All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World-Wide Web.
He's merchant-bank Bertie, who sets off at 6:30 . . . [crowding], with thousands of others, into a shabby train or bus headed for the City of London. Then they all swarm through damp, grimy streets to their skyscrapers. To do what? To cluster in some of the world's most expensive office space, to stare at flickering screens and pore over pages of numbers, and to talk on the telephone to other disembodied voices about this deal and that price. Bertie's business has its ups and downs, though it is mostly a good one. But why does he do it in London?--The Economist, 1991
[¶3.] Why in London indeed? Possibly because of the myth of the City as the place where transactions take place, and the idea that markets are made real and physical rather than enacted by his embodied presence. Before 1986, stocks in London traded on the floor of the London Stock Exchange, in a controlled and structured chaos of shouting, shoving, sweating, and dealing, as they had since the nineteenth century. But in October 1986, electronic trading came to the City of London with a "Big Bang." After several months of soul-searching and critical examination, operations on the floor were converted to electronic transactions.1 The market became the International Stock Exchange, and the exchange floor itself ceased to be the focus, replaced as the real locus of transactions by glowing squiggles on thousands of computer screens.
[¶4.] What happened in London has been repeated, before and since, in other markets and on other trading floors, as computerized trading and computerized databases replaced the traditional methods of trading and settling accounts. Why pay the high rents of traditional business districts, or put up with the awful winter weather of London or New York, when one can set up a computer and a data link and trade from Bermuda, Belize, or the Bahamas--or Spain or the south of France (if one wants to keep European hours)?
[¶5.] In the process, the entire range of activities connected with financial markets, from simple interbank transfers to the creation of entirely new instruments that represent possible movements of indices rather than tangible worth, became delocalized. Although the process is far from complete, most of the trading has now shifted from such physical institutions as banks and market floors to the interactive networks originally built simply to interconnect them. No other major human activity has moved so quickly to the edge of "cyberspace."2
[¶6.] Becoming expert in market trading and analysis was once a craft skill, acquired by apprenticeship and experience. Whether trading in stocks, in commodities, in futures, or in money itself, the ability to anticipate market movements and trends was the substance and hallmark of success. But the power and influence of the old traders, and of their institutions, is waning. As more and more of the action takes place over the newly global computerized networks, both the influence of the traditional, hierarchically ordered firms and the power of the regulatory structures designed around them are diffusing horizontally across the net.
[¶7.] It may be acceptable to allow the City and Wall Street to continue as social rather than functional centers, while the real players sit at their remote computers, improving their models and machines in search of a more elaborate model or an improvement in response time that will give them an edge on their competitors in the net.3 It will probably not be acceptable to assign the same course of evolution to either private or public regulators of market fairness and behavior. But even if it were, that would require monitoring and control of activities on the net, without hierarchical authority, by a new class of players sufficiently equipped and informed to allow them to enter on equal footing.
[¶9.] To those who trade in bonds, securities, and other financial instruments, rapid and accurate information has always had tangible value. New technologies for information and communication have been not only welcomed, but encouraged.4 The telegraph, first demonstrated in 1844, was already being used by 1846 to make markets and to speculate in stocks. Price quotations from London were being published in New York papers only four days after the transatlantic cable was completed. In the 1860s, the exchanges developed a specialized information technology of their own, the stock ticker, and raced to upgrade speed and capacity. By 1880, there were over a thousand tickers in New York, and the New York Stock Exchange (NYSE), in the first example of regulation of new information technologies, acted to consolidate and integrate the distribution of information.
[¶10.] Wall Street was the first major consumer of the telephone and the electric light. The U.S. financial industry was the first major sector of the economy (after the telephone system) to make widespread use of electronic data processing and related innovations such as computer-readable cards and optical scanners; the now-familiar technology for magnetic ink character recognition was introduced as far back as 1958. Banks were among the pioneers in business use of the new capabilities of mainframe computers; by 1980, almost every bank had at least one computer, accounting for half of their capital expenditures other than for buildings.5 It is therefore not surprising that financial institutions moved more quickly than most to make use of the new techniques for marrying information processing and analysis to the management and manipulation of large databases.
