Two points need to be remembered from the last Auction Countdown (5/1/99) in order to follow the thread into this one. Both contradict conventional wisdom. The first is that it is regulatory reform – not technology – that is driving the whirlwind of changes to our market structure. ECNs, day trading and demutualization are creatures of the reforms, not spontaneous responses to new technology. While regulators and their academic supporters like to portray such changes as the result of the inevitable march of technological progress, to which they must respond quickly with academically supportable regulatory adjustments, in fact it is the march of regulation itself to which they are responding. The second point is that, while the reforms are consistent with the antitrust tradition of breaking up network industries, breaking up the stock market network is disrupting its most important functions, such as liquidity and price discovery. Stock markets traditionally organized around membership structures do, indeed, appear to violate antitrust principles. It is, therefore, not surprising that reforms are pressuring them to demutualize. But I see this as evidence that antitrust should be re-examined, not stock market structure.

    This Commentary will focus on how reforms, by breaking up the stock market’s network, are also neutralizing the value of its network effects, such as those which drive trading to larger exchanges and away from smaller ones as investors seek liquidity. This powerful force for centralization, which also provided the basis for profitable intermediation, is being fragmented by reforms. One consequence for intermediaries is that the loss of the potential to exploit network effects is rapidly undermining the business models of almost all exchanges and broker-dealers. Where they used to generate value and brand recognition by offering unique pockets of liquidity to customers, they are now being forced to share their liquidity with everyone – including their competitors. As a result, ECNs can now launch new businesses without any liquidity of their own by being mere "front ends" to the public liquidity. This situation will not lead to the generally predicted consolidation, but rather to more ECNs and more exchanges. Few, if any, however, will find this role profitable, a predicament we will trace to the missing network effects.

    Call markets, which operate at fixed points in time, rather than continuously, could provide some relief from these problems. That is because continuous markets need membership organizations to create networks and network effects, while call markets do not. While reforms pressure membership organizations to demutualize, call markets – because they do not need those now verboten anti-competitive organizing methods – can restore some lost functionality to the markets by, in effect, reconnecting the network. Along the way, they may also help shore up some important business models. Being different in kind from the commoditizing continuous pool, they naturally distinguish themselves from the pack. Those continuous market vendors who specialize also in call market access may do so as well. Unfortunately, this will take time. Meanwhile, since almost all markets are now continuous, the disintegration of membership organizations will remain a big problem affecting the liquidity and stability of the markets, and the viability of Wall Street’s businesses.

    First let’s see how membership structures create network effects, positive feedback, and monopolies. As Auction Countdown readers know, I do not use the word "monopoly" pejoratively, as long as such business forms are the natural result of competition, not Government protection. True natural monopolies are essential, in my view, to the proper functioning of stock markets, among other things. And, as we shall see here, they are also needed to support durable differentiation and profitable brands for markets and members. Whether they form without deliberate cartelization and market division, as they do in call markets, or by use of those traditional (and now illegal) organizing techniques, as they do in continuous markets, monopolies are essential to centralization, liquidity, price discovery – and profitable intermediation.

    Traditional market organization is virtually synonymous with cartelization: it always involves some form of loyalty oath. A strong form of such an oath might say "We members agree to only trade stocks listed on our exchange and only with its members and to refrain from trading with the members of or dealing in the listed securities of any other exchange, or from engaging in any business whatsoever with those who do." While modern descendents of such rules – of which NYSE Rule 390 is the best known example – are never so strongly worded, every traditional exchange still has at least tacit prohibitions and inducements designed to prevent its members and the exchange itself from supporting competitors. Such practices promote "tipping." Any competition among two or more exchanges for members or listings will quickly tip in favor of the first or biggest. The bigger it gets, the more untenable all alternative exchanges become, as their prospective members and listed companies are effectively ostracized from the relevant community.

    Such membership tipping is a positive feedback loop that is normally strongly reinforced by a similar liquidity loop, which drives trading to larger and more liquid exchanges. These two positive feedback loops work powerfully in tandem to winnow the weaker and smaller exchanges until only one is left standing. But exchanges should never assume that establishing a monopoly position means they will keep it forever. In particular, exchanges should not fall into the trap of believing that, having won the day, they can rely henceforward on the liquidity loop alone and abandon those uncomfortably "anti-competitive" features of the politically incorrect loyalty loop. If the lack – or loss through regulatory prohibition – of effective loyalty provisions allows non-members or other exchanges to free-ride off the prices being determined on an exchange, even very large liquidity advantages can turn to sand.

