Modern stock exchanges couldn’t exist without wires. They are virtual, global, infinitely expanding. Their trading floors are humming servers. But no one ever planned this transformation, and it took many by surprise. This episode explores the long processes of automation throughout the second half of the twentieth century. We hear about engineers, screens, and how technology created a new stock exchange almost by accident.
Let’s take a walk through a stock exchange. In the 1980s, it would have sounded like this…
—– trading pit —–
That’s a trading pit, with the bell sounding, bodies crammed together, pushing, shouting. We have heard it a few times by now. In the late 1980s, when Tom Wolfe visited the trading room of Pierce & Pierce, he still found a terrible noise, ‘an ungodly roar, like the roar of a mob…an oppressive space with a ferocious glare, writhing silhouettes…moving about in an agitated manner and sweating early in the morning and shouting, which created the roar. It was’, he writes, ‘the sound of well-educated young white men baying for money on the bond market.’ But the market is only partly in this trading room, it is outside, absent, on the screens. And if you walk through a stock exchange today, it would sound like this…
——– ‘singing servers’—–
Isn’t that eerie? The sound of servers in a data centre, chattering to one another. A beautiful recording, too. These changing sounds are the background to the story in today’s episode, that of automation, the transformation from spoken markets to those of near instantaneous speed, a transformation that has made possible an increase in the volume and scale of financial transactions to a level that would have been simply inconceivable 30 years ago. Economists delight in pointing out how technological improvements in financial markets lead to socially beneficial outcomes through facilitating liquidity and choice. That argument, however, supposes that changing the medium of trade has no consequences other than making it easier. By now, we know this cannot be the case: throughout the first part of this podcast we have seen how the shape, function and purpose of financial markets are every bit as dependent upon their material structures as on regulatory regimes and global political-economic conditions. Through the 1980s and 1990s, automation turned stock exchanges inside out. That is today’s story – even if we don’t make it all the way into the cloud in one episode.
Hello, and welcome to How to Build a Stock Exchange. My name is Philip Roscoe, and I teach and research at the University of St Andrews in Scotland. I am a sociologist interested in the world of finance and I want to build a stock exchange. Why? Because, when it comes to finance, what we have just isn’t good enough. To build something – to make something better – you need to understand how it works. Sometimes that means taking it to pieces, and that’s exactly what we’ll be doing in this podcast. I’ll be asking: what makes financial markets work? What is in a price, and why does it matter? How did finance become so important? And who invented unicorns?
The last two episodes have focused on the upheavals felt in the world of finance during the 1980s, the decade when greed became good. We saw, in episode six, how shifts in the tectonic plates of global economic governance and the intellectual fashions around ownership and collective versus individual responsibility had led to the birth of a new kind of social contract, the individualism of Thatcherism and Reaganomics. We saw how – in the UK at least – that manifested itself in a new kind of investor, Sid, the archetypal blue-collar worker turned property owner who bought into the newly privatised industries and could consider himself a member of the rentier classes. In episode seven I explored the new deals imagined by those working on the cutting edge of finance – the invention of elaborate investment bonds fashioned from home mortgage repayments, and the leveraged buyout beloved of corporate raiders and asset strippers. This was when you forced your target to borrow money to buy itself, tore it to pieces and sold them off to pay the debt, and kept yourself a handsome profit in the process. You justified your actions by claiming that you were returning value to oppressed and voiceless shareholders, whom managers had apparently been robbing for years. But none of this would have been possible without steady, mundane, and often barely noticeable changes in the technological infrastructures of the stock exchange.
Of course, these changes were not always invisible. Some came with a big bang, as on Monday 27 October, 1986, when London’s markets finally went electronic. You may recall that regulatory changes put in place with Big Bang saw the end of single capacity trading and the role of the jobber, the end of fixed commissions and the liberalisation of ownership rules. The fourth and final plank of the Big Bang reforms was the London Stock Exchange’s decision to replace spoken trading with a distributed, screen-based system called SEAQ – S –E-A-Q. Market-makers – who replaced jobbers and were able to deal for clients and on their own account – published ‘two way’ buy and sell prices over the screens.
