earlier correspondence, I noted my dismay that the NYSE had not in any way acknowledged the relevance of NYSE Rules 76 and 91 to the discussion, and the need to amend those rules (however foolish that might be) to dispense with those rules' clear mandate that a subsequently arriving superdot order be exposed to the market (in situations where the rules are applicable), rather than simply be assigned a price by the specialist as contemplated by the NYSE's proposal.
Belatedly, the NYSE acknowledges these rules in its June 7 letter, but discusses them in a bizarre and unprecedented fashion. The NYSE's summary of the rules is overbroad. The NYSE makes it appear as though the rules apply in any instance in which the specialist effects a proprietary trade. The rules, in fact, only apply when the specialist is representing both a buy and sell order at the same time (Rule 76) or is intending to trade as principal with an order he or she is representing as agent (Rule 91). The rules do not apply, for example, when the specialist effects a proprietary trade with a floor broker, likely trading situations which render the NYSE's proposal particularly problematic, as the specialist would not necessarily, as the NYSE's June 7 letter would have it, have followed any order exposure procedure prior to effecting a trade. The NYSE's confusion about how rules 76 and 91 actually operate led directly to their crafting of the ludicrous Scenario 1 on page 3 of the June 7 letter (discussed below).
The NYSE then misses the point as to how Rules 76 and 91 are relevant. The NYSE provides the following paragraph, which I had to read several times before being able to suspend disbelief that the NYSE could have so ignorantly misrepresented its own rules:
"Because this procedure [crossing/order exposure] ensures that the specialist's bid (offer) is the best available price at the time that the dealer trade is orally consummated, any later arriving DOT order(s) to which the specialist must yield under proposed Rule 104.10(11) would, by definition, also be receiving the best available price in the market at the moment that the order arrived on the book. Put another way, because the specialist has just offered (bid) the security from his own account before yielding to a later-arriving DOT sell (buy) order, he has appropiately determined that the later-arriving DOT order would not have received further price improvement if it were exposed to the market. This is consistent with the intent of Exchange Rules 76 and 91."
It is difficult to know where to begin to deconstruct so credulity-defying a paragraph. The Exchange appears to be adopting a hitherto unknown ( and absolutely ridiculous, to say nothing of being totally inconsistent with the rules' plain language) interpretation of Rules 76 and 91 to apply them retroactively to a prior trade in the market.(Crossing procedure followed on first trade, its effect thereby carries over, somehow, to the subsequently arriving order). This retroactive application of the rules thereby obviates the need to follow the rules' clear requirements (these are very simple, straight-forward rules) when an order enters the market after the consummation of the prior trade. And, since these rules can be applied retroactively, the specialist can simply assign the price of the prior trade to the order, having determined, with respect to the prior trade, that the subsequently arriving order will be receiving the best price. And somehow the notion that the order exposure/crossing requirements don't have to be followed is "consistent" with the "intent" of rules whose plain language absolutely mandates that the crosssing procedure be followed in every instance, no exceptions.
This is total rubbish, shameful stuff really. Don't the NYSE's lawyers review this material before it leaves the hallowed halls? But, as we shall see, it gets even worse (more incredible) with Scenario 1 (discussed below).
The fact that the specialist may have followed the crossing procedure (or not, as in a floor broker trade) in a prior trade has no relevance whatsoever to a specialist's responsibility to expose the subsequently arriving superdot order to market interest existing at the time that order is received. The blanket asumption that the price of the specialist's prior trade is the best price for the subsequently arriving superdot order is simply untenable as a general principle. It may be, it may not be, but, in the event, the rules clearly require that the market existing at the time the order is received, not the specialist, determine the appropriate price for the order. The specialist has never been permitted under Rules 76 and 91 to act as the "price determinor" based on prior transactions; far from being the "price determinor", the specialist is simply required under Rules 76 and 91 to act as the facilitator, by means of the crossing procedure, of the price discovery process. The fact that the specialist has never been permitted to simply assign prices, but has been required to follow Rules 76 and 91, has been one of the great strengths of the NYSE market, as it results in prices validated by the marketplace. The sheer ignorant mindlessness of the NYSE's proposal can only undermine that historic strength, its trusted prices.
