Mayday, Mayday

(Originally Issued May 1, 1999)


The apparent consolidation of U.S. stock markets masks a reality better characterized as disintegration, an antitrust-induced breakup of this nation’s liquidity, price-discovery and capital formation network. Although the impending loss of these critical functions – of which escalating volatility and valuation distortions are early signs – is unintended, the breakup of the network is not. Breaking up networks is what antitrust does. From Standard Oil through AT&T to Wintel, sacrificing the coordinating effects of networks to promote a misguided vision of "competition" is meat-n-potatoes for trustbusters, and these ongoing stock market reforms are vintage antitrust. Since neither energy crises, nor telecom chaos, nor crumbling technology standards have gotten trustbusters to recognize antitrust for what it is – The Mother of all Mistakes – there is little reason to suppose they will wake up to their error on their own before disaster strikes the stock market. The only solution to this dire situation, therefore, is complete repeal of all antitrust laws and abolition of the bureaucracies that administer them. The rapidly spreading dysfunctionality of the stock market is both the best evidence of the urgent need for repeal and the choke point at which capitalism will fail in its absence. How could capitalism conceivably survive without capital?


    While recent stock market reforms are achieving many of their stated goals – narrower spreads, lower trading costs for individuals, transparency, automation, opportunities to trade without intermediaries etc. – the actual result beneath the surface is closer to the disaster pointed to by the displaced intermediaries. Virtually every newspaper or magazine these days carries an article or two about certain strange aspects of the current investment landscape: wild IPO volatility, locked and crossed markets, weird valuations seemingly divorced from any conceivable reality, drastic price swings over very short periods that are unconnected to any news or information, clearly valuable companies that go nowhere, while companies that have no business yet skyrocket, and, perhaps most telling of all, old stocks whose names or ticker symbols are mistaken for some new company that has no business yet – also skyrocket. Many of the articles contain an increasingly popular reference to "unintended consequences," giving voice to suspicions that somehow all those good effects of reform may have come at a price.

    I believe we are, in fact, witnessing the disintegration of the U.S. stock market as a liquidity, price discovery and capital formation network. That the breakup of the most important structure in the world of capitalism is proceeding pursuant to textbook antitrust enforcement principles – clear violations of the most egregious "per se" kind, addressed with standard multiparty competition remedies – makes the situation all the more dire. If we do not reverse course soon, recent troubling trends – rampant volatility, diminishing liquidity, grossly distorted valuations, the proliferation of earnings manipulation scams, accounting fraud, pyramid schemes, short squeezes and shell games – will intensify. Within a few years, the market will serve only the needs of the gamblers and scamsters, the casino atmosphere having chased off all the legitimate investors and real companies in need of capital.

    This Commentary will turn later to a discussion of how reforms are causing the breakup of the stock market as a liquidity network, and how this promotes all the unsavory characteristics described above. But let us turn first to the antitrust roots of these reforms, because until it is recognized that antitrust plays no legitimate role in a capitalist system, it will be impossible to challenge the reforms, however detrimental they may be, since they proceed from correctly applied antitrust principles. Markets are not coming apart because we’ve got the wrong increment, or an imperfect balance between auction and dealer market structures, as if a trip back to the comment period drawing board could get it right next time. They are coming apart because antitrust always results in massive Government intrusion into the design and operation of network industries, which always destroys their ability to perform efficiently. And, while we may survive energy crises, telecom chaos, and crumbling technology standards – results caused by antitrust action in some industries discussed below – we could not last long as a leading nation without efficient capital formation. In short, antitrust and stock markets do not mix; one of them must go.


The Two Rockefellers


    In his new biography of John D. Rockefeller, Titan, Ron Chernow repeatedly notes the conventional bewilderment – which Chernow shares – over how Rockefeller could have been both model citizen in his private life and evil incarnate in his business dealings. How could this devoutly religious man, who gave money to the poor even before he had any, going on to become the greatest philanthropist in American history as he gave away the great majority of his fortune before he died, who never smoked, drank or cheated on his wife – how could such a man have been so unscrupulous in business that a whole body of antitrust laws was created to contain his predatory practices, and to permanently outlaw the trust structure he largely invented? The contrast between the two Rockefellers was all the more perplexing to Chernow because Rockefeller, himself, saw no contrast. That is, he considered his business conduct to be as honorable and moral as he tried to be in his personal life. When asked, even many years after the Government had broken up the business he had so painstakingly consolidated under Standard Oil, what he would have done differently, his answer was: "It was right before me and my God. If I had to do it tomorrow I would do it again the same way – do it a hundred times." Convinced of the sincerity of the great monopolist’s profession of innocence in the business realm, and having thoroughly dispatched the conventional cynical explanation that his apparent personal virtue was a sham to distract an angry public and cover up his business sins, Chernow ends his excellent biography still bewildered at the seemingly contradictory process by which "[t]he fiercest robber baron had turned out to be the foremost philanthropist . . . [and] . . . so much good had unexpectedly flowered from so much evil."

    But what if Chernow and the conventional wisdom are wrong? What if Rockefeller is right? Could it be that all those predatory practices were in fact legitimate business initiatives creating productive improvements in the organization of economic activity, enabling constructive evolution of a more efficient socioeconomic infrastructure? Never mind the confused gibberish pouring from every accused monopolist’s (or his lawyers’) mouths about not being a monopoly. Given the bedrock legal status of antitrust and its associated regulatory tar pits, the dissembling is understandable, for by it the monopolists may escape those tar pits on technicalities by pointing out the many contradictions of antitrust as applied over the years. But assume away, for a moment, both the inconsistencies of antitrust and any doubts that the great reputed monopolies of our time – Standard Oil, AT&T, Microsoft, Intel, etc. – are indeed monopolies as charged. Is it possible that we should be applauding and encouraging the very practices of which they are accused? Could it just be that it is the robber barons – and not the trustbusters – who are on the side of the angels?

    There can be no doubt that Rockefeller guided Standard Oil to practices that were considered restraints of trade even before the Sherman Act was passed in 1890, and became illegal thereafter: secret deals with railroads to receive rebates on Standard’s shipments of oil, and even "drawbacks" on the value of oil shipped for competitors; threats to withhold product from distributors who insisted on also carrying competitors’ kerosene; threatening dry goods stores that would not exclusively carry Standard kerosene that Standard would put them out of business by opening nearby competing stores and underselling them in every product; buying up all the locally available barrels or tank cars that competitors might ship their product in; using the local tank car monopoly to pressure railroads into favorable shipping arrangements; buying up almost all competing refiners in Cleveland, who were reeling from Standard’s transportation advantages, then using this monopoly and allied rail and pipelines to pressure the rival Pittsburgh refiners into submission, too – and on and on and on.

    As clearly in violation of what became antitrust law as such practices were, Rockefeller believed that bringing a reliable supply of safe and affordable kerosene to the world was a boon to business, the economy and society. Standard’s actions, whether one approves of them or not, fostered coordination among otherwise raucous competitors, who previously had engaged in ruinous price wars, causing wildly fluctuating prices, inconsistent product quality, and uncertain supply. While such conditions are not unexpected as new industries get underway, for those industries to mature into efficiently functioning elements of the economic infrastructure, they need organization, some means of bringing order to the chaos.