The best prices for any security were highlighted by a yellow strip at the top of the screen and a broker who wished to deal would call the market maker on the telephone and strike a bargain. London had borrowed this distributed-trading model from NASDAQ: even the name showed a debt of gratitude, the Stock Exchange Automatic Quotation, echoing National Association of Securities Dealers Automatic Quotation. The new system looked so like the American over-the-counter market that the New York Stock Exchange put itself in a perilous political position by banning its members from trading on the London Stock Exchange, just as they were banned from NASDAQ. A week later New York retracted, a spokesman conceding that, ‘If the British Parliament says it is an Exchange, that’s good enough for the Big Board.’ (I should say, as always, that full references are available in the transcript that accompanies this podcast).
Those designing the new market had no particular wish to disrupt the old one. The system was built with continuity in mind and made it possible for people to trade on the Stock Exchange floor, just as they had always done. Many firms took leases to pitches on the new floor, refurbished and upgraded at the cost of several million pounds. But the jobbers knew that their world was changing. While the big firms were buying long-term leases, the jobbers knew that they would never set foot on the floor of the house again. On Friday 24 October, the last day of spoken trading, the floor of the house hosted a day of wild festivities. Jobbers chased a pantomime horse containing two clerks round the floor, and the Spitting Image puppet of Chancellor Nigel Lawson made an appearance. In all, says one historian, it was more a ‘rowdy Irish wake’ than the solemn, final day of a mighty institution. Managers, expecting business as usual, were caught out: ‘Within five minutes of Big Bang,’ says one, ‘on Monday morning, it was clear to me that the floor was dead. I’m not bragging. I was the last person in the City to figure it out.’ But there was no reason to loiter downstairs, struggling to elicit prices from a seething crowd of traders when one could survey the whole world of prices from the comfort of one’s desk. The crowds just moved to their offices upstairs, so promptly that, by mid-morning on Monday it was clear that the trading floor was finished. In January 1987 only a hundred people traded regularly on the floor – just a tenth of the crowd that had traded there a year previously – and the Financial Times was speculating about whether the new six-sided pitches might become a ‘Hexagonal Wine Bar’. The trading floor closed three months later.
—- keyboard and typewriter sounds, here and below —-
If technology merely improved informational efficiency, why was there such inevitability to the collapse of floor trading? And why couldn’t the banks and investment houses themselves see it coming? It was not just more comfortable to trade from one’s desk, but also safer. Traders were now obliged to trade at the prices offered on the screen, for these were ‘firm prices’. But if the telephone was ringing, it was always possible to check the screen before picking it up. In fact, one of the great complaints about screen trading was that during sudden market collapses – when lots of people simultaneously want to sell – dealers stopped picking up the phone. Traders could have more screens on their desk, bringing in all kinds of information from the outside world, and placing them at an advantage to others; office organisation could deliver the same benefits, with salespeople, analysts and other experts easy to reach. Moreover, everyone in the office knew the news first – the technology inverted the relationship between floor trader and clerk, between front office and back. And moving to screens did not mean abandoning all those social relationships that had sustained trade on the floor. Those young men in Peirce & Peirce’s trading room are shouting into telephones, making deals with others that they spoke to, as one trader wryly pointed out, more frequently than they spoke to their spouses. Telephones formed a useful bridge between the bodies of the floor and the disembodiment of screens. Under the SEAQ system, brokers still dealt by phone, or by direct lines connected to an intercom known as the box. These devices were crucially important in the operation of major stock markets in the late 1980s and the 1990s: ‘If you don’t have your brokers in the box, you are not in the market’, said one Parisian trader.