[¶11.] The need to make efficient use of the new information capabilities while still conforming to the understood rules for trading and for transaction settlement led to concentration and hierarchy. Activities were increasingly dominated by a few large and powerful firms, concentrated geographically, along with their less powerful cohorts, in relatively few locations such as Wall Street and the City. For the most part, the hierarchical distribution of power and influence among the financial centers, among the increasingly international firms, and between markets, firms, and national regulatory authorities was stabilized in the 1930s; although the distribution of power and wealth underwent radical changes through the Second World War and its aftermath, the structure itself changed little for fifty years.
[¶12.] Centralization was encouraged by structure and by intent. Information flows outside of financial capitals were too limited and often too slow to allow for much interactive trading, especially across national borders. Even when advances in communication technologies allowed for national integration, the information flowed in and out of the exchanges, to be acted upon by traders on the floor. And with the memory of the market collapse of the late 1920s constantly refreshed by reexamination and re-analysis in schools of business and economics, governments sought to keep markets nationalized and localized, so that regulators could ensure honesty, control economic fluctuations, and avoid overseas flights of capital and ownership.
[¶13.] From the 1950s to the present, the structure and behavior of markets were relatively stable, needing little more than regulatory trimming to encourage wider participation, particularly by small investors. But the recent rate of technological change has begun to create visible signs of upheaval.6 As traders moved from mainframes and dumb terminals to faster and smarter mainframes and individual computers, increasingly rapid access was required to different sources and types of information. The direct effect was to make the market more active, and more volatile. The indirect one was to further increase trading in secondary instruments, such as futures, and to encourage the development of a host of new financial instruments that represented bets on the future movement of markets rather than the markets themselves.
[¶14.] By the mid-1980s, multiplexing and video switching made possible both consolidation of physical equipment and expansion of access to other sources. Satellite transmissions and digital data feeds made information accessible worldwide, increasing both complexity and the pressure to respond quickly. More powerful desktop computers fostered the development of sophisticated programs that further encouraged the expansion and activity of trading on secondary markets, and in derivative instruments. And, as will be discussed more fully in the following chapter, the eventual linkage within and between markets has created a global net of interactive trading capable of operating twenty-four hours a day, seven days a week.
[¶16.] What markets need to operate well and fairly are liquidity and efficient payment mechanisms.7 Both of these have always required accurate, up-to-date information, at first to make sure that the posted value of stocks, bonds, and currencies was a good measure of the current performance (and presumably value) of the underlying industry or enterprise. In addition, getting more accurate or timely information has a special value to those seeking to play markets for short-term profit rather than invest in them.8 Given the technologies available before the 1980s, these requirements tended to favor increased centralization, increasing the power and importance of the traditional exchanges.
[¶17.] The first collective approach was the "call" system, in which small numbers of traders gathered (often behind closed doors) to hold auctions. Unlike the system that preceded it, in which trading occurred informally and privately in bars and coffee houses, the call system encouraged market prices set by supply and demand. However, the call room was usually closed to outsiders. In Sydney, where the call system persisted into the 1960s, traders were locked in the call room, with no access to phone or telegraph, and all forms of the media excluded, until the call was over and prices established.
[¶18.] Closed systems provided few opportunities for small investors, and were too vulnerable to manipulation to encourage the entry of those not well socialized into the circle of traders. In an effort to open the system to the growing middle class, in 1871, the NYSE adopted the present "post" system, in which trading takes place on an open floor, and almost every other major exchange soon followed. Listed securities or futures are each assigned a specific floor area, or post, and buying and selling take place in auctions held at the posts on the spot. Trading is continuous and parallel, and traders have ready access to information. At first these were provided by telephones and telegraphs on the floor, and from public services such as stock market tickers, radio, and television. More recently, these have been augmented by computers and display terminals, providing access to a wide range of global information, to colleagues, and, in many cases, to the home office.
[¶19.] The manifest openness of the post system encouraged even relatively small investors to enter markets, greatly increasing the volume of trade and the pool of available investment capital. Moreover, the increased centralization that resulted provided more efficient and timely mechanisms for transaction settlements. When it became clear that a manifest demonstration of fair and equitable treatment encouraged further participation as well as being in the best interests of the majority of traders, centralized trading in concentrated financial centers also afforded a convenient means and venue for the expansion of government regulation.
[¶20.] The increased concentration of activity also created new opportunities for indirect manipulation of markets, through control of information flows or the misuse of inside information. After several identified abuses, preventing the exploitation of markets by insiders possessed of knowledge not yet made public became another motivation for regulatory intervention. Regulations were subsequently put in place to protect investors from false or misleading information and from the issuance of paper that had no underlying real value. Over the years, the relationship between markets and regulators became a symbiotic one, fostered and nurtured by the concentration of major activity in a few large sites, and by the desire of large, centralized firms to avoid public reprimand or punishment.