    This is roughly the situation faced today by most of the world’s largest exchanges, including the NYSE and Nasdaq. Not only have their property rights in trading information been undermined by mandated public tape and quotation systems, but the latest order handling rules mandate direct access to trading opportunities, too. Thus, retail customers of e-brokers and ECNs in Nasdaq stocks have direct terminal access to trading on an equal basis with the professional members of that market. Because the rules require that all ECNs be allowed to free-ride on each other’s and the general market’s liquidity through Selectnet, the concept of proprietary liquidity on a market whose members are loyal to it is already out the window in the Nasdaq market. Once ECNs become exchanges, as several plan to do, this situation will also apply in NYSE-listed stocks.

While the press often portrays these ECN applications to become "full-fledged" exchanges as if they were moving up to a higher level of regulation, in fact exchange regulation is defining itself down to them. There is no practical difference between ECNs linked through Selectnet for the trading of Nasdaq stocks, and those same entities linked through ITS (Intermarket Trading System) for the trading of Listed stocks. The fact that in order to do the latter they need to register as exchanges under new Rule ATS does nothing to change their basic structure or business model. And Rule ATS will also allow old markets, such as NYSE and Nasdaq, to become demutualized, for-profit "exchanges," effectively to compete as ECNs, too.

    In the end, merging definitions under the new regulatory reforms will force competitors to converge on a business structure similar to what ECNs trading Nasdaq stocks look like now, and all competing entities – the NYSE, Nasdaq, all current and new exchanges, ECNs, and brokers – will adopt that form. Whether through merger, alliance, association, multiple registrations, or other means, all continuing competitors will have to be all things to all people. In order to get tape revenues and full access to trading Listed stocks – including an ITS link and a pass on NYSE Rule 390 – all ECNs will become (or associate with) an "exchange." In order to compete at electronic access with ECNs in Nasdaq stocks, all exchanges will demutualize and become or sponsor ECNs. All these ECNs/exchanges (probably under some new name or acronym, since the categories will have formally merged) will free-ride off each others’ and the general market’s liquidity via Selectnet, ITS or successor systems. And they will all provide customers with direct electronic access to all information and trading facilities.

    One doesn’t need to follow this line of thinking very far to see that even the most fundamental and familiar distinctions in the market could soon disappear. With the NYSE planning to launch an ECN to trade Nasdaq stocks and Nasdaq planning to register as an exchange to trade Listed stocks, and both of them planning to demutualize their members out of any say in whether their systems become fully electronic, before long there may be no difference between auction markets and dealer markets, NYSE and Nasdaq. They will, for all intents and purposes, be the same market using the same system to trade the same merged list.

    It is a matter of no small irony that the regulatory juggernaut leveling categories and distinctions all around has been fueled for several decades of National Market System reform by an effort to legally define the difference between exchanges and brokers. This effort was not necessary in the first place and has utterly failed. Fraught with perilous competitive issues and beset by vested interests on all sides, regulators went through the Institutional Investor Study in the ‘Sixties, NMS hearings in the ‘Seventies, the Delta case in the ‘Eighties, and Market 2000, Order Handling, and ATS in the ‘Nineties – only to arrive at ever more incomprehensible "exchange" definitions that moved progressively away from accomplishing their distinguishing task. The first irony is that there is no longer any practical difference between exchanges and brokers, as tacitly recognized by Rule ATS, which gives them a choice of how to register. The double irony is that the obvious answer all along has been that a stock exchange is a membership organization for the trading of stocks – a definition never considered by modern regulators.

    The triple irony is that this is not too far from the original definition in the Securities Exchange Act of 1934, whose authors apparently considered what an exchange was to be so obvious that they dispensed with the matter in circular fashion, essentially saying an exchange is an exchange "as that term is commonly understood." While the rest of the ’34 Act definition could apply to both exchanges and brokers, does anyone really believe that the Congress of 1934 was unaware of the difference? To imagine that all the subsequent re-defining was necessary, one has to imagine that Congress was both ignorant of the obvious differences between exchanges and brokers and willing to let the SEC defy the law by requiring exchange registration of membership exchanges only and never of brokers. Since neither of these assumptions is conceivably justified, much less both of them, one is left to conclude that the entire NMS exercise that has led to the currently collapsing categories was, at best, unnecessary – an argument over how many angels can dance on the head of a pin.

    As unnecessary as NMS has been, its result – that exchanges and brokers must compete with each other as if there were no difference between them – is consistent with antitrust principles. Market division is an antitrust evil, along with cartelization, and the first purpose of traditional market organization was to divide the role of the exchange itself from that of its members. This first act of market division then provided the backdrop for the further differentiation of the businesses of the members, a sort of tacit market division enabling each member to develop a unique and, therefore, profitable niche (more on this below). Members were profit-seeking businesses who chose to organize their exchange so it could maintain a favored environment for their pursuit of profit. Exchanges were non-profits because their only purpose was to maintain that profitable environment for their members and never, under any circumstances, to compete with them. How things have changed.