    Under Standard coordination, the oil industry became a unified network, from drilling through refining to distribution, delivering a steadily increasing supply to a continually expanding market at reliably declining prices. As essentially proprietary owners of an industry, the Standard crew was constantly improving its efficiency through the application of research and productivity-enhancing initiatives, the boldness of which could hardly have been attempted by any of its smaller competitors. This is particularly so, since the benefits from any improvements initiated by non-dominant companies would most likely flow to their competitors, including Standard, who would copy them. Antitrust, in contrast, strives to make all companies in an industry non-dominant, thereby undermining their incentives to improve the overall industry. As leader of the Standard trust, Rockefeller himself was ever willing to make bold personal bets, from buying his first oil company at auction for $72,500 (although he "feared for my ability to buy the company and have the money to pay for it") to using his own money to pay for research into refining technologies when he had failed to convince his fellow board members of its value, to standing ready to buy out the shares of any wavering partners. Rockefeller’s trust strategy caused not only the oil industry to coalesce into a unified network, but was a significant contributor to the knitting together of rail and pipeline networks, too, providing valuable transportation infrastructure improvements even before oil’s "killer app" – the automobile – came along.

    While Rockefeller was not hurt financially by the breakup of his company in 1911 (the automobile’s use of oil multiplied his wealth several times over even after he retired), the same cannot be said of the nation. Without the organizing influence of the Standard Oil monopoly, Government stepped in to fill the void with a neverending string of allocation schemes, price controls, subsidies, small refiner entitlements, state-by-state prorationing, station-by-station gas rationing, 55 mile-an-hour speed limits, energy-saving tax credits, mandatory import controls, use of giant salt domes as Government storage facilities, and so forth. While the goals of all these interventions were well intentioned, such as price stability, reliability of supply, national energy independence, and the everpresent "fairness," the results were in every case the opposite of those intended. According to Dominick T. Armentano in Antitrust and Monopoly – Anatomy of a Policy Failure:


The predictable result of this intervention process was the piecemeal creation of a crazy-quilt system of regulatory privileges and punishments that made no economic sense whatever, but necessarily generated great uncertainty and a vast misallocation of energy resources. In short, business and government interventionism led inexorably to the energy crisis of the 1970s and early 1980s.


So, instead of an American monopoly, we have OPEC. Instead of reliable supply we have gas lines and salt domes. Instead of steadily declining prices reflecting productivity improvements, we have prices shooting from $3 to $40 per barrel. Then, just when economists predicted that "deregulation" (a misnomer that really means the continued application of antitrust – more on this later) would push them over $200 in a neo-Malthusian apocalypse, they fell to under $9, at which point economists predicted they would go below $5 just before they shot back up toward $20. And this instability continues to this day, with oil, according to the April 22, 1999 Wall Street Journal, "behaving like an upstart Internet stock" — (a characterization that should not surprise, as we shall see). Last Fall, once again, just when gloom was so deep that economists were predicting prolonged sub-$10 and maybe sub-$5 prices, and industry profitability was so low that re-merging some of the old pieces of the Standard monopoly was contemplated, crude began, according to the Journal, "a stunning 60% rise in only two months [as] analysts predicted that $20 isn’t far off." One wonders what Rockefeller would have thought watching one U.S. energy "Czar" after another fail to wrest energy independence from a foreign cartel. Seeing us schizophrenically switch sides between petty potentates every few years depending on which one’s zealots seemed the greatest threat of the moment to the flow of oil through the Dardanelles, would it have occurred to him that all of these continuing crises in the Middle East and elsewhere might not be so intractable were it not for our dependence on foreign oil?

    It is unknown, of course, whether – in the absence of antitrust – Rockefeller or his successors in oil and related industries could have parlayed Standard’s dominance into a freer hand for the United States in world affairs, or organized its energy resources more efficiently than the Department of Energy. Perhaps so – he could hardly have done worse. But the more important question is this: Should the right to win a monopoly, and thus take a proprietary interest in the efficient operation of a whole industry as a unified network, be considered a fundamental property right? Antitrust effectively confiscates monopolies in the public interest, under the assumption that multiparty competition to provide the network service is worth more to society than the efficient functioning of the network as a network – unified and coordinated. But is this assumption justified?

    Rockefeller was proud of how his trust substituted "coordination" for competition, and admired other "modern" industrialists, particularly railroad magnate Jay Gould, for his understanding of its value. But he did not speak much publicly about such matters, sensing that "modern" would not beat "monopolist," nor could "coordination" trump "anti-competitive"– much less "robber barons" – in the court of public opinion. Jealousy was then, and is now, a powerful source of antitrust support with the public (however much antitrust purists deny that wealth, or even monopolies "fairly won" are violations). Rockefeller’s reticence to mount an intellectual defense of his practices is, therefore, understandable. Besides, he was not an academic, and even they did not come across the concept of "network effects" until many years later. According to Carl Shapiro and Hal R. Varian in their new book Information Rules, network effects were first discussed in a 1974 paper by Jeffrey Rohlfs entitled "A theory of Independent Demand for a Communications Service." Probably not Rockefeller’s cup of tea even if it had been written a century earlier. But the fact that antitrust laws were formulated without any understanding of how important and valuable networks, and, therefore, monopolies, could be, explains the mystery behind Robert Bork’s statement in his 1978 book, The Antitrust Paradox: A Policy at War with Itself:


One reason for the stifling solidity of received opinion about antitrust, why counterarguments make so little headway, is that most of us accept our first principles and even our intermediate premises uncritically, as given, because we assume that they were established theoretically and confirmed empirically by legislators and judges long ago. Discussion begins from there. . . What we all "know" is wrong. We are working from an intellectual base that does not exist. . . . If modern results often appear sensational, that is . . . because we never really understood the sweeping implications of the old ideas.


Thus, as hard as it is to imagine that such an important new policy as antitrust was only a shoot-from-the-hip attack on wealth, it was just that. Although elaborate theories of "perfect competition" were eventually developed by economists to justify what was then, and is now, primarily motivated by an essentially redistributionist impulse, they clearly took no account of network effects and, thus, were oblivious to their real-world impact. That is why, time after time, we see antitrust actions cause surprising results – the infamous "unintended consequences" – while policymakers continue to trumpet their trustbusting triumphs with circular reasoning that assumes their efficacy in the face of all evidence to the contrary. Even those who now recognize the potential benefits of coordinated networks have not yet begun to question antitrust’s "first principles." And so we have the bizarre prospect of industry after industry suffering deterioration of its core functions under antitrust, while academics, regulators and the public at large continue to believe that more and more of this bitter trustbusting medicine is just what the doctor ordered.

    To gain a better understanding of how breaking up networks under antitrust is bad policy, both because it undermines our fundamental rights of economic action, and because it always destroys the value of the network product, let us turn now to an examination of these issues in an industry that is much easier to recognize as a network: telephony.


Phony Deregulation


The Telecommunications Act of 1996 was hailed by all and sundry as a bold "deregulation" that would shock the presumably sleepy local telephone monopolies into rapid innovation, aggressive competition, and improved service. Their known desire to get into long distance, the reasoning went, provided regulators with the basis for a deal they presumably couldn’t refuse: Give your competitors – such as local service providers from other regions and long distance companies – access on reasonable terms to your local networks of customers, and we’ll let you pursue long distance. The expectation was that this supposedly brilliant quid pro quo would lead to a general free-for-all with everyone getting into everyone else’s business, thus improving the quality of each service through multiparty competition for that service.