[¶21.] Although the system clearly was still subject to abuse (as in the recent junk bond scandal), the centralized exchange system, centrally regulated and centrally controlled, remained stable for more than fifty years, from the Great Depression through the 1980s. As volume grew, so did the ability to provide and analyze information. And as the techniques for managing and processing information became more powerful and rapid, that in turn spurred more frequent trading, further increasing the volume.
[¶23.] As access to up-to-date information broadened in the 1960s, large, institutional funds began to acquire their own analysts and in-house expertise, and to explore the notion of buying and selling directly rather than acting through (and paying high fees to) brokerage houses. But the brokers still had a monopoly on trading on the exchange floor, and the funds needed a substitute for the liquidity the exchange normally provided.9 In 1969, seeing an opportunity to broaden its information services, Reuters created Instinet, an institutional network that allowed subscribers to make and accept offers electronically, and directly.
[¶24.] As its computers grew more capable, Instinet further broadened its growing services, providing at first a computerized system for matching buy and sell orders and finally, as exchanges themselves became computerized, a mechanism for placing and executing direct orders electronically, bypassing the brokers completely.10 By the 1980s, electronic trading, at first offered only to institutional investors through vendors such as Reuters, was transformed by desktop computers and improved data systems into an electronic trading system that provided information and liquidity to anyone with access--and access itself broadened considerably as market deregulation came into vogue.
[¶25.] The final step toward breaking down the former hierarchical control of markets and trading came about in the 1990s as a combination of regulatory and technological innovation.11 The deregulation of telecommunications removed the blocks once placed on the entry of telecommunications firms into the computer arena, and computer firms into telecommunications, just when powerful desktop computers connected to local, national, and global networks were becoming standard desk accessories in the financial world. The result was a functional merger of communications and data processing. The "data terminals" originally put on desks to display and process information could now manage and settle transactions as well.
[¶26.] Once electronic trading removed the monopoly of financial centers on information, liquidity, and settlement, and provided access to expertise and information dispersed across the network, the hegemony of the centralized institutions was broken. The electronic program trader staring intently into a computer screen is increasingly as important a market actor as the floor trader or the traditional expert with a line to a broker. The stock ticker spewing forth its yards of yellow tape is now a recorder of trading history rather than a guide to investment strategy.
[¶27.] It has also replaced the old set of problems and potential abuses with an entirely new set. Near-instantaneous trading has spawned a new breed of "paper entrepreneurs" who seek profit through exploiting small market shifts in currencies or interest rates, or through other methods of manipulation of corporate paper, reacting to such stimuli as changing political conditions or immediate news, or gossip, with little or no concern for the production of goods and services.12 And nearly every firm engages in program trading or other forms of automated stock-index arbitrage, seeking to generate short-term returns from price discrepancies between markets.
[¶28.] Electronic trading has increased both the rapidity and the volume of large market movements--both by providing automatic responses and by removing the physical limits of executing trades on actual exchange floors.13 Stability once depended on large investors with a vested interest in long-term performance and both the time and the resources to assure it. Ensuring market stability may now be moving beyond the scope of those who understand, or care about, the tangible part of economic activity, or the companies, industries, and institutions that constitute it.14 Even in the market crash of 1987, large investors were able to stem the tide only with the help of massive intervention by regulators.
[¶29.] Traditional modes of regulation, both private and public, were designed to prevent gross manipulation by powerful people using their special accesses or special privileges to cheat, disinform, or mis-inform their peers and the investing public. They are likely to be far less effective against networks of program traders, freed of market floors, who can make nearly simultaneous moves in huge baskets, rather than as individual entities. The resulting increase in volatility would further discourage small investors, who already find it nearly impossible to enter the market effectively without also buying in to the new technology.15
[¶30.] Computers were introduced into centralized markets and exchanges to make them better informed and more efficient. The unanticipated secondary effect has been to render them increasingly obsolete. Control of markets is increasingly passing out of the hands of large institutions and dispersing into the network of computers and data links that make up an ever larger share of the trading volume. But the decentralization of markets, the dispersal of activity and deconstruction of centralized power, has not worked to the advantage of the general public. Instead, it threatens to remove from effective control both the mechanisms for market access and the means to regulate market behavior.