Mechanisation had become a preoccupation of stock exchange officials worldwide. This interest stemmed from the middle of the twentieth century. Often, it had egalitarian underpinnings: if mechanism could reduce manpower, wrote one author, ‘we might even reduce the costs to such an extent that small orders became profitable and the ideal of the Cloth Cap Investor at last became a reality.’ Fischer Black, the economist whose option pricing theory was to transform the financial world, had dreamed of a fully automated securities market. His pamphlet was illustrated with a line drawing of an enormous machine straight out of B-movie science fiction, the market machine drawn as a riveted dustbin on stilts with enormous tendrils, like vacuum cleaner tubes, reaching down onto the desks of bankers and traders. It is hard to read the expressions of those occupying the desks, but they certainly are not joyful. Thinking such as this was never entirely benevolent: it also had roots in the desire for effective supervision of market participants, whose dealings by handshake and conversation could be easily hidden. But we should be careful of reading the history of automation as a smooth transition from lumpy, inefficient bodies to sleek, efficient machines at the hands of strategically visionary management. Juan-Pablo Pardo Guerra, who has written extensively on the topic, asks why – bearing in mind the comfortable, profitable market positions held by senior players within the organisation – did automation happen at all? He argues that the process is haphazard and diffuse. It begins, inevitably, with the routine tasks of settlement and clearing; in London, the post war years saw mechanical calculating devices, and then computers, introduced to streamline what had been a labour intensive, time consuming process. Crucially, according to Pardo Guerra, these early machines allowed a new kind of participant, the technologist, into the closed world of the LSE. Calculators and computers demanded technical expertise, and the technologists who worked on them built their own quiet and often invisible networks of power within the organisation. The members of the exchange (the brokers and market-makers) were used to treating back-office workers as staff, secondary in status and in access. They treated the technologists the same way. Pardo Guerra passes on a story about a member meeting the Exchange’s new technical director – a senior appointment – in the lavatories of the sacred seventeenth floor, a space reserved for members, and expressing his displeasure about sharing the facilities with the staff. One can hardly blame the technologists for pushing changes through, until, one day the members woke up to find that they were not in charge any more.
The details of automation are complex, and are exhaustively covered in Pardo Guerra’s book. Change was incremental. In 1970 the London Stock Exchange introduced its Market Price Display Service to show middle prices on black-and-white television sets in offices throughout its newly constructed concrete tower block. The service was a manual-automatic hybrid that relied upon Exchange representatives patrolling the trading floor, physically collecting prices. The blue buttons were happy to delegate this work to them and began quoting prices verbally rather than chalking them up on a board. MPDS prices often differed from those made available by the Financial Times and Extel – rival data producers – so the Exchange banned these organisations from the trading floor, thus creating itself a monopoly in the new and lucrative commercial market for data. This early analogue computer, data carried in coaxial cables, was soon outdated. The LSE implemented a database called EPIC (The Exchange Price Information Computer) able to hold a limited amount of price information for every single stock traded. Then, in 1978, it launched a new system named TOPIC (or, less snappily, Teletext Output of Price Information by Computer) based on the Post Office’s proprietary teletext system, named Prestel. ‘TOPIC,’ writes Pardo Guerra, ‘was not simply a scoping device, a way of seeing the market: it was, rather, a common platform, a standardized mechanism for displaying market information – from prices and company announcements, to charts and tailored analytics – and reacting to it from afar.’ As Pardo-Guerra points out, the crucial advantage of this system was that data could flow both ways – from the trader’s terminal to the central hub and back. TOPIC made possible new modes of visualization and calculation. It was, in other words, creating a new market place: the screen. In the early 1980s the looming Big Bang provided the technologists with an opportunity to cement their grip on the organization of trades, and they set to work to render the sociality of the exchange into cables and screens, a utopian endeavour that simply never came to fruition. Forced to adopt a quick fix to meet the deadline, the Exchange hammered TOPIC and EPIC – its two existing systems together into a new combination, named SEAQ, which underpinned the change to dual capacity trading in October 1986.
So a series of incremental improvements, driven by political concerns, attempts to grab a bigger share of an emerging market for data provision, and the struggles between managers and technologists, eventually coalesce around a system that makes the trading room redundant. Nobody had expected this, and certainly no one had planned it. It caught many off guard. Those who had spent their careers on the floor of the house had learned to read bodies, not numbers. They did not really need to know the long term prospects for a company, how much its dividend might be or whether the bank was likely to foreclose. They simply needed to know who wanted to buy stock, and who wanted to sell; even better, to know who wanted to sell, and who had to. Bodies were enough for that. Eyes, sweat and movement, the look of tension on the junior’s face, these things told an experienced jobber everything they needed to know. Screens project a new kind of market. There are no people, no bodies: no scent of greed or fear, no recognition of friends or foes. The screen trader must make sense of strings of numbers, learning to read the market in an entirely different way. Screens make possible a global market, unrolling through an electronic network that circles the globe from bridgehead city to bridgehead city: Tokyo, Frankfurt, London, New York. Screens are devices that visualize and create the market; the sociologist Karin Knorr Cetina describes them as ‘scoping devices’, analogous to the instruments of a laboratory. Traders arriving at work, she writes, ‘strap themselves to their seats, figuratively speaking, they bring up their screens, and from then on their eyes will be glued to that screen, their visual regard captured by it even when they talk or shout to each other, and their body and the screen world melting together in what appears to be a total immersion in the action in which they are taking part.’ Making sense of this vast world of information means building new kinds of calculators, and prices tracing across screens are the perfect material for doing so. Traders’ tools are the graphs and spreadsheets of the Bloomberg terminal, with its endless, varied representations. At first, innovative computer programmers sought to recreate the bodily world of the trading floor. Programs simulated crowd noise, rising and falling in line with activity, but these were never successful. Other prompts and shortcuts grew to fill the space instead. In London, for example, the Exchange introduced the FTSE 100 ‘trigger page’. This showed the code for every single stock in the FTSE 100 on a single, teletext screen. A blue background to the code signified the share was moving up and a red that it was moving down. You no longer needed to hear the crowd to know how the market was faring; the information one needed was there, brightly coloured, on a single screen.