    Many observers of the demutualization trend have concluded naively that it is being driven by technology. It is not. While the presumed need to rapidly insert technology into the market provided the nominal justification for NMS – as if modernization would not happen without Government intervention to force it – the real reason that exchanges are demutualizing is that NMS trust-busting has made the exchange role, with all its market division and cartelization, untenable. In other words, it is reforms – not technology – which are forcing exchanges to demutualize. Technology is only delivering the coup de grace.

    The regulatory consequences of having never considered the obvious – that exchanges are membership organizations whose first purpose is to divide the roles of markets and members – are many. Having ejected the membership structure from its official definition even as the SEC’s entire registration and regulation methods were based on it, the Commission is now facing several dilemmas. How do you accomplish the registration of electronic brokers (ECNs) as exchanges, when their structure as brokers is fundamentally inconsistent with exchange structure? The answer appears to be – a la ATS – that you alter the regulation regime for exchanges to allow non-membership, for-profit brokers to fit into it. If you do that, however, fairness dictates that you let the old membership exchanges out of their regulation regimes by letting them become for-profit, non-membership entities, too.

    Although such dilemmas will continue to perplex regulators, their practical import for competitors is that reforms will push markets rapidly toward a situation in which stock market intermediation becomes a commodity and essentially disappears. Such "democratization" is understandably popular with the e-crowd, especially day traders. But it is bad news for the exchanges and their members, because it erodes the exclusivity value of membership, with all its cartelization and market division benefits. Although seat prices may stay aloft as long as members expect the cash out value of their seats in an IPO or merger to reflect the strength of their exchanges’ powerful brand names and critical mass of trading, conversion to for-profit status will in reality be accompanied by loss of the foundation on which the brands were built and consequent rapid dispersal of their critical mass advantage. With reforms effectively mandating direct public access to all its facilities, for example, it is difficult to imagine what the residual value of "permanent" access for members in the NYSE’s reported IPO plan is. Not only will the public have equivalent access without needing to be members, but – without brand-building and brand-protecting loyalty provisions – the world’s largest exchange will become just another ECN competing with those created by its erstwhile loyal members.

    Worse, the exchange will have to allow its competitors to free-ride – read "feed" – off its critical mass via NMS linkages like ITS. Taken to extremes – where we are clearly headed – all markets and former members will be relegated to roles of being mere windows on the "public" market, front-ends to a system designed and effectively run by the SEC and its academic advisors. With the loyalty loop dead by law and the liquidity loop transformed into a public utility, such once differentiable competitors as NYSE and Nasdaq, Instinet and Merrill Lynch – even newcomers like E-Trade and Archipelago – could all become mere six-of-one-half-a-dozen-of-another access points, differentiable by little more than their screens’ color schemes.

    Some economists have noted that what antitrust theory calls "perfect competition" is, in the end, a profitless condition. That condition is upon the stock market now. As hard as it is to imagine that such a traditionally profitable industry as securities trading could suddenly become profitless, the threat is real. To understand why, we need to look first at how things used to work. Until recently, each member was able to develop a distinct book of customers and a distinct set of procedures backed by a distinct mix of agency order matching and capital commitment, such that the combination for each firm was a distinct liquidity pool, different from its competitors. Salomon’s brand of liquidity was different from Goldman’s, which was different from Merrill’s, Schwab’s or Instinet’s. And upstairs discussions about order flow – the kind that were banned by the antitrust settlements – enabled the discovery of good prices given those flows, thereby facilitating order matching and capital commitment.

    How? In the pre-electronic, pre-anonymous days, the unique personal relationships of each firm’s individual traders with customers and other dealers put a high reputational premium on honesty and integrity. Buy-side traders were reluctant to mislead dealers about their orders’ true sizes, because a burned dealer would not commit capital for them again. Moreover, because traders do talk among themselves, the dissembling buy-side trader could get such a reputation that no dealers would deal with him. Similarly, sell-side traders were careful to avoid harming a valued customer through whatever upstairs discussions resulted in arranging a large trade. Failure to do so – in the non-anonymous environment – would result in the dealer losing that customer. And, again, because traders do talk among themselves, a front-running dealer would probably lose many more customers than the one whose trust he abused.

    Such detailed and unique counter-party auditing in the pre-anonymous days enabled the liquidity and price discovery network to function very effectively. Considerations ranging from the P&L to individual career prospects to a firm’s standing in the industry all hinged upon – and enforced – honest communication of accurate information about order flows. Importantly, the depth and strength of this reputation-based information network increased with size. In another example of increasing returns in a positive feedback loop, the larger institutions sought out the largest and most honest dealers, and the larger dealers sought out the largest and most honest institutions so they could service their liquidity needs – and see their flows. For institutions, the bigger they got, the more their size could command the best liquidity servicing by the largest dealers. For dealers, the bigger they got from servicing the larger institutions, the easier it was to credibly sell their liquidity services to them – and, therefore, the bigger they got.