[¶32.] When the "Big Bang" came to London in 1986, The Economist reflected on some of the implications and concluded that some of the resulting changes (most notably the sharp increase in index arbitrage) were in fact inevitable, now that technology allowed the kind of quick movements needed to make profits out of small, transient movements. Its conclusions, however, merit quoting:
The fact is that none of the regulators knows exactly what the new techniques are doing to the markets. It looks likely that America's Security and Exchange Commission will soon mount a formal inquiry into the causes of September's mini-crash, which some have blamed on computer trading. Because the SEC has the power to demand documents from brokers, it may find out who did (and does) what and why. But for a fair verdict, wait to see how Wall Street's computerized investors handle a bear market.16
[¶34.]
[¶36.] In August 1987, the Dow Jones average had risen to over 2700. A gentle decline that continued into the fall raised no alarm and little concern. Then, without warning, the Dow dropped by more than 95 points on Wednesday, October 14. On Thursday it declined by another 58. On Friday, the market was down by 60 points by 3:00 p.m., but still orderly. At that point, the bottom fell out. Swamped with sell orders, many from automatic trading programs, the Dow dropped another 48 points in the last hour.
[¶37.] The 108-point decline for the day (and 235 points for the week) made headline news. But far more important was the perception by traders that the market had become disorderly. Had the closing bell not intervened, program and automatic trading would probably have been joined by other, slower to respond, investors seeking to cut their losses, taking it down further. Other markets, however, were still active, and the response continued.
[¶38.] By late Sunday, markets in the Far East were already down sharply; London was down by 10 percent before the market even opened in New York. At the opening bell, NYSE's "state of the art" mainframe computer was already overloaded. Mutual funds began to try to dump stock to get out from underneath. And every program trader and automatic trading program sent only one message to the NYSE: Sell!
[¶39.] The result was chaos. The NYSE dropped by over 100 points in the first hour, and another 104 in the next thirty minutes. With electronic trading systems active, and transaction reports lagging, the market staged a series of rapid and seemingly unmotivated rallies and falls that blocked attempts by more seasoned traders to act.20 This erratic behavior broke the confidence of many large traders and index arbitragers, who then opted to close down their high-tech systems and wait the market out. This removed the largest players from the market, and left the old-fashioned trading specialists on the floor of the exchange as the sole buyers struggling against the tide of automated selling programs.21
[¶40.] What followed was a disaster. By mid-afternoon, portfolio and fund managers began to dump large quantities of stock again, and the market went into free fall. As each automated program read the increasing rate of decline, it triggered its own emergency dump program, effectively adding to the wave of sell orders. Under the pressure of the tremendous volume, even the automated trading systems began to break down.22 It became increasingly difficult to get accurate quotes, let alone to make transactions.23
[¶41.] By the closing bell, the Dow had declined a record 508 points to 1738. Moreover, market orderliness, fairness, and access had gone the way of stability. Individual buyers and small traders, seeking to step in and take advantage of the low quoted prices, could not get through. Neither could old-fashioned specialists seeking to stabilize the system, even large ones. Indeed, financial expert Felix Rohatyn thought the market came "within an hour" of total and complete disintegration.24
[¶42.] Tuesday morning, a combination of restrictions on program trading, and the withdrawal from the market by specialists trading in specific players, reimposed some semblance of stability. Around mid-day, the old-fashioned experts staged a bald-faced market manipulation. According to the Wall Street Journal, a "small number of sophisticated buyers," using very little cash, engineered a rise in the Major Market Index futures.25 Automated trading programs read this as a clear signal to buy, and slowly but surely buyers returned. By mid-afternoon, in what may well have been another deliberate manipulation, they were joined by major corporations offering stock buybacks.26 Over the next few days, program traders, faced with dire threats of severe regulatory interference, monitored their electronic trading more carefully, while large buyers continued to intervene selectively to stem prospective runs. Faith in the market was gradually restored, and with it the myth that prices were based on the underlying economy.
[¶43.] In a TV show analyzing the crash, many expert analysts blamed the exaggerated response largely on computerized trading programs, which kicked in at the first steep decline, turning what would otherwise have been only an unusually bad day into an automated panic, driven entirely by electronic transactions.27 "The fundamentals of the market changed," said one analyst, "We became part of a computer instead of part of a system, and the liquidity in our marketplace was taken out of the hands of the professionals and run through a computer."