    The result? Not even close. Instead, there has been legal wrangling over who can compete with whom and under what circumstances, fights over which regulatory bureaucracy will make the call, customer confusion over the many potential combinations of service offered by a proliferation of fly-by-night providers making annoying solicitation calls at all hours offering repackaged versions of other companies’ products, a bunch of 10-10 "dialarounds" inviting us to benefit somehow by memorizing even more digits and the special circumstances in which using them might save us money, threatening letters from your long distance company to call 1-800-BLA-BLAH "immediately to complete your account" as if not switching to them for local service, too, will have dire consequences (and perhaps it will), "slamming" customers to new providers without their permission, "cramming" them with services they didn’t order, bills no-one can decipher, time wasted dialing extra digits, redialing them when memory fails, fumbling in phone booths with cards and slips of paper to get the right number sequences for the situation, enduring endless "thank you for using tra la /&/ la la"s when often you had no intention of using them, did not realize you were using them, or had never heard of them – and in any case don’t know what they charge or how they are billing you. And, while the average consumer has clearly lost touch with who is responsible for service or can fix a problem, there are no visible signs yet of any improvements in service in spite of the many promising new technologies. To top it all off – and with magnificent irony – instead of competing as the quid pro quo envisioned, all the companies are aggressively attempting to merge or otherwise reassemble something like the old AT&T monopoly.

    What’s going on here? How could all the experts have been both so sure and so wrong about what deregulation would bring? What claim have they to launch such grandiose schemes when they clearly know so little about their purported area of expertise that their schemes cause the opposite of what they expect, mergers rather than competition? The answer, again, is that they failed to take account of network effects in making policy, or when assessing the results of policy. This enables real-world effects perceived by the average consumer to diverge entirely from the picture seen by bureaucrats. While the former feel disaster, the latter see that, sure enough, they are successfully encouraging competition – so everything must be OK, or at least on the road to OK. But a realistic assessment of the effect of antitrust would begin by forgetting all that nonsense about how antitrust allows the free market to function with the least amount of interference. The first thing to recognize is that antitrust, itself, is incredibly interventionist. Telecom deregulation is far more tangled by red tape and bureaucratic bungling than any ministry of the Soviet Union would have made it. No wonder we have so many unintended consequences, when our intentional reforms are based on so little understanding of economics, regulation – or, for that matter, the English language – that we can refer to a massive increase in Government intrusion as "deregulation."

    A more accurate and realistic definition of what is meant by "deregulation" of an industry is that Government will more aggressively enforce the antitrust laws in that industry. Ignore the nominal removal of formal Government rate-setting or rate-approval processes, such as in the deregulations of air travel or crude oil, and ignore the apparent relaxation of Government-maintained barriers between potential competitors, such as in the current telecom deregulation and in the electric utility deregulation that will be unfolding soon. These are mere sideshows, trumpeted for headline value, which seldom actually result in diminished roles even in the specific areas of rate-fixing or barrier-maintenance from which Government is purportedly removing itself, much less in an actual diminution of Government’s role net of the usually massive increase in antitrust activity. The real story of deregulation is that, since antitrust runs counter to the natural process of business formation, which often tends toward concentration and monopoly, a plan to deregulate is tantamount to a plan for Government to take over and manage any further design and operation of the industry’s service. By declaring the service to be an "essential facility," Government basically expropriates the private property of whatever company or companies had originally created it, and, from that point forward, through law- and rule-making, effectively doles out industry operating licenses to those companies that meet Government criteria as approved competitors.

    The assumed benefits of deregulation, like antitrust, include lower prices due to competition, more innovations in service delivery from the proliferation of competitors, improvements in the core essential facility’s features, and all these with minimal or no disintegration of the core service due to lack of coordination among the competitors. On inspection, none of these assumed benefits pan out:


Inappropriateness of Price Control. Price control is an issue for the Federal Reserve, not trustbusters. On a macro level, looking at prices across the economy, attempts at inflation control by the Antitrust Division, the FCC, the FTC, etc. only complicate the central bank’s job. General swings between success or the lack of it in antitrust enforcement can cause disturbing swings between deflationary commoditization and inflationary gouging, influences that can be as complicating to the Fed’s goal of price stability as, say, OPEC. While, over time, such disturbances can be offset through appropriate monetary policy, they can, as OPEC proved, be very disruptive in the short run. On a micro level, looking industry by industry, how could it be beneficial to only suppress prices of those industries the trustbusters can successfully attack and commoditize? And what about the price signals that come from our paying more for those things we want more, and less for those things we want less, as industries – including some monopolies – compete for our dollars? Given the importance of price signals to the efficient allocation of capital, economists are way off task when they invoke antitrust "efficiency" to suppress them.


Retarded Innovation In Service Delivery. Once the core service of an industry has been effectively nationalized by declaring it an essential facility, there is no reason to suppose the competitors licensed by the Government to deliver it will engage in further innovation – quite the opposite. Not having a proprietary interest in the core service, their primary incentive is to free-ride on the public expectation of quality promulgated by the Government and their competitors, while cutting corners to reduce their own expenses. Since any innovations they make in delivery of the core service can be copied, any capital spent in support of such innovation will inure largely to the benefit of their competitors. So they make few significant innovations in service delivery, and generally engage in a race to the bottom in service quality. A good current example is air travel, where, according to a New York Times article on April 2, 1999, customers are seething with anger over "poor service, scant leg room, lost or mishandled baggage, incomplete or misleading information about delayed flights, overbooking and frustration with frequent flier miles that are easy to accumulate but next to impossible to use." Here is another industry where deregulation has produced so much disappointment that the Vice President and many Senators and Congressmen are supporting "passenger rights" bills simply to require the basics: honesty, courtesy, fair treatment etc. And once again, the experts in their conventional wisdom are oblivious to the possibility that it is Government intrusion through antitrust-managed deregulation that is the cause of the problems they are now tripping over each other to solve with more Government intrusion.


Retarded Improvements in Service Features. As to the ancillary features of the core service, since any improvements made by any of many competitors will benefit everyone else, too, no single competitor has an incentive to spend capital on such improvements, unless in the context of an industry-organized and Government-approved program. Since such cooperation is very difficult to obtain, and since, if obtained, it often looks highly suspicious to trustbusters anyway, improvements are next to impossible to bring to fruition. Thus, taking the air travel example again, such common-to-all determinants of the quality of the air travel experience as airports, parking, airport food service, baggage handling, and even critical components, such as passenger and baggage security screening, air traffic control systems, or oversight of safety maintenance – all of these are bound to continue deteriorating, since no single or feasible group of competitors has a sufficient proprietary interest in the industry’s core service to address them.


Disintegration of Core Service. The problems described in the above two paragraphs relate to the ineffectiveness of antitrust deregulation to improve the delivery of the core service (e.g., courtesy, reliability, consistency, responsiveness to problems etc.), or to improve the ancillary features of the core service (e.g., in air travel, more legroom, better food, less lost luggage, or, in telecom, smaller, lighter phones with more bells and whistles, higher speed lines for data, better voice quality etc.), whether such improvements are initiated by individual competitors or Government-blessed coalitions of them. As much as antitrust undermines delivery and ancillary features, such damage pales in comparison to the harm done to the core service itself, which, believe it or not, it is the policy of antitrust to effectively dismantle. This result follows inevitably from the antitrust fixation on multiparty competition for all products and services, even those whose value derives primarily from their monopoly status as unified networks to which everyone is connected. No friendly service or fancy features, even if they existed, can make up for your frustration if your phone network only includes some of your friends, family or business associates. But multiple competing networks will always produce this result, in spite of all efforts to counter that fragmentation with mandatory linkages to other networks, fair access requirements, "must carry" provisions, and all the other bureaucratic pipe dreams meant to make a refereed contest work like a network.