This positive feedback loop enhancing size and concentration is important, because the larger the institution, the larger the orders, and the larger the orders, the more price discovery import they would have. Thus, "size talking to size" led to an information pyramid or hierarchy in which the largest and most important flows were handled by the largest dealers, who would work as needed with peers or down the chain to smaller dealers to complete a trade. With so much capital and reputation on the line, these ad hoc syndicates had to discover accurate prices – and they had the order flow information to do so.

    In yet another positive feedback loop, the more size-talking-to-size enabled accurate prices and liquidity based on them, the more viable the businesses of ever larger institutions became, which, in turn, enabled the further aggregation of meaningful order flow and better price discovery. The one word that best captures this size aggregation process is "institutionalization." But, while many have debated the pros and cons of aggregating individual interests in the hands of large institutions, the debate has turned almost entirely on the populist question of whether bigness is bad per se. Neither side in the debate – from the time the SEC’s Institutional Investor Study kicked it off – seems to have noted the potentially critical connection between institutionalization and price discovery, much less the fact that, if such a connection exists, de-institutionalizing the market could be highly destabilizing.

    In any case, regulatory reform has suddenly turned the institutionalization process upside down. By banning as anti-competitive such personal relationship-based practices as holding non-disclosed orders in the crowd at the specialist’s post and engaging in upstairs dealer discussions about customer orders, reforms have destroyed the reputational infrastructure which constituted the price discovery network. And by fostering electronic trading as a replacement, reforms have inadvertently encouraged participants to drop their old honest ways and anonymously engage in a variety of at least misleading and sometimes manipulative order placement practices. As a result, price discovery has turned into a crap shoot, making dealing in size at any one of those fleeting prices too dangerous for both institutions and dealers, and forcing even the largest of them to contemplate emulating the day traders they complain about, even though their average trade size is less than 1/1000^th the size of the average institutional order.

    The academic pipe dream underlying NMS is that connecting all traders to an electronic system in which everyone has exactly the same information is itself a full liquidity and price discovery network. The problem is that neither viewing the continuous tape of small trades nor the quotes of traders’ willingness to trade in such small sizes conveys any of the meaningful information that characterized the old network. Moreover, expecting thousands of day trading truck drivers, hair dressers and dentists to make up for the liquidity lost when NMS prevents professionals from doing their job is an academic conceit of gargantuan proportions. The playing field may be level now, but the old players are gone and the new ones can’t do the job.

    While it is not inconceivable for the old network to form again in spite of the presence of the black box, that is unlikely. The volatility engendered by the black box renders disclosure of real information about size far more dangerous than before. Moreover, with the regulatory environment so strongly biased against private information, all the legal departments of the large firms are simply advising their traders to stop complaining, stop talking – and just use the black box. The reality is that the old methods were a complex and detailed ecosystem. Expecting it to regrow after NMS’s black box mandates would be like expecting the arctic tundra to redevelop if all of it were covered by a massive oil spill.

    The transformation of the market from one in which a relative handful of dealers at the top of an information hierarchy did the heavy lifting in price discovery and liquidity provision to one in which that role has been dispersed among thousands of individual on-line traders is perhaps the world’s best example to date of what antitrust critics call "atomization." Competition is now so "perfect" that there is no longer any chance of collusion – tacit or otherwise. There is also no differentiation and no pricing power. Reforms have turned the only naturally differentiable aspect of the Wall Street value proposition – the organization and packaging of liquidity – into the equivalent of a public utility run by regulators. Instead of providing liquidity itself – a differentiable and brandable activity – competitors today can only provide access to the common liquidity pool. Both access and the common pool are inherently undifferentiable commodities.

    In an increasingly desperate bid to retain some relevance in this Brave New World, virtually all brokers now offer or plan to offer the on-line service that is turning customers into competitors. While this would seem to validate the position of the e-brokers riding the new paradigm, there will be few opportunities to durably differentiate them either and, thus, precious little pricing power for anyone. Going forward, the e-competitors will not only include all new and old brokers, but ECNs and exchanges, too, all vying to provide their brand of access to the same public liquidity.

    The old broker brands naturally differentiated themselves by virtue of their trading reputations, which were almost impossible to precisely copy once a firm began riding the increasing returns to size in its personalized niche. Moreover, those trading reputation-based brands did not need much in the way of expensive advertising, because word of mouth was so efficient at reaching the key decision-makers at the firms you wanted to do business with. Most important, your brand was not just an image trumped up in an ad, it was a unique reality that, in the truest sense of the term, advertised itself.