[¶44.] Neoclassical markets are supposed to measure the performance of the economic activities whose representative paper they value and trade. In the weeks that followed, regulators and major traders moved to restrengthen that legitimating myth. Supposedly adequate restrictions and restraints were put on program trading, particularly during large market moves--treating the symptoms rather than the disease. Public confidence was thereby restored. Over the next few years, the market moved generally up, eventually recapturing most of what had been lost during those frantic days. But in the new era of electronic networking and program trading, the decoupling between markets and real-world activities remain; markets have remained more volatile, and less predictable, since.28
[¶46.] Finding itself at a loss to respond to public demands for better protection, Congress requested from the Office of Technology Assessment (OTA) a report on the growing role new communications and information technologies play in the U.S. securities market. The OTA report, whose title I have shamelessly borrowed for this section, set out in some detail the pertinent market mechanisms and methods, including options, futures, settlements, and market clearing, dealing with twenty-four hour, global trading in a separate report.29 OTA found U.S. markets to be basically healthy, liquid, efficient, and fair.30 But they acknowledged that the stress being caused by technological change not only affected market structures and procedures, but put a major strain on existing methods for regulation and oversight.
[¶47.] The direct impacts of technical change were clear. But they could not be easily separated from other, less direct factors such as the changing role of large investors.31 These technological factors, arising from often indirect and seemingly contingent interactions between the social and technical dimensions of markets, are changing both rules and structures.32 OTA suggested that this may be the most lasting effect of the integration of computers and information systems, but was unable to predict what the long-term outcomes might be.33
[¶48.] OTA made three main observations. The emergence of an interactive global economy makes it increasingly difficult to limit trading to those whose actions may be constrained by their involvement in the national economies that markets are said to represent. New modes of trading are fostering the development of global institutional investors with enormous resources and open access to national markets and exchanges. The changes that are made possible by the computerized means of interaction are seemingly radical, but still far from complete.
[¶49.] The regulatory situation was found to be far behind the pace of change. The two primary agents of federal regulation, the Securities and Exchange Commission (SEC), established in 1934 to regulate the primary exchanges, and the Commodity Futures Trading Commission (CFTC), established in 1974 to regulate commodity and futures markets, operate separately, with different goals, different means, and different objectives, even though securities, futures, and options markets grow increasingly interdependent because of the new technologies that link them.34 Neither has either the jurisdiction or the authority to deal with trading across national borders, and neither is able to keep up with the rapid, nearly automatic trades made by computers over networks.
[¶50.] When program trading began, the major concern was that rapid
[¶51.] In October 1987, experienced traders were able to intervene to stabilize the market when regulators could not. But their power, too, is waning. As skill with the computer and its models becomes more important than accumulated experience, power and influence are shifting to a generation of younger traders more familiar with electronics than trends, less concerned with preserving long-term market stability than with developing and mastering better, quicker models with which to outmaneuver the competition.
[¶52.] The increased rapidity of market action and reaction also requires constant attention, which will not only disadvantage the older generation of insiders but gradually shut out other traditional investors who still have to act through brokers instead of dealing directly with the electronic networks. It also presents enormous problems to regulators seeking to guarantee that the improvements in information and liquidity are not compromising the other central value of fairness that characterized the historical exchanges.