Network Nirvana


    Perhaps a simple diagram will help clarify this point. Imagine that Network #1 was created in response to desires among A, B, C and D to communicate with each other. Assume that there are no other people who want to communicate with each other, or with A, B, C or D, so that ABCD Network Corp. has a 100% monopoly in network communications. While A, B, C and D are seemingly happy with the service, trustbusters make the argument that its monopoly is enabling ABCD Network Corp. to engage in some combination of price gouging and failure to innovate, or that it is using its monopoly to gain advantage in another industry that ABCD Corp. wants to enter. To any or all of these problems, trustbusters say, network competition is the answer. So they encourage the entry of Networks #2 and #3 into the network communications business. That is when the problems begin.

    Network #2 only connects A, C and D, while Network #3 connects only A, B and D. So Network #2’s customers who want to communicate with B, and Network #3’s customers who want to communicate with C, must rely on trustbusters to compel Network #1 to let its new competitors use its connections to customers they don’t have – or don’t have yet. Invoking the essential facility doctrine, the regulators make ABCD Corp. lease its network at nominal rates to its competitors. Ignoring issues of equity, property rights, bureaucratic sprawl, legal snafus and other obvious problems with such arrangements, focus for a moment only on the network itself: Is it likely to give communications service to A, B, C and D that is equivalent in quality to what they had before the trustbusters got involved? Not if there are any incompatibilities in the protocols or technologies used by the various networks. Not if there are any hiccups as one network switches to another to complete a call. Not if the switching process slows data delivery or hampers voice quality or introduces "noise" on the line that can break a connection and force the caller to start over. And not if there are redundant capital costs that must be recovered because three different networks have built connections between A and D.

    What if these new competitor networks manage to sign up all the customers: A, B, C and D? This is, of course, Nirvana to trustbusters – multiparty competition for the exact same service. But, for the reasons just described, this network Nirvana to trustbusters is actually a network nightmare to customers, who find that the more choices they have, the more their calls seem to get dropped, the more their bills become indecipherable, and the more they seem to have the wrong friends and family. The reality is that "perfect competition" to antitrust economists is fundamentally at odds with the very concept of a network, which implies the unified, seamless connections among each other that A, B, C and D had found so attractive. So long as they must choose among competing networks, the consequent network-switching to complete calls and the resulting service degradation are inevitable, because no amount of quasi connections and access rights mandated by regulators can reassemble the smooth functioning of the original. With their inherently different technologies, equipment, protocols etc., the competing networks under deregulation can only produce a fragmented, Balkanized hodge-podge that actually separates those they are trying to connect.

    Worse still, the nightmare caused by breaking up ABCD Corp. in a domestic deregulation will be multiplied many times over in a globalizing economy. As trustbusters from every country – and international competition tribunals, such as those at the World Trade Organization and the European Union – have their say, the technically feasible linking of everyone worldwide in an A to Z network will be forever prevented by fear of a dreaded AtoZ Corp. monopoly. While individual companies like MCI-WorldCom will continue to brag of their unified global networks, these will always fall short of potential, because only small subsets of global customers will be allowed on them. Too much market share and the trustbusters will step in to stop them, thus blocking the main benefit communications technology has to offer: unified, unhindered connections to as many people as possible. Oh sure, they’ll maybe let another Baby Bell or two merge, maybe even Deutsche Telekom and Telecom Italia (perhaps out of guilt or worry among regulators that Deutsche Telekom lost 30% of its long distance market share in the 16 months since Germany deregulated its telecom market). But the technically possible dream of a seamless global communications network – perhaps even the Internet – will continue to be frustrated by misguided antitrust concepts.

    And that is before factoring in the intrusive bureaucracy trampling on our economic freedom that is needed to produce this result, before the redundancy costs from multiple lines, switches, companies and capital applied to providing the same service, before the enervating effect on risk capital that comes from requiring companies to lease expensive facilities and technologies to their competitors at an always falling replacement cost, and it is before considering that such purported antitrust benefits as price reduction and greater innovation are either nonexistent or were inappropriate goals in the first place. Even if all these unintended consequences could be ignored, antitrust would still cause a dead weight loss of economic efficiency, because the intended effect of multiparty competition destroys the networks needed to organize an increasingly complex and an increasing global economy.

    Although real industries are not as obviously subject to network effects as that monopolized by ABCD Network Corp. in our fanciful example, many others besides telecom share its features and problems. It may not be as obvious, for example, that the customer who described in the Times article "a lack of concern from most of the six airlines he dealt with" on one trip was suffering from the inability of the industry under deregulation to connect up as an air travel network. But the tipoff to the similarity between his situation and that in telecom can be seen in the nearly identical actions in Washington to address them. There, activists promote such paternalistic solutions as House Transportation Committee Chairman Bud Shuster’s "Airline Passenger Bill of Rights" that would mandate the return of lost items with names on them. Similarly, in telecom, according to the April 12 Wall Street Journal, "[t]he Federal Communications Commission is expected this week to propose a set of guidelines requiring phone companies to issue clearer bills."

    Less trivially, the Clinton Administration has proposed a "Patients Bill of Rights" to deal with the inconsistent service and escalating cost of health care. But what would you expect, when years of intervention through Medicare, Medicaid and competition policy have finally divided the health care industry into multiple HMOs? As in airlines and telecom, their incentives are to free-ride off the expectations of quality care fostered by their competitors and Government guarantees, while cutting their own costs by making it difficult to get the expected and advertised care. Why else would they treat supposedly valuable customers so badly that comedians now routinely mock them and movies and TV shows feature HMO villains? It may be hard to recognize the old fee-for-service system as an effective network that was broken up into competing HMO gatekeepers, and certainly there are differences between how network effects operate in industries as diverse as telephony, airlines and health care. But note the similar frustrations: switching on one trip between six airlines that don’t care about you, fuming that your phone bill seems to reflect charges from multiple sources, each protected from your complaints by interminable phone trees, and the inconvenience, mountains of paperwork, and suspicious stares that must be endured just to go to a doctor from another HMO. All three of these are characterized by multiple competing sources of a common service and diffuse responsibility for its quality.

    Even less trivially, the misery of health care consumers may be overtaken early next century by that of electric utility customers. Just as antitrust regulators and politicians have touted the efficacy of consumer choice among competing airlines, phone companies and HMOs, the Clinton Administration has proposed, according to the April 15 Wall Street Journal, to deregulate electric utilities so that "consumers [will] be able to choose their electric companies by 2003." Why? "Department of Energy officials predict that the legislation would help the economy by increasing competition." Given that this is the same Department of Energy that has presided over both competition in oil and repeated energy crises, its predictions regarding the benefits of competition and customer choice in electricity are hardly reassuring. Nor is the fact that "[t]he bill would also create a new board, the Electricity Outage Investigations Board." Nonetheless, following the same deregulation pattern as telecom and other industries, the bureaucrats at FERC (Federal Energy Regulatory Commission) would assume vast new powers to organize and referee the competition and, thus, to dole out operating licenses. According to the Journal, "[a]mong the commission’s new authorities will be the power to order competitive access to transmission lines, create regional grid-management agencies and review the effects of utility mergers." But do consumers really want to shop for electricity? Or should we expect yet another "bill of rights" to crop up in Congress as the confusion, price swings and outages begin?