    In contrast, few if any of the new elements of broker brands offer either increasing returns to size or unique niche-building potential of any kind. Moreover, because the audience has changed from a few key decision-makers easily reachable by word of mouth – and each one personally loyal to the person who services him – to millions of individuals your traders will never know, any successful brand can and will be copied. The best broker that is also a financial portal? Copyable. The best discount broker, period? Copyable. The best broker that projects an image of fast day trading? Or the best that encourages a responsible approach? Both easily copyable. Research? Entertainment? Real time account updates? Level II? Psychological counseling for compulsive traders? While all of these seem potentially brandable, efforts to build brands around them will be characterized by old-fashioned diminishing returns, not the increasing returns that naturally protect brands. So every success can and will be copied and commoditized. The situation will require ever escalating advertising costs to keep the brand fresh, ever-escalating development costs to keep the product up-to-date, and continuous price discounting to stay in the game.

    That almost all competitors in the stock market business are likely to face grave difficulties soon is not generally recognized. This lack of recognition stems largely from the view also held by regulators and academics that change is just the result of the inevitable march of technology, rather than radical regulatory reform. This view allows its holders to assume that, after a few difficult years of transition, new structures and competitors will establish themselves – better and stronger than the old ones – and we will all happily say good riddance to those antiquated, anti-competitive membership exchanges. Typically, those who hold this view seem to focus only on the liquidity-begets-liquidity phenomenon, apparently having forgotten the traditionally tight relationship between success generated by an exchange’s liquidity and that generated by its loyalty oath features. The oversight has led to confident predictions from leading figures at exchanges and ECNs – as well as from prominent regulators and academics – that will almost certainly prove wrong.

    Their predictions appear to be based on the presumption that, once the anti-competitive loyalty oaths are dissolved, the network effects will again reassert themselves, allowing positive feedback in the liquidity loop to re-centralize trading and bestow incumbent monopoly status on a new winner. Thus, the heads of several ECNs predict that all the new ECNs and exchanges will consolidate soon into only one or two trading venues (presumably including the speaker’s own). What they are missing is that, without loyalty provisions, any liquidity advantage (in a continuous market, anyway) can and will be dissipated by free-riding. In that the reforms as much as mandate free-riding forever, none of the brokers, ECNs or exchanges that expect to gain, ride or keep a liquidity advantage will be able to do so.

    A scenario more consistent with current regulatory reality would see continued fragmentation, indeed, accelerating proliferation of trading systems, as more and more firms realize that the reforms encourage free-riding on the liquidity and prices of the incumbents who are now powerless to prevent it through loyalty oaths. Going forward, the National Market System reforms will play out in either weak form or strong form. Weak form NMS is essentially what we have now: linked trading venues in which the linkage is imperfect enough that different liquidity pools can maintain some semblance of uniqueness and independent liquidity, but not so imperfect as to enable one or two of them to dominate. Strong form NMS is a "hard" CLOB (consolidated limit order book), to which every system is connected so perfectly by CLOB-wide price and time order handling priorities that it makes no difference whatsoever what system you come in through: the liquidity is identical.

    If liquidity anywhere becomes liquidity everywhere in the National Market System, then it no longer matters whether you are big or small; you can get into the game even if you have no liquidity to offer, and you can’t be knocked out by competitors who have more. This environment will breed more, not fewer, "exchanges," and more, not fewer, ECNs. Although the names of the categories could certainly change again, the number of entities providing official access will probably continue to include not only all of the current entrants, but perhaps many others in related businesses, too.

    The H&R Block purchase of Olde and the CNBC purchase of a piece of Archipelago made clear that the definition of "related businesses" is becoming increasingly attenuated. In addition to all the brokers, ECNs and exchanges, who will be better able to compete in their category if they also own, associate with or register as the other category, some in even less related businesses will be sucked in by that calculus. More TV stations and tax preparers may enter the "exchange" game. Eventually, newspapers and magazines, entertainment companies and Internet portals, e-commerce behemoths like Microsoft, Amazon and eBay – perhaps even the Wall Street Deli – will discover they can better attract customers to their main product by also being an exchange, an official window on the National Market System. When Microsoft gives away free browsers or ISP access to turn you into a more loyal Windows customer, such actions put pressure on the prices Netscape or AOL can charge. Similarly, if you are in the business of trying to sell stock market access against many competitors selling exactly the same thing – some of whom view your core business as a loss leader – you are not likely to have much pricing power.

    One implication of all this is that the announcement-a-day consolidation craze is chasing the wrong rabbit. Big old firms, and giant new e-brokers, are all frantically trying to buy into the right system or coalition, so as to be part of the new central market when it finally forms. Many of them are buying into multiple ECNs and multiple coalitions, just to make sure they have the winning ticket in their pocket. Since, by regulatory design, none of them will be able to consolidate liquidity, none of these coalitions or systems – nor all of them together – will "win." The fancy prices being paid for these portfolios of lottery tickets could perhaps be reduced, therefore, by starting negotiations with the Wall Street Deli. Mandatory public access means that almost any old firm will do the trick to be part of the new NMS. In fact, by the time all the categories finish crumbling, many of the firms ponying up for multiple tickets now may find they had a perfectly serviceable one in their pocket all along.