[¶53.] The only response thus far has been to seek ways to monitor the relation between information and communications networks and actual trading activities. Some markets have already moved to incorporate limited electronic after-hours trading systems rather than leaving themselves at the mercy of global traders working in the wee hours; even the traditional NYSE is beginning a phased program that will take it to full around-the-clock, global trading by the year 2000.36 But the SEC has no real authority over financial information vendors or others who threaten to move a large part of the trading volume outside of regulatory structures.37
[¶54.] A more effective response would be to recognize that what is emerging is a radical and not an incremental transition. Although current trends to decentralization and deregulation have resulted in major changes in many industries, and will probably affect many more, what is happening to electronic markets may be the first important example of a structural reformation of industrial societies that goes beyond the near-chaos of postmodernism to something just as radically different from traditional, large, powerful firms based on vertical integration of functions and skills, but more systematically organized. The vertical modes of organization that once governed markets and trading will not fragment into disorder, but rather will progressively devolve their powers, and their roles, to more horizontal modes of organization and functional differentiation in which authority, purpose, and skill are widely dispersed and integrated by computer networks rather than bureaucratic hierarchies.38
[¶56.] On Wednesday, July 19, 1995, technology stocks led American stock markets down, and then partially back up, in another of the dramatic roller-coaster rides that have marked the era of electronic trading.39 The Dow Jones average was down by over 130 points before recovering to close down 57 for the day. Trading volume was the third highest ever, more than 480 million shares; phone lines were jammed and home-computer trading systems were almost locked out completely.40 The invocation of program-trading circuit breakers and other mechanisms did manage to stabilize the Dow, and phone lines and switchboards expanded in the wake of the 1987 crash were for the most part able to handle the tremendous volume of calls. But the smaller Nasdaq Stock Market had a terrible, horrible, no good, very bad day indeed.41
[¶57.] Now the second-largest stock market in the United States after the NYSE, the newer Nasdaq has a higher proportion of the high-technology stocks whose trading drove the Wednesday activity. Lacking a traditional market floor, it also depends heavily on its new and powerful computerized system for handling trades. On Wednesday, July 19, the rocketing trade volume exceeded the computer's capacity, causing at first delays and later almost a complete halt in trading. Customers trying to shift to phone lines instead also overloaded the switchboard capacity of Nasdaq traders; many could not get through at all during the critical hours when the market bottomed out before starting its rise. It was not a repeat of October 1987, but it came uncomfortably close. It would seem that we are not yet quite finished learning all the lessons of the potential risks of computer trading.
[¶59.] Because markets trade in wealth rather than creating it, because what they deal in are symbols and promises rather than tangible goods, they have been able to exploit the new opportunities of computers, computerized databases and information systems, and computerized networks almost as quickly as the technical means have progressed. In the process, they have transformed the structure of the financial world in ways that were never anticipated. What started out as technical improvement and technical change quickly evolved into systemic, technological change, in the broadest sense, reconfiguring the institutions that developed and deployed them.
[¶60.] As computerized networks and powerful satellite communication links make markets accessible from almost anywhere on the globe, experienced traders with a vested interest in making and stabilizing markets are increasingly being displaced by a new breed of "paper entrepreneurs," computer wizards who depend on their computers rather than experience, seeking to make short-term profits by anticipating small market movements rather than long-term ones by creating a pattern of stable investments, driving the market in response to trading programs with little or no concern for the underlying economic activity.42
[¶61.] With access to rapid, accurate information and fast response times, the opportunity to make many small profits by quick trading to capitalize on time lags, small movements between and among markets, or rapid transactions in secondary instruments, is a tempting alternative to the traditional processes of seeking sound investments. Why invest large sums and wait for market movements in response to the comparatively glacial change in economic indicators when larger profits can be made more quickly by investing in derivatives, or speculating on future movements on margin?43
[¶62.] Thoroughly modern Bertie, sitting on a beach in Belize, may trade in relatively real things, such as securities or money, in a variety of new secondary instruments that exist only because the power of networked computers makes trading in them possible, or even seek to profit on minuscule differentials or time delays between markets separated by continents. He does not need to submit to traditional exchanges, nor to the trading restrictions imposed on national markets by exchanges or governments.
[¶63.] The computerized networks of information and communication established to allow powerful, centralized, vertically organized firms and banks to exert better control over financial markets are now transforming into networks of action, dispersing their power and decreasing their control. Over time, the large firms and banks may simply become convenient nodes for such physical inputs and outputs as are required. Traditional government regulation was designed and implemented to impose order on the system through its leverage on these concentrated nodes. As their importance decreases, so will regulatory effectiveness.
[¶64.] To adapt, regulators would have to accept that they can no longer regulate effectively from the traditional position of an outside authority, able to control a few large firms by rewarding good behavior and punishing bad. To do so in the newly dispersed markets would involve a degree of intrusiveness and control that would be neither effective nor affordable.44 The alternative would be to devise regulatory structures adapted to the new circumstances. Horizontally rather than vertically oriented, these regulators would be charged with entering into the networks as co-equals, using people and equipment as sophisticated as those of the traders to monitor behavior by participation and involvement.45 As with the case of the California Highway Patrol being visibly present on major highways over holiday weekends, their presence alone would have a considerable effect.