    Seeming to reflect deep philosophical confusion, these last two industries appear to be moving simultaneously in opposite directions: health care careening toward nationalization, electricity lurching toward deregulation. Part of the problem is our inordinate fear of "natural monopolies," and our lack of appreciation for the organizing value of network effects in their creation. Network effects, if recognized or understood at all, are considered dangerous, rather than beneficial, because they lead to these natural monopolies, which, it is assumed, must either be 1) nationalized, 2) regulated as single utility companies, or 3) "deregulated" by trustbusters. The philosophical confusion comes from the fact that, although all three remedies result in effective Government takeovers, deregulation is incorrectly touted as a free market solution. Since deregulation actually requires the most intervention, the normal philosophical alignments are confusingly reversed. For example, it is the Republicans and conservatives – normally associated with smaller Government – who unwittingly support the more interventionist approach (deregulation), lured by a naive understanding of what their free market rhetoric actually promotes in this context, while the liberal Democrats – by resisting deregulation – unwittingly push more laissez faire approaches.

    As if that weren’t confusing enough, because both nationalization and utility regulation have a reasonable chance of retaining some network effects, these admittedly interventionist approaches can easily work better than deregulation – the supposed free market approach. In any case, every new "bill of rights" proposed to mollify frustrated consumers in a deregulated industry is pitched as needed to solve the problems caused by unbridled free market capitalism, when the problems are actually the result of an unrecognized turn to socialism. Clearly, given the growing number of consumer complaints about service in deregulated industries, it is time to consider whether the policy that has been in place while these problems proliferated might be the culprit. The answer may be outside the box where no one is even looking, perhaps even what one politician recently called "the worst of all worlds, which is an unregulated monopoly."


The Dynamic Duo


    Such facile snippets of Washington wisdom are expected to be swallowed hook, line and sinker. But investors, at least, might want to note the similarity of the situation it describes to that in the high tech sector driving the economy and bull market. As even casual followers of the current Microsoft and Intel antitrust cases can readily see, both companies are monopolies (although, for reasons of legal liability and the aforementioned regulatory tar pits, neither can admit it), and – until now, anyway – both were largely unregulated. Sure, they’ve had their run-ins with trustbusters, and Intel reportedly puts its employees through regular antitrust training to school them in how to avoid the tar pits. But both have had relatively free hands until now to engage in a variety of predatory practices, such as were exposed to the satisfaction of most observers in the ongoing Microsoft trial, and not disputed by Intel in its recent settlement with the FTC. And, as both of their defenses convincingly argued, their practices were – and are – common in the industry, although others are not large enough yet to have wielded them so effectively.

    They clearly bullied suppliers, distributors, competitors, software developers and even, on occasion, partners, into giving their products favorable or exclusive positions versus their competition, often in flagrant enough fashion as would have made old man Rockefeller proud. In one such example, Microsoft once sold its operating systems in deals requiring distributors not only to pay Microsoft a low bulk rate for every computer they sold with a Microsoft system, but also for every computer outfitted with a rival’s operating system. In their confidentiality, in their discriminatory discounts, in their inducements to not deal with competitors, such arrangements contained all the essential elements of the secret deals on rebates and drawbacks that Rockefeller cut with the railroads, and for which he was roundly vilified for the rest of his life. But, like them or not, just as Standard unified the oil industry through such tactics, they enabled the Wintel standard to ride herd over the most explosive expansion of economic potential in human history. The prodigious network effects both exploited and generated by this Dynamic Duo not only created two super successful companies – as well as hundreds of merely very successful companies drafting behind the standard – but contributed to the commercialization of the Internet, which may one day spawn competitors that could replace Wintel.

The essence of their strategy was to gain market share by wielding the interlocking network externalities that resulted from coordinating their respective products: Intel’s chips with Microsoft’s operating systems, and both of them with applications software developed by Microsoft or independents. The strategy apparently included at least tacit understandings of divided territories: Microsoft would not develop chips nor easily deal with Intel competitors; Intel would not develop software nor easily deal with competitors of Microsoft; independent applications programmers were welcome to write for DOS or Windows, as long as they were able to accept the fact that Microsoft’s own applications developers probably knew a little more a little sooner about OS changes than they did, and – if an independent did come up with a hot new idea – Microsoft might eventually want to drag and drop it into the OS. Harsh as such tactics may be, they did not prevent many of those disappointed designers from becoming millionaires. In any case, the success of this alliance meant that PC customers preferred Wintel, which made it possible to commit the multibillions for newer and better OS releases and chip fabrication facilities, which, in turn, brought an even more loyal installed base of Wintel consumers, which caused OEMs and peripherals manufacturers to conform to the standard to get their pieces of the installed base, which compatibility, again, brought more installed base and a greater willingness to make the next multiyear multibillion dollar investments in OS upgrades and chip fabs.

    Such powerful network effects rolling up increasing returns in a positive feedback loop often result in a winner-take-all monopoly. That is because an installed base of customers is not valuable only for its size, but because it can be a complete network exclusively encompassing all related parties. The closer such a network has to a 100% market share, the better it is for participants, and the more beneficial its effect on the economy as a whole. This notion goes farther than Metcalfe’s "law," (after Bob Metcalfe, inventor of the Ethernet and founder of 3Com) which says that the value of a network is equal to the square of the number of people connected to it. While this rule of thumb neatly captures the great and exponentially increasing value of ever larger networks, the value of a network – particularly to the economy – is related both to its size and to its market share, from which springs its ability to compel the coordination of economic activity. More important still, because the capacity of standards to compel coordination is so critical to growth, the value of a network is also related to the extent of its reach into and across the economy.

    The definition of "standard" to this point has been loose – and it’s about to get looser. We speak of Wintel as a standard. But we could also speak of Microsoft or Windows or Office, or Intel or Pentium II, or any one or any combination of their products or product lines. A broader understanding of "standard" would also include peripherals and related compatible products, like printers or faxes, and any of the commonly used communications protocols or connection devices that enable their operation. We could even include the loyalty of consumers "locked in" to buying only standard-compatible products, because that’s what they already own and know how to use, or software developers not wanting to waste their time writing to a wannabe standard with little market share and, therefore – in a positive feedback, increasing returns environment – a high likelihood of failing altogether. All of these factors are either visibly part of, or intimately involved in, creating, defending and extending the standard, however defined.

    But, it should be clear when contemplating those parts of the standard that are far away from its center, that this game could go on and on. Why not include remote externalities that affect and are affected by the success of the standard, such as the educational facilities that sustain a pool of trained developers, or the economic environment that nurtures demand for the standard’s products? While these may seem far-fetched with respect to narrow definitions of any standard, it is necessary to consider them, too, if one is to gain an appreciation of 1) the importance of standards to coordinating, energizing and sustaining economic activity, and 2) the danger to all three posed by antitrust. In particular, by promoting never-ending multiparty competition for each component of the standard, or, worse yet, Government-monitored standard-setting committees, antitrust both retards the formation of commanding standards, and boxes in the ability of those that do form to extend their benefits widely into the economy. Now let’s consider how the primary intended effect of antitrust – prevention of the pursuit or extension of monopolies – undermines business and consumer confidence and, thereby, the economy.


Standard or Stranded


    For any given standard, once non-conforming competitors to any of its elements are brought to heel, the risk of leaving its users "stranded" disappears. Once they no longer have to worry about wasting time, money, training, research or capital on any activity, project, technology or system that depends on the standard’s continued dominance, consumers are emboldened to buy, and businesses are emboldened to hire more people and install new productivity-enhancing systems. Their purchases, hires and installations, in turn, generate further network effects off the original standard to which they conform, which builds confidence for everyone in the long chain of economic activities proceeding from it. In effect, they, too, are part of the network, even though they may have no knowledge of the standard’s core components, and even though their participation in and contribution to its network effects are indistinct and remote. But any adoption of the original standard or its derivatives extends confidence in its continuation and, therefore, engenders more aggressive economic engagement. Thus, the stronger the standard’s position, and the broader and deeper its reach, the more the imprint of its presence blends into the general economy. At some point, its penetration becomes so significant that its power to organize, energize and sustain economic activity is simply felt as consumer confidence.