Screen-based markets make it possible to trade without any human help at all. In many ways, this was the dream of visionaries such as Fischer Black, using machines to cut costs and trim trading margins until a truly efficient, democratic market was achieved. According to a certain line of thinking, the proliferation of trades that machines bring creates liquidity and benefits all market participants. The jury is still very much undecided as to whether computerised trading leaves us better off – Michael Lewis’ Flash Boys argues passionately that it does not, and we’ll return to the topic in due course. But it is undeniable that computers react more quickly than people and without any sense of restraint. At the time of the Big Bang, computerised trading had nothing of the sophistication of modern algorithms. Robots followed a simple set of rules designed to launch sales if the market fell too quickly.
Programme trading, as this was called, soon came to the world’s attention when global stock markets suffered their ever worst day of falls: 19 October 1987, Black Monday, just a year after Big Bang. We’ll pick this up next week.
It turns out that technological processes have overflows far beyond their creators’ expectations. In fact, technology can start a stock exchange almost by accident, and in 1995 it did just that. The exchange was called OFEX, and if we are interested in the possibilities of small-scale exchanges for the funding of social goods, we should take good notice of its story.
You may remember from episode six how the Jenkins family established a small jobbing firm in London, specialising in dog tracks and holiday camps; how John Jenkins grew to be senior partner; how they made £1 million in five minutes of trading when the British Telecom issue came out; and how the firm was sold to Guinness Mahon and thence a Japanese investment bank. In the bear market that followed the crash of 1987 the trading desk was closed and Jenkins found himself unemployed, bruised and battered by a difficult period in a toxic working environment. But John had not just traded dog tracks. He had also developed a specialist expertise in the London Stock Exchange’s little-known Rule 163.
The rule, which later became Rule 535, and then Rule 4.2, allowed members to conduct occasional trades in companies not listed on the London Stock Exchange. Trades had to be conducted on a ‘matched bargain’ basis. This meant that the jobbing firm had to line up a buyer and a seller and ‘put through’ the trade, taking a commission of one and a quarter percent on each side. Each bargain had to be reported to the Stock Exchange and was carefully noted and approved by the listings department. It was clearly not meant as a volume operation. But Jenkins & Son already traded like this: jobbers in the smallest stocks could not rely upon a steady flow of buy and sell orders so were reluctant to hold stock on their books, tying up capital, possibly for years. Instead they would build up lists of potential buyers and sellers, and only when they could make a match would they trade. It was fiddly work, says John, though lucrative: ‘Nobody else wanted to do it, nobody else wanted to fill the forms out, run round and you would fiddle about in those days, would the client take 1,049, well I know he wants to buy 1,000 but will he take 963 and then you would have to piece it all together and do it…But for a grand a day, in those days!’
In the early 1990s John was twiddling his thumbs and missing his old trading days. He fancied starting a new firm but his application to the London Stock Exchange was twice turned down. John was on the verge of giving up but his blue button – his apprentice – from a few years before, Paul Brown, was made redundant as well, and this moved John to a final try. Brown remembers the conversation:
‘I rang John up and I said to him, “Look, John, just to let you know, before you hear it, I have been made redundant.” And he went, “Okay”. I’ll never forget it. He said to me, “Okay, Brownie, I’ll come back to you”. And that was it. And he rung me back the next day and he said, “Look, I went for a walk along the river, and I’ve thought about it. I’ve had this idea, trading what was 535(2) stocks then. How about you and I give it a go?” He said, “I can’t pay you a lot of money but it’s a start-up, we’ll get an office, just you and me, and we’ll give it a go.” So I said, “Yeah, fine.”’