[¶65.] Unfortunately, no such entity is yet on the horizon. The coordination costs would be enormous, and so would the knowledge burden that would be imposed on regulators or other agents, who would have to become at least as proficient in the uses and misuses of the new technology as those whose behavior they are trying to identify and detect.46 Such self-organization and self-regulation would require at least tacit acceptance by traders of the validity and legitimacy of both the rules and their enforcers. But that in turn would require the formation and maintenance of precisely those kinds of social exchange relations that characterized the traditional, centralized markets the new computerized trading systems are systematically deconstructing.
NOTES:
1 "Anyone Know a Cure for Hiccoughs?"
2 Compare, for example, to Hans Moravec, "Pigs in Cyberspace." The term first gained wide currency in the cyberpunk science-fiction literature, e.g., Gibson, Neuromancer.
3 See, for example, "The Screen Is the Future, Master."
4 See, for example, Garbade and Silber, "Technology."
5 Landauer, Trouble with Computers, 33. As an interesting sidenote, Landauer also cites Franke ("Technological Revolution") as showing that bank productivity stagnated during the 1970s, when computerization was taking place rapidly for the ostensible purpose of increasing it.
6 U.S. Congress, Office of Technology Assessment, Electronic Bulls and Bears.
7 Liquidity means that exchanges should provide a means for every seller to find a buyer, and vice-versa. The larger and more varied the trading volume on an exchange, the more likely that such transactions could be completed. But for markets to be efficient, transactions should be rapid. Here also, increased capacity and sophistication of constitutive technologies (even such simple ones as the telephone) helped the exchanges to grow without delays or bottlenecks that would impede liquidity.
8 This is, of course, especially true for secondary markets that trade in futures or other paper rather than issued securities or bonds.
9 Salsbury, "Emerging Global Systems."
10 Although Salsbury points out that this was suspended following the October 1987 debacle because of the risk it posed to market makers in a rapidly falling market, Reuters and others are still pursuing the idea.
11 Meyer and Starbuck, "Interactions Between Ideologies."
12 Harvey, Postmodernity, 163. The growth of paper entrepreneurialism had already been noted in the early 1980s: See, e.g., Reich, Next American Frontier.
13 Some Wall Street analysts maintain that electronic arbitrage is actually to be preferred. Writing in Science, M. Mitchell Waldrop quotes trader Michael Alex on the counter-argument: "Think what it means when you execute a program trade without using the computer system. The broker has to take the orders for (500) different stocks and give it to all these runners. The runners go screaming out onto the floor, pushing up to the specialists' desks and shouting. That alone can create panic" (Waldrop, "Computers Amplify Black Monday").
14 Remarkably, and despite the concern shown by technology assessors and some market analysts, the computer science and other technical communities seem to have overlooked these issues almost completely. Many recent books and articles surveying the role of computers in the future "information society" fail to mention security, bond, or currency markets at all. See, for example, Weinberg, Computers in the Information Society.
15 Computer trading programs based on historical trend analysis tend to converge, and are therefore likely to behave similarly given similar inputs. Market responses are therefore likely to be exaggeratedly large, perhaps dangerously so, despite regulatory attempts to suspend program trading on large movements.
16 "Is Your Stockbroker User-Friendly?"
17 Crudele, "Volatility"; Sloan, "Why Stocks Fell."
18 Bennett, "More Wild Stock Swings Expected."
19 A good deal of this chronology is adapted from Rosen, "Crashing in '87." Also see OTA, Bulls and Bears.
20 Rosen, "Crashing in '87."
21 A "specialist" or market-maker is an exchange member who takes special responsibility for providing liquidity and smoothing transactions for one or a few specific stocks in exchange for a unique and profitable role as dealer. They not only find buyers and sellers, but use their own capital to absorb shares to ensure a smoothly functioning market. A growing concern is that specialists are not sufficiently capitalized to stabilize against the large blocks being moved by the trading programs of large institutional investors, as was nearly the case during the Black Monday crash of October 19, 1987. See, for example, OTA, Bulls and Bears, chapter 3.
22 For example, Wells Fargo alone acting on behalf of a giant pension fund fed in thirteen separate bundles of more than $100 million each. By the end of the day, the $1.4 billion represented more than 5 percent of the total trading. "What Caused the Meltdown?"