    None of this means that competition with the original standard – even a very large, dominant and entrenched one – is over; far from it, as the Internet is teaching Microsoft. What it does mean is that, rather than perpetual antitrust-refereed battles producing neither victors nor standards, the contestants – including monopolies and aspiring monopolists – compete for a greater share of the "value proposition" perceived by consumers when they purchase a product. In this kind of competition, aspirants try to increase their relative value by such means as horse trading their installed bases with other monopolists to more felicitously align component combinations, whether by merger or strategic alliance, or by cutting prices. Ultimately, they all hope to latch onto a major monopoly of their own – even those who already have one – a new killer app that can displace the reigning gorilla, and confer on the newcomer all those strategic advantages for defending and extending a monopoly that come with standard-setting.

    Think of the relative amounts received on a PC purchase by Dell, Microsoft or Intel, or of the value to each of them of the customer loyalty that arises from that purchase, or how those might change depending on whether it is used primarily for, say, business, personal finance or Web surfing, or how all of those answers would change over time. Or look at what an observer in the April 27, 1999 Wall Street Journal called "one of the most complicated chess games I’ve ever seen" among Comcast, Microsoft, AOL, Time Warner, AT&T, At Home, TCI and Media One, all jockeying for positions in related, but different, services, like digital television, broadband Internet access, local and long distance telephone services and video. Who or what combination will hit on the right concoction of inclusive or exclusive arrangements, cooperative or competitive stances, to carve "mind share" and installed base out of what is, after all, a relatively fixed pool of potential consumers? Such constantly shifting borders and intricate dances between potential competitors and allies will always bring about turnover among standards and monopolies far faster than the settled precincts that congeal around rent-seeking and red tape. But more important than the rate of turnover among monopolies and standards is the protection that monopoly-seeking behavior provides consumers and businesses that progress and turnover can occur without leaving them stranded with unfinishable projects, orphaned technologies and obsolete products. This cannot be said of the upsetting changes that are induced when networks are broken apart by antitrust and standards are not allowed to form.

    Perhaps the most important benefit of high tech in recent years has been the inability of trustbusters to figure out how to bust its practitioners. Poor initial grasp of such factors as the complexity of and rapidity of changes in the technologies, the unfamiliar dynamics of an information economy riddled with network effects, and the difficulty of demonizing nerds – even billionaire nerds – combined to give the Dynamic Duo and its followers a nearly free run at capitalism the old fashioned way. Like the robber barons at the end of the last century, they made the most of it. Even with new technologies, products and services proliferating at a blinding pace, the expected confusion and fragmentation was held in check – indeed, organized and harnessed – by Wintel. This alliance has been, in all likelihood, the prime reason for the long stretch of powerful and accelerating economic growth we have experienced, which continues to confound and surprise all the experts. And, to be perfectly clear, the value came not primarily from all the great products or productivity-enhancing applications that this monopoly produced, but from the very fact that it was a monopoly with the market power to compel coordination and instill confidence it its continuation. But the jig, it appears, is up.

    The ultimate dispositions of the individual cases against Microsoft and Intel are no longer important. What is important is the fact that trustbusters now know how to prevent the predatory practices the Duo used so successfully to organize and energize the economy. Like kids caught on a caper, they are now being interrogated separately (Microsoft by Justice, Intel by the FTC), told they can’t hang out together anymore, or ever again do any of those bad things that so upset the judge’s neighbors. And, as regulators no doubt intended, the example set by their punishment will intimidate all their friends, too. So, while native ingenuity continues to rev up the forces of fragmentation, none of the old gang will risk what happened to their leaders by trying to get everybody together again. And, even if they – or their younger brothers – did, the cops now know where to find them.

    Whether Microsoft is split into Baby Bills, Windows is nationalized, or Justice merely reviews all contracts and new features for Windows – just to take three recent suggestions – the good times are over. Old standards are disintegrating and new ones won’t form, so the fear of making bad choices is rising rapidly and the risk of getting stranded is soaring. Multiple suddenly credible operating systems, each building its own installed base of devotees in separating development communities; various versions of Unix (still); resurgent Macintosh (if you can believe it); Java (that supposedly runs on anything – but doesn’t) dividing developers over where to devote their time; free and open source systems like Linux, and, reportedly, maybe even Windows (perhaps to show Microsoft’s contrition, or maybe because it doesn’t matter anymore) – all of these fragmenting features of the new landscape, while demonstrating clearly the continued vibrancy of the industry’s potential for innovation, also raise the specter that, without the most important innovation of all – coordination – much of that potential will turn to sand. Although any of these systems could emerge as the new standard, it is now enforceable policy that such a result will never be allowed to happen again. And, lest you think that the monopolists will simply migrate to the Internet and set up the new standard there, remember that the Internet is a communications network, and communications networks are – unfortunately – familiar territory to trustbusters.

    Meanwhile, Intel is no longer swearing off software. In fact, the chip giant appears ready to move far away from its concentration on microprocessors. So far away that, according to the company’s prediction in an April 23, 1999 Wall Street Journal story, "in two or three years about 90% of Intel’s revenue will come from e-commerce." The rapid commoditization of the chip market – which the reformed Intel is now virtually powerless to defend against – is apparently not as attractive as its planned "new Internet services and networking ventures." These will provide outsourcing services to ISPs, called "Internet hosting." While that may prove an interesting business, it is unlikely to have the coordinating value to the economy that Intel’s half of Wintel did. One analyst quoted in the article wondered "what advantage Intel has over everyone else who wants to be in this business."


The Last Bust


    Mayday – May 1, 1975 – is the day deregulation began for the U. S. stock market, ending the NYSE’s fixed commission regime, and requiring that competition with the Big Board’s specialist by other dealers be allowed in a "National Market System." To facilitate that competition, regulators required the creation of automated, publicly available transaction and quotation reporting systems. While the conventional wisdom holds that these reforms ushered in a more modern and efficient U. S. stock market, it is my belief that, like other deregulations, they are actually undermining the quality of stock market services by dismantling its network – the means by which it finds liquidity, discovers price and raises capital. As I have argued in previous Auction Countdowns, efforts to improve transparency are actually destroying it, efforts to increase liquidity are undermining it, and efforts to sharpen the accuracy of price discovery are causing it to come undone with wild volatility and dangerous valuation distortions (see particularly those dated 10/26/98, 6/23/98, 3/9/98, 11/30/97, 6/19/97, and 3/24/97 – all available at Rather than repeating those detailed arguments here, this section will focus first on the similarity to other industries of the network-busting and standard-fragmenting effects of applying antitrust to stock markets, and then on some little-noticed results of these actions.

    The 1975 deregulation, as disruptive as it was then, was only a warm-up act for the micromanagement of market structure that is unfolding now. In recent years, we have seen various antitrust settlements by Nasdaq and its members, new order handling rules to facilitate compliance with the agreements made in those settlements, the new ATS (automated trading systems) rule governing regulation of exchanges and proprietary trading systems, and many other regulatory initiatives requiring detailed and drastic changes in market structure. For exchange-listed stocks, the NYSE’s specialist auction – an effective network that once centralized information about liquidity needs and pricing preferences – has been fragmenting since 1975 as members internalize orders upstairs – or sell them to other dealers – without bringing them to the floor. More recently, for over-the-counter stocks, the network comprising Nasdaq dealers and Instinet (another Dynamic Duo), has been broken up into about a dozen new competing markets, mostly called ECNs (electronic communications networks), that rely on clumsy mandated connections that seem to work better at separating buyers and sellers than at enabling them to meet – at least compared to the original network. Thus, stock market deregulation has had the cumulative effect of replacing unified networks with regulator-refereed contests among multiple competing entities poorly connected by "linkage." Sound familiar?