    Since fixed time call markets are in many ways the opposite of continuous markets, it may not be surprising that call market providers are not generally subject to the problems now facing purveyors of continuous trading. There are two basic reasons for this. First, calls do not need membership organizations or loyalty oaths to prevent free-riding and are not, therefore, targets of the current antitrust-based reforms. Second, because calls centralize trading in the temporal dimension, their liquidity pools are naturally distinct from any others, whether they be other calls or continuous trading venues. This means that – while calls cannot free-ride their way into existence the way ECNs can – once established, others cannot free-ride on them either. Both the uniqueness and the independence of call markets enable their providers to naturally differentiate what they do from the crowd providing access to the homogenized continuous pool.

    Continuous trading naturally encourages free-riding, because it is basically a free-riding process itself. Without temporal aggregation of interest, every trade keys off old trades – the last print, the close, the VWAP etc. Since theoretical equilibrium is an unknown, both parties essentially accept – i.e., free-ride on – the reference price, give or take a spread and/or a commission. This natural free-riding on continuous prices forms the basis for such practices as payment for order flow or other forms of dealer guaranteed liquidity, block trades, crossing networks and principal program trades. It also is the basis for all the linked markets in the NMS, particularly the Selectnet and ITS systems.

    This natural tendency toward free-riding in continuous trading was countered by membership exchanges in two ways. The first line of defense was their loyalty oath features, which originally simply prohibited non-members or other exchanges from seeing their prices. Reinforcing the loyalty oath was the fact that, by promoting the development of a strong reputation-based order flow information network, membership exchanges made customers want to come to them rather than help competitors free-ride off them. With size talking to size, the network was able to discern meaningful order flow well enough to discover good prices, which emboldened members to provide sizable liquidity at those prices. While membership exchanges, loyalty oaths and upstairs information networks worked very well at providing good price discovery, good liquidity and protection against free-riding, reforms are extinguishing all of them. In the absence of the membership organization, I know of no other way to prevent free-riding in continuous markets.

    Even a CLOB could not do it, because, as mentioned, continuous trading is itself a free-riding process. Forcing all trading through one continuous CLOB would only accelerate the process of spreading trading out temporally, thereby putting an even greater premium on getting the prices that everyone else is getting – say VWAP. By increasing the number of trades and decreasing their relative size, a CLOB would push you to shrink and spread out your own trades so that they do not stand out against all the others. And, because you are shrinking your size to free-ride off them, others will shrink theirs to free-ride off you.

    Call markets do not have the free-riding problem. Because they centralize all trading interest into a point in time, they discover true equilibrium very directly and accurately. Moreover, anyone who wants in on that price can do so only by participating in the call. There is no incentive to free-ride, and those who try will risk missing the consensus price they could have easily had. For example, since no one knows precisely what the price is going to be until the call ends at its appointed time, attempts to free-ride on it prior to its conclusion would risk missing the real price. Similarly, attempts to free-ride off its price by waiting until it is over would risk not finding counter-parties willing to trade at it anymore.

    The fact that a call naturally defines its own liquidity opportunity, and, thus, naturally defends itself against free-riding, is not just the source of its price discovery power. It is also a natural protection against commoditization. Two calls at the same time cannot exist for long as businesses. Just as membership exchange competition would tip in favor of one exchange – the one with the most members and liquidity – the competition between two or more markets running competing calls at the same time would quickly tip in favor of the biggest. The winning call would not only be able to easily differentiate itself in the crowd. At least some of those continuous markets and brokers described earlier, who are now so hard pressed to differentiate themselves, could, by offering that call periodically to customers, stand out, too.

    Some might argue that the theories and terms bandied about here – path dependence, positive feedback, increasing returns, tipping, lock-in – are awfully esoteric to use to criticize such settled doctrines as antitrust, the National Market System, electronic trading and transparency. And how could the loss of network effects they never heard of so severely threaten the livelihoods of experienced stock market hands? Wouldn’t the academics and regulators have warned them if such a threat existed?

    The answer is that it is the academics and regulators – in a sort of mutual admiration society – who have spun theories that don’t hold water. Settled doctrine or not, it makes no sense to turn the stock market into a testing laboratory for anyone’s theories. But the fact is that, after all the tinkering, the academics and regulators have become the chief designers of the stock market today. This fact alone should make us question their theories. But, blind to their errors – and perhaps enjoying the power and recognition – academics and regulators keep egging each other on to the next level of intervention and experimentation, as if each policy disaster (fragmentation, day trading, demutualization) justifies more meddling to fix it. The latest example is the increasingly shrill cries for a CLOB as the only way to fix the fragmentation that somehow sprang up in the wake of the order handling rules. Long since banished from the debate is whether their theories made any sense in the first place. This, of course, precludes consideration of whether the actions they have already taken based on those theories may be the cause of the problems they are now trying to fix.