And yes, a subsequently arriving superdot order could in fact receive a worse price, as it clearly should if that is the market dynamic existing at the time the order is received. That is absolutely the way all markets (including, most especially, the NYSE) have always worked: orders are executed at prices existing when they are received, not assigned the price of a trade consummated before the order even entered the market.
The notion that the price at which a specialist has effected a proprietary trade must of necessity be the best price in the market for a subsequently arriving superdot order (the NYSE's position) is absurd in many real world market situations. A simple example should suffice, although dozens could be provided. (I'm grateful to an NYSE floor broker here, who assures me this is a "garden variety" example). Suppose the market is .50 bid - .60 offer, with the size of the bid being 2000 shares and the size of the offer being 1000 shares. The last sale is .49. The bid consists of the following: a 1000 share bid made by a floor broker, who has priority, and a 1000 share limit order on the public limit order book. The floor broker is working a larger, "not held" market order, but has only shown the 1000 shares reflected in the priority bid. Consistent with typical practice, the floor broker has not indicated to the specialist the total size of his buying interest, nor has he indicated at what subsequent prices he might trade.
Assume the specialist determines to sell 1000 shares from the dealer account to the floor broker's priority bid. Assume the specialist did not first offer 1000 shares at .51 to try to obtain price improvement for himself. The specialist would probably feel this is pointless, as the floor broker has priority at .50 (and, in the event, there is no legal requirement that the speialist make an offer of .51 here before trading at .50). The specialist effects the trade at .50, but before it can be reported the specialist receives a superdot order to sell 1000 shares at the market.
Under the NYSE's proposal, the specialist would be required to substitute the later-arriving superdot order as principal to the contract with the floor broker representing the 1000 share trade at .50. The NYSE would argue that the specialist, having just traded at .50, would of necessity have determined that .50 is the best price for the later-arriving superdot order and would have no responsibility to expose that order at .51, as Rule 76 (the applicable rule in this case) would otherwise require on its face.
Here's the obvious problem with the NYSE's position. There's a floor broker with a not held market order who would, in fact, probably pay up a penny (the broker no longer has priority at .50) if the specialist offered the stock at .51, as clearly he is required to do under Rule 76. But the crucial point here is that the market should decide (in the form of the floor broker), not the specialist, as to what the appropriate price is for the later-arriving superdot order, and that is why the specialist is required in every instance to expose the order to the market. What kind of prices do later-arriving superdot orders receive? Sometimes better, sometimes worse, sometimes the same, which is as it should and must be, because the then-existing market, not the specialist, determines the price.
If the NYSE is serious about its newly concocted "interpretation" of Rules 76 and 91, it obviously needs to amend those rules accordingly, because the rules' language flatly contradicts the "interpretation." Let the SEC then publish that proposal for public comment. I'd wager it won't pass the most lenient "giggle test."
And speaking of "giggle tests"....
Scenario 1: The NYSE's Entrant for Inclusion in the Guinness Book of Records As "Stupidest Statement Ever Made by an American SRO to Its Regulator"
In its June 7 letter, the NYSE offers two "scenarios" purporting to illuminate the operation of Rules 76 and 91. The following factual predicate is common to both scenarios:
"Assume that the specialist has a buy limit order at .50 and a sell limit order at .60, which constitutes the current market. The last sale was at .49."
In scenario 1, the NYSE has the specialist selling from the dealer account to the .50 bid. In scenario 2, the NYSE has the specialist following Rule 76 and exposing a sell order limited to .50 by making an offer of .51 on its behalf before completing a trade at .50.
Both scenarios purport to show the specialist's diligence in establishing .50 as the best market price, thus, in the NYSE's view, obviating the need for the specialist to expose a subsequently arriving superdot sell order (notwithstanding the rules' clear requirement that the order be exposed). Both scenarios are eqally objectionable in this regard, for the reasons stated above.
The real problem here, though, is the manner in which the NYSE has described the specialist's actions in scenario 1 in effecting a proprietary sale to the .50 bid. If not already convinced, this should disabuse the SEC staff once and for all of the notion that the NYSE has any credibility whatsoever in its handling of this proposal.
According to the NYSE, a specialist wanting to effect a proprietary sale to the .50 bid would be required, under NYSE rules, to first make a bid of .51 to try to attract sellers. (The bid is the specialist's proprietary bid. Obviously, the specialist cannot bid .51 on behalf of a buy order limited to .50). Attracting no sellers to trade with the specialist's bid at .51, the specialist then consummates a transaction by selling to the limit order at .50.