    Here, too, although many believe that deregulation and subsequent reforms have been successful, such apparent success is often the result of self-referential assumptions about the efficacy of antitrust. If price fixing is an antitrust evil, then "deregulating" the NYSE’s fixed commissions must be good. Right? If so, then so would requiring Nasdaq members to cease colluding to fix fat increments and spreads. How about banning clear discrimination against non-members in the dissemination of trading information and access to trading facilities by requiring trade and quotation tapes and customer-first order exposure rules? All of these reforms seemed like slam-dunks – indeed, they were slam-dunks – to improve the market, assuming antitrust is correct. Since that assumption is also conventional wisdom, it is not surprising that most people view the wrenching changes to the market under NMS as all for the best. Clearly, ridding those archaic old markets of their anti-competitive restraints of trade could only improve them. Right? Wrong, as we have seen in so many other antitrust-based deregulations. We’ll turn shortly to some signs of consumer dissatisfaction that are related to the application of antitrust in the stock market. But first, let’s look at why it is not at all surprising or incorrect for antitrust to be applied here, for there is no other industry that is more clearly in violation of its precepts.

    For all its vagueness, certain principles of antitrust stand out pretty clearly. Price fixing, for example, is a "per se" violation, meaning that all a prosecutor has to do to prove a violation occurred is to show that prices were, in fact, fixed. Violators cannot argue mitigating circumstances, good intentions, or that some kind of efficiency would result. Similarly, cartels are per se violations, such as when a group of competitors collude to fix prices, divide territories, or agree to trade only with each other. One needn’t know much of the history of stock exchanges to recognize many anti-competitive aspects of their founding agreements. And these were not mere side codicils, but were, in fact, the main event in these agreements. As strange as it may sound, stock exchanges were not formed to create liquidity, discover price, raise capital or to do any of the other beneficial things they do. They were formed so their members could engage in antitrust violations (or what, for many, would later become violations, since many major world markets predated the Sherman Act of 1890, and the newer ones merely copied their structures). The very notion of membership is predicated on an unlevel playing field, one in which members have privileges unavailable to the general public regarding such matters as access to information and trading facilities, control of trading rules so that members would be favored, and determination of who can be a member.

    No wonder regulators have been increasingly active and engaged in reforming and redesigning the stock market. Ever since Section 11A of the Exchange Act – the National Market System amendments – gave them the authority to do so in 1975, they could hardly have avoided reforming the markets under antitrust, because the law appeared to explicitly require it. Indeed, one could argue that they could have gotten involved sooner. The violations were always there, and even though, as regulated entities, the markets were exempt from antitrust attack by the Justice Department, their regulator – the SEC – presumably could at any time have used antitrust to force changes upon them. But, regulatory politics being what they are, the Commissioners probably would have feared being outgunned, had they insisted on such drastic changes prior to the years of studies and hearings they sponsored to buttress their case for the formal authority of Section 11A.

    But now that they have that authority – and are two and a half decades into using it – the markets are finally showing signs that the old structure will collapse. Traditional exchanges really have no options left, having been effectively declared illegal as antitrust reforms squeeze the life out of their raison d’etres. Just what is a market without its membership? What is the point of an exchange that not only no longer seems to represent the interests of its members, but actually wants to go off and compete with them? Desperate members around the world appear to be thinking: Why not "demutualize" and do an IPO while they still remember our brand? Meanwhile, various visionaries, pundits, and chanters of the competition mantra excitedly anticipate a world without those old anti-competitive dinosaurs, replaced at the Millennium by some supersonic electronic system-of-systems, perhaps a series of linked for-profit exchanges created out of the mergers of old auction and dealer markets who’ve thrown in the towel, maybe with a few ECNs sprinkled in -- ? Well, this all sounds very hopeful and modern, but does anyone know how it will work?

    I should point out here that I have nothing against for-profit exchanges – AZX is one. But I believe it is incredibly dangerous to so rapidly extinguish by Government fiat a structure – the membership market – that has been the most important component in the machinery of capitalism since its creation. In terms of the coordination concepts discussed earlier, this particular intervention destroys stock market networks and standards in two ways: 1) it eliminates by rule the means by which dealers and brokers had communicated so they could process the information that enabled them to provide good liquidity at stable and accurate prices and 2) because the rules that do that are draconian and detailed, requiring unnatural linkages and consistent integration with NMS processes, it is very difficult for any new market to emerge that might solve the liquidity problems with a new network. Fixed-time call markets, for example, which operate so differently from the prevalent continuous markets, must spend inordinate – and potentially interminable – amounts of time seeking authority to operate. Because they are different, regulators must ponder (and, often, ponder and ponder and ponder) whether their operations can be or need to be made consistent with the rules that were designed for continuous markets – and that includes all of them.

    Many of the regulations bear considerable resemblance to the "bill of rights" approach to consumer complaints that we have seen in other deregulated industries. Investors now have "rights" –defended by regulators – to "best execution," "price improvement," "transparency," "equal access" on a "level playing field," and, coming soon, to "decimalization." That each of these nostrums is backed up with specific rules and obligations is a source of comfort to some, confusion to others. The inherent contradiction, for example, between best execution and price improvement is so unresolveable that at least one senior regulator has been known to refer to best execution as a "loosey goosey" regulation. No wonder they refuse to define it, even as they set up ever more elaborate procedures to ensure compliance with its requirements. But what is an investor to think, if, on the one hand, he knows that regulators are assiduously watching over his executions to make sure they are all "best," but sometimes he is able to benefit from their ability to get him an even better one? How do you get better than best?

    In spite of all the efforts of regulators to improve the quality of execution, the fact is that survey after survey identifies distrust of their brokers as a prime concern of investors. It appears that, once again, we have a deregulated industry in which competitors have little incentive to actually improve service. Rather, once 11A declared the stock market to be a "national asset" – similar to the "essential facility" concept discussed earlier – the primary incentive of each competitor has been to free-ride on the expectation of quality service provided by their competitors and Government guarantees like best execution, while cutting their own costs. In keeping with the notion of a race to the bottom in service quality that we have seen in other deregulated industries, it is often alleged that brokers will sell their customers’ orders to the highest bidder, who – one can assume – is able to win bids by giving the customer the worst execution he can get away with; so much for best execution. As in airlines and telecom, it is difficult to know whom to blame when you get nailed by a practice that everyone knows is common in the industry. Since, with or without payment-for-order-flow, cost-cutting in this context can take the form of trading against your customer, it is not surprising that we have seen so many scandals in recent years, nor that customers are flocking to on-line trading venues where they bypass brokers altogether.


On Line Locusts

    The shift from markets composed of either floor-focused specialist auctions, or upstairs dealers on telephones, to a situation where all legal trading information is carried on screens viewed by millions of investors – has been a dream of academics and regulators for over 3 decades. It is the embodiment of many concepts underlying NMS, such as the Consolidated Limit Order Book, or CLOB. And, although there are certainly more shoes to drop, such as decimalization, the fact is that the current market structure is already pretty close to the ideal one envisioned by so many supporters of the National Market System. We have automation, transparency, competition, equal access, narrow spreads, low costs for individuals, and many other aspects of the market we wished for.