    The public is not so gullible. They understand network effects and related concepts instinctively – although perhaps in less esoteric terms – because they are based on plain old common sense. Voters in the real democracy, for example, would not fall for any phony "democratization" schemes claiming to use "modern telecommunications technology" to put them "on an equal footing" with the "bosses in the smoke-filled rooms." Voters understand why we prohibit reporting election results prior to the time polls have closed. They recognize that West Coast voters are disadvantaged if East Coast returns – or even news stations’ reports of their exit polls – are made available before they finish voting. They can see that early reports of winners would build their credibility and discourage opposing voters (displaying increasing returns). They see that early success is dependent on who votes first (displaying path dependence), and that continuous transparency would build momentum (positive feedback), perhaps delivering (tipping) elections to early favorites, even if a simultaneous or non-transparent vote would produce a different consensus and winner.

    Transparency? How absurd. That would require a running tally so that each voter could see the "tape" of how voting was going prior to casting his ballot. Many voters are even aware of problems in primary voting due precisely to what amounts to continuous transparency. With each state voting separately in a set sequence – so there is plenty of time to transparently report the results prior to the next primary – there is considerable path dependence, increasing returns, positive feedback and tipping. Former President George Bush called it "The Big Mo" (for "momentum"), by which early success in Iowa or New Hampshire becomes a self-fulfilling prophecy (exhibits positive feedback) convincing later primary voters to consider a candidate viable and, therefore, worthy of their consideration. His son, it is said, became the early frontrunner for the Republican 2000 nomination by topping early polls when many of those polled accidentally confused him with his father. Riding increasing returns since then – and long after those careless (devious?) pollsters had corrected their error – Governor Bush appears to have sewn up (locked in) the nomination.

    The primary process itself is one of the world’s best examples of lock-in. No candidate or party can risk challenging it or the sequence of primaries for fear of alienating the very states most needed to gain that initial spark. But these are not esoteric or difficult concepts. Ordinary Americans can easily understand how this locked in process of continuously visible voting can lead to distortions relative to the true will of the electorate. How else to explain the disproportionate number of presidential candidates who express support for ethanol subsidies (to get through Iowa) and sign a pledge to not raise taxes (to get through New Hampshire)? Regardless of the merits of these policies, the fact that these states come first gives their voters disproportionate influence on candidates’ views. These states also seem to have more than their share of voter manipulation incidents, such as the use of "push polling," in which candidates ask leading questions in "polls" of far more voters than necessary to sample their views, with the intent of smearing their opponents. And early primary states are also home to various questionable tactics to create the appearance of more support than candidates actually have, such as paying "supporters" to attend rallies or busing them in from nearby states.

    The point here is that the theoretical factors that might improve or harm electoral process are not so esoteric that voters cannot understand them. Investors, too, can understand such matters as the potential importance of privacy in the interactions between large institutions and dealers. Investors, too, can understand that there might be some natural order or value in a system of membership exchanges where the individual does not trade "on a level playing field" with Salomon Brothers. And investors would certainly understand that, just as Hertz probably gets cars at better prices than individual buyers do, mutual funds might get quantity discounts, too.

    Network effect theory is not rocket science. It is common sense. A brief look at either would convince any reasonable person that the settled NMS doctrine of continuous electronic transparency is bound to produce non-consensus prices, distorted valuations, volatility and manipulation in markets. Not that there is anything wrong with any piece of that doctrine taken separately. Continuous trading, electronic trading, and transparent trading all have very powerful and valuable uses. Electronic and transparent even combine very nicely in call markets. Putting all three together, however, is a formula for the disaster that is unfolding now as a disintegrating stock market structure.

    But rare is the academic or regulator who does not feel that intervention to improve "fairness" is always good. And what could be more popular, populist and "fair" than combining those three key playing field levelers to benefit the small investor? The resulting screams of intermediaries only seemed to confirm that these populist academics and regulators were, indeed, doing good. And there is no question that the SEC’s populist attack on the status quo struck a chord with many – Utopian schemes always do. But it is also clear that even average investors can see that their ultimate interest does not lie in egalitarian redistributions of the assets of intermediaries, but in a stock market that works properly. The populism behind regulators’ and academics’ radical reforms not only enables them to ignore all these obvious points, but also to avoid acknowledging any potentially valid theories that might call into question the wisdom of their massive interventions. This populist ethos, for example, appears to make regulators and academics studiously unaware that it is the reforms that are the principal cause of the day trading phenomenon.