The third submission was accepted by the London Stock Exchange, and on 11 February 1991, Jenkins and Brown set up JP Jenkins Ltd with a mandate to trade unquoted stocks ‘over the counter’ under the Stock Exchange rules.
There followed a period John remembers as one of the happiest in his working life. JP Jenkins occupied a small office above the ‘Our Price’ music store in Finsbury Square. A friendly Dutchman on the floor above would descend on their office mid-afternoon bearing a bottle of gin. It was just ‘two guys and a sofa’ trading with pen, paper and phone.
‘John had this old computer,’ says Brown, ‘so he brought it in, so it sat on the desk, but we never used it. We just had it there for show… it was a sofa and a computer that didn’t work. It did absolutely nothing. I mean it did nothing. It just sat there.’
Business was about making lists and matching, and the firm was soon known for the catchphrase “I’ll take a note”. They never said no, they just made a note; they had a good name, and they did well.
In 1992 the firm moved to Moor House in Moorgate. There was a separate room for the back office. Shares traded did not fall under the London Stock Exchange’s Talisman regime, so trades were settled in house, by the ‘manual XSP’ method. A typewritten catalogue of stocks includes some well-established entities such as Rangers and Liverpool football clubs, National Parking Corporation (NCP), breweries such as Daniel Thwaites and Shepherd Neame, Yates’ Wine Lodges, and even Weetabix. Alongside these were the stocks of smaller, high-risk, or less frequently-traded entities: Pan Andean Resources, Dart Valley Light Railway and the Ecclesiastical Insurance Office, to name three at random. Trading business grew steadily and the firm was profitable; John Jenkins’ horizons were not much bigger – no ‘delusions of grandeur’ as he put it.
No man is an island. Nor is any small market-maker, and the tendrils of automation soon began to wind their way into the comfortable life of these traders. Ironically, John was always an early adopter of technology. Even before the Big Bang swept terminals into London, he had travelled to the USA, visiting a broking firm named Herzog Heine Geduld, and watched the computer-based NASDAQ. He returned one of the few believers. His new firm soon got rid of the broken computer and installed its own bespoke system. Processes of automation bring existing taken-for-granted practices and assumptions to the surface, so we shouldn’t be surprised that John’s new computers simply mimicked what he and Brown had been successfully doing with pen and paper. But the big story was outside of John’s office.
Alongside SEAQ, the Exchange set up a ‘non-SEAQ board’. It was just another set of teletext screens, a home for Rule 535 stocks. It published rudimentary data and also historic trades. In doing so it made the traders’ margins visible, a matter made worse by screen’s long memory. John’s son, Jonathan, explains:
‘[It] didn’t show any live prices, didn’t show mid-price. It showed the previous day’s close and it would show you the price at which trades had happened. It used to piss people off because you’d get someone saying, “I bought them off you at nine and it prints on there you bought them at six.” It showed everybody exactly what we were doing.
But it was the market’s place. At some point in the early 1990s, JP Jenkins took over the operation of the LSE’s non-SEAQ notice board. The LSE had threatened to discontinue the service and the firm could not imagine life without this central, public space. To be excluded from what Knorr Cetina calls the ‘appresentation’ of the market – the electronic production of a virtual form – is to be excluded from the market itself. Alongside the non-SEAQ board the firm created ‘Newstrack’, a rudimentary news service for the small companies that it traded, displaying prices and a limited amount of company information over the Reuters network – Jenkins struck a chance deal with Reuters, then looking to expand its content. The service provided market capitalisation and some volume information. A rudimentary connectivity between the market makers and Newstrack meant that that if the price moved the market capitalisation would also move. Firms released final and interim results through the pages, published dividends and were encouraged to make trading announcements. In other words, Newstrack consciously mimicked the London Stock Exchange’s Regulatory News Service (RNS). JP Jenkins realised that there was money to be made here, too, and started charging firms to use the service. It had inadvertently stumbled into that new and growing revenue sector for stock exchanges: data provision.