23 See, for example, French, "Efficiency of Computer Systems." Ironically French's firm survived October 19 very well by shutting down its computers and trading almost exclusively from the floor, using its floor-trained, experienced personnel. There is a lesson in there somewhere about the displacement of such people by computer systems in the future, but few, if any, of the market firms are paying attention to it. Also see OTA, Bulls and Bears, chapter 3.
24 Rosen, "Crashing in '87."
25 "Terrible Tuesday."
26 Rosen, "Crashing in '87."
27 For an analysis of the role of arbitrage on October 19, see, e.g., "What Caused the Meltdown." A similar fluctuation occurred in the spring of 1988 (de Maria, "Dow Up Just 1.99 in Wild Day").
28 Historically, market pundits have never lacked for causal explanations of market activity. It is therefore quite remarkable how common, and readily accepted, it now is for news reports of the day's market action to attribute the moves solely to program trading--accepting without question the notion that computerized trading is now as much determinative as responsive.
29 U.S. Congress, Office of Technology Assessment, Trading Around the Clock.
30 This somewhat Pollyanna conclusion is not uncommon among market traders and analysts, who see the new technologies as adding to their capabilities, efficiency, and profits. The same set of beliefs is expressed by the many authors collected in Lucas and Schwartz, as well as repeatedly in journals such as The Economist and the Wall Street Journal. The underlying myths are powerful and enduring.
31 Indeed, it is not even clear whether the link between technological change and increased dominance by large and powerful parties is indirect rather than direct. Many critics of capitalist economic systems go even further and argue that technological change is a means by which large and powerful actors directly seek to increase their dominance under the guise of sharing authority or power with the less wealthy and less advantaged. See, e.g., Noble, Forces of Production; Harvey, Postmodernity; Noble, America by Design; Jameson, Postmodernism.
32 In formal terms, this closely resembles the dynamic models of structuration developed by Giddens (Central Problems in Social Theory; "Structuralism"). For other applications of similar ideas to socio-technical systems, see, for examples, Thomas, What Machines Can't Do; Barley, "Technology as an Occasion for Structuring"; Barley, "Alignment of Technology."
33 This is hardly surprising, since the predictive tools of policy analysis are rarely able to deal with extensive social transformations, particularly those in which technical and social changes interact strongly.
34 The SEC regulates trading in securities and assets, and tends to be more cautious in dealing with innovation than the CFTC, which, as regulator of contracts generally used for hedging and speculation, has tended to seek to be flexible and responsive to new approaches and ideas.
35 Crudele, "Market Being Manipulated." The opening of the Chicago futures exchange to electronic trading, for example, allowed a maneuver by which a large investor could rapidly buy large quantities of futures, hoping thereby to trigger the programs of other traders to buy the underlying securities on the NYSE. When the stock rises, the manipulator can quickly sell his futures, getting out with a profit before the NYSE can respond.
36 OTA, Bulls and Bears, 13, and chapter 3.
37 Ibid., 12.
38 Barley, "New World of Work."
39 Although it was not until May 8, 1996, in the middle of what several traders described as the most volatile week in memory, that the NYSE was to invoke both the uptick and the downtick limitations on computer trading in the same day.
40 Louis, "Heavy Stock Trading Jams Broker's Lines."
41 The full story of Nasdaq on that July day very much resembled the trials and tribulations set out by Viorst and Cruz in Alexander and the Terrible, Horrible, No Good, Very Bad Day. I thank Alexander Rochlin for having brought this to my attention.
42 Harvey, Postmodernity, 163; Reich, Next American Frontier.
43 Secondary instruments such as futures or "derivatives" such as option indices are essentially bets on the future performance of interest rates, or the indices of exchanges dealing in "real" goods, securities, bonds, or other financial instruments. Although derivative markets are by definition zero-sum (every seller must find a buyer, and anyone's loss is someone else's gain), they have attracted considerable concern because of their leveraging effect and their potential volatility. In principle, there is never any net gain or loss in a derivative market; in practice, individual investors or firms can lose large amounts of money very quickly.
44 Which does not seem to have stopped Congress from at least considering attempts to regulate and control the accessibility and flow of pornography over the Internet, or the use of certain forms of data encryption.
45 Rochlin, "Trapped by the Web."
46 For an exemplary discussion of the knowledge burden, the costs to an organization of keeping itself informed and up to date on the activities it is trying to monitor or control, see Demchak, Military Organizations.
The Crash of 1987
Electronic Bulls and Bears
Nasdaq's Very Bad Day
Conclusion