    An article in a recent Barron’s defended day trading as nothing more than the modern electronic version of what professional traders have always done. That’s true, to a degree, but beside the point. You can’t do what professionals did electronically, at least not the size talking to size part. In effect, the SEC gave day traders a license to steal from Nasdaq’s dealers by free-riding off their prices. Happy to play Robin Hood helping the SOES bandits break through Nasdaq’s loyalty oath, the SEC launched the day trading profession, never giving a thought to the possibility that enabling truck drivers to play Wall Street professionals might have consequences beyond simply giving both the opportunity to change careers. The irony is that now regulators are treating day trading like some plague of locusts that blew in by chance on their watch. These are not investors, or even speculators – they are gamblers, say regulators in their highest moral dudgeon. They need counseling, and the firms that house them are violating suitability, capital adequacy and margin lending laws. There is no admission that day trading is a phenomenon 100% created by the SEC’s desire to "level the playing field." The real tragedy of this Frankenstein from the NMS Labs is not that it exists, but that the rest of the market’s most important functions and businesses had to be sacrificed to give it life. Like price discovery.

    To promote the democratization that is giving day traders their opening, all the underpinnings of good price discovery in continuous markets are being pushed aside. Membership organizations are demutualizing, dealers no longer talk to each other about order flow, and trade size is breaking down into a thin stream of meaningless prices all free-riding off each other, none of which can be trusted to support a sizable trade. No wonder tech and Internet stocks are so high that Microsoft president Steve Ballmer calls the whole industry overpriced. No wonder some subsidiaries of listed stocks are trading for free or less, according to a recent Wall Street Journal article. And no wonder it’s so easy to promote "pump and dump" schemes on the Internet that regulatory roundups are reported seemingly every other day, reflecting a virtual explosion of fraud. The core problem giving simultaneous rise to "irrational exuberance," "irrational indifference," and a rich playground for crooks is the loss of reliable price discovery. Yet so sure are regulators and their academic supporters that their egalitarian theories are correct that they have, again, never considered the possibility that the policies based on them are causing the problems they are now trying to solve with more of the same.

    Democracies, it is said, are stable because their governments reflect the consensus of the governed – no need for revolution. But it should be clear by now that electoral process can promote or inhibit the formation of that consensus. Amending the original thought to reflect this, one would say: Democracies are stable in the degree to which their governments reflect the consensus of the governed. Similarly, stock markets are stable in the degree to which their prices reflect the consensus of investors. At accurately determined prices, great volumes of transactions can be confidently done, allowing assets to change hands and capital to flow to productive enterprise. But stock markets, like elections, are dependent on the effectiveness of the process used to discover consensus. If prices are very inaccurate, crashes can happen – just as revolutions to change non-consensus governments can happen. If the price discovery process is built on a poor theoretical foundation, the market is unstable. Worse, if very large markets had developed naturally on top of a relatively correct theoretical foundation – but suddenly had that foundation pulled out from under them – the result could be disastrous instability.

    In that several decades of flawed theory are now pulling the rug out from under continuous markets, it is time to consider how fixed time call markets can solve the problems investors and intermediaries face. Call markets produce a consensus price; so they have a stabilizing effect on the market. They can stop a crash in its tracks, being natural circuit breakers that don’t even need to shut the market down in order to work their wonders. They can alleviate the danger that mutual fund liquidations will cause the market to "melt down" or "seize up" with illiquidity. They can also do more mundane tasks, like opening or closing markets without the accustomed volatility, manipulation and sense of unfairness. They can allow stocks to be added to or deleted from the S&P500 without disruption. Because call markets allow many people to trade at once – and at one price – they have deep liquidity at very low cost for institutions and a feeling of true "empowerment" for individuals. Neither the largest institutional orders nor the wildest packs of day traders whipping up momentum can upset them.

    In terms of how calls may provide business opportunities for intermediaries, because calls’ liquidity pools are naturally distinct, intermediaries associated with their introduction and operation will differentiate themselves from the competition. And, because call markets cannot be free-ridden on, they need no loyalty oaths, market division or other anti-competitive agreements to become effective. This means that calls will not present the antitrust targets that continuous markets did. Even antitrust hawks acknowledge that it is not illegal to be a monopoly; it is only illegal to engage in anti-competitive practices to become, defend or extend one. Because call markets can rely on liquidity alone to create, defend and extend their position, they will have no need to engage in questionable practices. Those intermediaries who build the call market will, therefore, build an asset for themselves that is durable even in today’s commoditizing technological and treacherous regulatory environment. Most important, call markets will spring up naturally to solve the market’s price discovery and other problems. There is no need for studies, regulation, legislation or other interventions to implement them.