    But there are problems, too, none more insistent than the fact that these changes – and, in particular, the shift to online trading – have been accompanied by rising volatility. I have written of this phenomenon before (in the Auctions Countdowns referenced earlier), and so will focus here only on the connection between this volatility and the breakup of the stock market as a network: The specialist auctions and the upstairs dealer market were unified information and trading loops that could carry infinitely more information than could ever be contained on a screen, and, far more importantly, they could effectively coordinate the application of that information to liquidity and price discovery. While it may appear that electronically connecting millions of individuals to trading screens is a consolidating act, it is, in fact, just the opposite. It separates them all so that there are no means by which the order-flow import of trading information can be brought to bear on liquidity provision or capital commitment by dealers. Consequently, the new regime satisfies only those who have no liquidity needs larger than the few hundred shares available at the narrow (and narrowing) spread. Everyone else is out of luck, and must deal with increasing impact and volatility.

    This is why security analysis is shifting to flow analysis: who cares what a stock is worth, if its price is determined by day-trader flows? If the old dealer capital networks adjudicating supply and demand are gone, and new ones, like call markets, are tangled in red tape, there is little to keep prices from going anywhere at anytime, driven by sudden flows coming out of chat rooms, false information distributed on the Internet, or even by old fashioned "earnings management." Just like a pyramid scheme, you can keep your stock rising, if only you can keep the flows coming. Stock splits, putting ".com" in your name, shrinking your float to maximize illiquidity and, therefore, the effect of flows – all of these tools are used regularly today, demonstrating that we are, indeed, in a New Era. Recent Wall Street Journal coverage has pointed out that the worse your earnings are, the better you seem to fare in the market. Having earnings, or merging with a company with real earnings and a history that can be valued rationally, have both been shown to hurt you. Why bother trying to determine whether p/e ratios are too high, or big stocks are too expensive relative to small ones, or value stocks too cheap relative to growth stocks. Why use any of the old Graham and Dodd methods, when the most successful predictor of stock returns over the last 5 years was Laszlo Birinyi, who called it right over and over just by looking at flows? With due admiration for Mr. Birinyi (and he now has many other admirers, as well as imitators) there is something wrong with a market that cannot adjudicate value. At the very least it is highly vulnerable to scams, as the PairGain joke made clear. If there is no means to manifest true supply and demand, or an aggregate opinion of value, hasn’t the market really become just a casino?

    There is an interesting positive feedback loop that is driving trade size down: dealers without the old information networks (because they are not allowed to talk to each other anymore) are not as able to read the flows that would allow them to commit capital to size bids, so institutions are breaking their trades into smaller pieces, and spreading their trading out in order to get close to VWAP (the volume weighted average price), so reading the flows is getting even harder for dealers, and they are even less willing to commit capital to make the size bids. Throw in online trading by swarms of individuals driving price so that it becomes even harder for dealers to read the flows, and it becomes clear that dealers cannot make the size markets anymore, nor – even if they could – can institutions afford the risk of any block trades looking bad against VWAP. So trading must break up into little pieces, and, the more it does, the more it is likely to happen further. In another particularly scary dimension of this loop, nobody cares about price anymore, because it simply doesn’t matter where any one of your hundreds of 200 share prints are priced, as long as they average out close to VWAP. Neither do the day-traders, with their 30 second holding periods – and they positively love the volatility. Frequently, as recent chat room transcripts that have hit the mainstream media demonstrate, they even deliberately cause it. The greater the volatility, of course, the more risky it is for a dealer to commit capital to making a size market, and the more a trader risks missing VWAP unless he trades all the time in tiny pieces – and we’re back again to the loop at the top of the paragraph. In short, volatility drives trade size down, and declining trade size causes volatility.

    And here is another interesting positive feedback loop: With day-trading in Internet stocks splintering flow into smaller and smaller pieces, institutions, who would need to do more than the 400 share size of an average "block" in one of them, are frightened away. They are also frightened by their wild volatility, lack of history, lack of earnings and lack of visible value – but this doesn’t seem to have hurt the Internets’ market caps. Ironically, the lack of institutional ownership may be one of the reasons for their strength. Because institutions don’t own them yet in anything approaching market weightings, they are effectively short, and will need to buy the Internets – like it or not – when they go into the benchmarks. So, the more valueless the company and the more difficult its trading situation, the more institutions will shun them. But, the more institutions shun them, the more their short situation will cause the Internets to rise – and their illiquidity, volatility and valuelessness to get worse.

    Imagine a whole industry of brand new Bre-Xs suddenly sustaining market caps that force their inclusion into the S&P 500. Even if all current Internet stocks are worth their prices, there are dozens or hundreds of new ones waiting in the wings. With the trading situation favoring those that don’t have any recognizable value, much valueless capitalization could be drawn into the indexes. And, as AOL proved, these are not ordinary adds. AOL came in as the 40th largest S&P 500 stock, an event on the last day of last year that seems to have marked the end of the index’s long run beating the other averages. But that could be just a warm-up, as these newer .coms are getting off to much faster starts. On their recent trajectories, many of them could be among the largest in the index when added. This short squeeze eventually could decimate the value of traditional stocks, as they are sold to make room for those that appear to have no value – and probably don’t. A handful of them, of course, may eventually have huge value. But there is no way at this time to tell even who is in the running, because value is on holiday. The street name for this – I mean the real street – is "shell game."

    The dramatic decline in average trade size as a percent of the total is, in my view, a pretty good measure of the disintegration of liquidity and price discovery that is giving rise to these positive feedback loops and valuation distortions. Even if the absolute dollar value of trades were increasing (which it is not; in fact it declined substantially between the 3rd Quarters of ’97 and ‘98, according to ITG) the fact that each trade is becoming less and less representative of the pricing and liquidity situation tells me that the market is becoming more and more unstable. Table 1 shows the percentage decline in average trade size as a percent of total daily volume in four stocks – GE, Intel, Microsoft and – between 1994 and 1998. Interestingly, a similar decline occurred in a spot check of every first Wednesday in September for four popular stocks between 1925 and 1929 – RCA, Bethlehem, GM and Sears (Table 2).


Table 1: Trade Size Changes in Late ‘Nineties

Year From Prior Year From 5 Years Previous

1995 -51%

1996 -8%

1997 -41%

1998 -31% -82%


Table 2: Trade Size Changes in Late ‘Twenties

Year From Prior Year From 5 Years Previous

1926 -17%

1927 -42%

1928 -48%

1929 -33% -83%


Tables show percentage changes in average trade size as a percent of total daily volume in each year from the previous year, and the cumulative 5 year change for the latest year. ‘Twenties data are for RCA, Bethlehem Steel, General Motors and Sears; thanks to NYSE archivist Steven Wheeler and NYSE visiting academic Marc Lipson. Nineties data are for General Electric, Microsoft, Intel and; thanks to Robert A. Wood, Distinguished Professor of Finance, University of Memphis Fogelman College of Business. All raw data on number of daily trades was converted by me to percents of total volume for tables; any errors are mine.


    Whether there is any significance in these similar declines or not, sustaining vigorous capital formation in the presence of unstable markets is difficult, to say the least. As Professor Robert Haugen pointed out in Beast On Wall Street, little capital was raised during the 4 years following the Crash of ’29, which were characterized by wild and persistent stock market volatility. Haugen disputes the conventional wisdom that the Crash was somehow forecasting those difficulties and the Depression. His view puts causality the other way around: it was persistent volatility engendered by a stock market reacting to its own price swings, rather than to any new information or analysis, that caused both the Crash and – by drying up capital formation – the Depression.