This is jaw-dropping stuff to read, flabbergasting in its monumental inanity. Fortunately, the procedure the NYSE claims the specialist should follow is not at all the actual rule, or the NYSE floor would never see another order. What the NYSE is saying here is that if the specialist wants to sell stock, he or she must first, instead, attempt to buy the stock (which, I'm sure, will come as a great shock to the specialist community). Not only must the specialist try to buy stock when all he or she wants to do is sell, but they must do so in a manner that "penny jumps" a public limit order they are representing as agent. If, in the NYSE's example, the specialist had attracted sellers to trade with the specialist's proprietary bid of .51, the specialist would have been required to buy the stock, penny jumping the book and adding to an inventory position when all the specialist really wanted to do was reduce that position.
I'm not in the remedial education business, but let me help the NYSE out here. (I'm grateful for some help from an NYSE specialist on this). The purpose of Rule 91 (the applicable rule here) is to try to obtain a better price for the order against which the specialst seeks to trade, which in scenario 1 is the public limit order at .50. Yes, the specialist should attempt to attract sellers, but sellers who will possibly give a better price to the public limit order, not sellers who would trade against the specialist's penny jumping proprietary bid. What the specialist would actually attempt to do under Rule 91 is "draw" the .50 bid, make a new bid of .49 on behalf of the public limit order (to try to obtain a better price for it) and make a proprietary offer at .50. Absent price improvement for the limit order, the specialist would complete the trade at .50. Basic Auction Market 101, a process that has been a great strength of the NYSE over the years.
The NYSE's treatment of scenario 1 is not just mind-blowingly dumb, it is totally clueless with respect to the purpose of its own fundamental rules and the actual dynamics of its trading process. It is appalling that an institution of the stature of the NYSE would shovel such moronic drivel into a public document and assert that it is providing "useful information" to the Commission. I say this perhaps only half in jest: maybe OCIE needs to audit the NYSE's rule submission process.
The NYSE's Proposal Would Permit the Specialist to Trade at Better Prices Than Those Received by Subsequently Arriving Superdot Orders
In my previous correspondence, I indicated a situation in which the specialist effects a proprietary sale at .50, but before the trade can be reported the specialist receives a superdot market order to sell. I posited a situation in which, just after the specialist became aware of the later arriving sell order, NYSE autoquote published a new bid of .55. I indicated that, under the NYSE's rule as drafted, the specialist would substitute the later arriving sell order into the trade at .50, and then turn around a make a proprietary sale to the new .55 bid, grabbing the improved price that in today's market would go to the sell order as a matter of course.
In its June 7 letter, the NYSE contends that this would not occur, but its discussion in fact confirms that this absurd result is exactly what would occur. According to the NYSE, the critical consideration is the time that later arriving orders enter the display book. Thus, in the NYSE's view, if the later arriving sell order entered the display book before autoquote reflected the new bid of .55, the sell order would be properly substituted into the trade at .50. This is, of course, ridiculous. The critical consideration cannot simply be the time of entry of orders into the display book; the critical consideration must be the actual market dynamic at the time the specialist becomes aware of the orders and is prepared to take action with respect to them. If the specialist becomes aware of the subsequently arriving sell order and the new bid before the specialist has taken any action with respect to that sell order, common senses dictates that the specialist handle both orders consistently with the new market dynamic (i.e., offer the sell order at .56 for price improvement, and, if not forthcoming, complete the trade at .55. This is absolutely how it would work today. The NYSE's position, however, is that since the sell order (though unnoticed and unattended to) entered the display book as little as one second before the new bid at .55, the sell order should receive a price of .50. Clearly, the NYSE needs to go back to the drawing board on this. (The discussion in this paragraph is relevant only if the SEC clears the broad policy hurdle discussed above and gets down to nuts and bolts practicalities).
I made a number of other important technical objections to the NYSE proposal that the NYSE did not address in its June 7 letter. I ask that they simply be incorporated by reference herein.
The NYSE's rule proposal is clearly problematic on both a conceptual and practical level. But the NYSE's defense of the proposal, as reflected in its June 7 letter, is a pathetic embarrassment on every level. The only honourable recourse for the NYSE is simply to withdraw its rule submission. Absent that, it must be rejected out of hand by the SEC.
Very truly yours,