Do you see what’s happening here? All of a sudden JP Jenkins is operating something that looks very like a small-scale stock exchange. It offers a venue where smaller companies can have their shares bought and sold, and where they can achieve some of the publicity and regulatory kudos that comes with a public listing. They can even raise money, for entrepreneurial corporate finance firms have spotted this thing that looks very much like a market and have begun to issue documents for fundraisings. JP Jenkins is making a tidy profit from its market-making, and starting to make inroads into the data sales sector. And all of this under the LSE’s regulatory banner. Remember that exchanges are themselves businesses, and that they operate in a competitive market for exchange services. It’s not surprising that the LSE starts to become really rather uncomfortable, so much so, that it gives in to political pressure on another front and sets in motion a process to set up another market for growth stocks. You must forgive me jumping around here, but that’s another story… What matters is that in 1995, the LSE closed both its Rule 163 reporting and the non-SEAQ board. It was an overtly defensive measure, but it was too late, for the path dependencies of organisations cannot easily be rolled back. Many of companies traded by Jenkins did not want to go to the LSE’s new venue. They petitioned John who – naturally – was keen to keep his business going. But he was confronted by another problem, the loss of his public venue, of his market place. What trader can manage without a marketplace? He had no option but to build his own space onto his existing data infrastructure. He called it OFEX (for off exchange). At first, it was nothing more than a label. Bolted onto the exiting Newstrack service, running through Reuters’ wires, OFEX was technically a trading facility. But taken as a whole, the assemblage – the wires, the screens, the trading mechanisms and networks of corporate financiers – could be seen as a capital market. On the basis of walks like a duck, talks like a duck (as one executive put it) it was a stock exchange. OFEX, specializing in the stocks of start-ups and small companies, was ready and waiting for the dotcom boom years of the late 1990s. But that’s a story for another episode.
So what have we learned today? That technological change – automation – shapes markets in ways participants do not expect, and that exchanges have histories and path dependencies that count for at least as much as regulation and global politics. And that, if you do want to build a stock exchange, the easiest way to do so seems to be by accident. Well, who said it was going to be easy?
I’m Philip Roscoe, and you’ve been listening to How to Build a Stock Exchange. If you’ve enjoyed this episode, please share it. If you’d like to get in touch and join the conversation, you can find me on Twitter @philip_roscoe or email me on email@example.com. Thank you for listening, and see you next time when, in the last episode of this first section, I’ll finally answer that question I’ve been asking all along: what’s in a price, and why does it matter?
 Sound recording from ‘touchassembly’ via freesound.org, under a creative commons attribution licence https://freesound.org/people/touchassembly/sounds/146268/
 Recorded by Cinemafia, https://freesound.org/people/cinemafia/sounds/24080/
 Norman S. Poser, “Big Bang and the Financial Services Act Seen through American Eyes,” Brooklyn Journal of International Law 14, no. 2 (1988): 327.
 Elizabeth Hennessy, Coffee House to Cyber Market: 200 Years of the London Stock Exchange (London: Ebury Press, 2001), 184.
 Eric K. Clemons and Bruce W. Weber, “London’s Big Bang: A Case Study of Information Technology, Competitive Impact, and Organizational Change,” Journal of Management Information Systems 6, no. 4 (1990): 49.
 Poser, “Big Bang and the Financial Services Act Seen through American Eyes,” 325. Quotation taken from Clemons and Weber, “London’s Big Bang: A Case Study of Information Technology, Competitive Impact, and Organizational Change,” 49.
 Sounds from freesound.org. Keyboard sound https://freesound.org/people/imagery2/sounds/456906/
Typewriter sound https://freesound.org/people/videog/sounds/240839/
 ———, “London’s Big Bang: A Case Study of Information Technology, Competitive Impact, and Organizational Change.”
 Interviewed by Fabian Muniesa, “Trading Room Telephones and the Identification of Counterparts,” in Living in a Material World, ed. T Pinch and R Swedberg (Cambridge: The MIT Press, 2008), 295.
 A Mr M Bennett, writing in the Stock Exchange Journal of 1959, and quoted by Juan Pablo Pardo-Guerra, “Creating Flows of Interpersonal Bits: The Automation of the London Stock Exchange, C. 1955–90,” Economy and Society 39, no. 1 (2010): 93.
 ———, Automating Finance: Infrastructures, Engineers, and the Making of Electronic Markets (Oxfoird: Oxford University Press, 2019), 128.
 K Knorr Cetina and U Bruegger, “The Market as an Object of Attachment: Exploring Postsocial Relations in Financial Markets,” Canadian Journal of Sociology 25, no. 2 (2000): 146.
 Pardo-Guerra, “Creating Flows of Interpersonal Bits: The Automation of the London Stock Exchange, C. 1955–90.”
 for more detail on this history see my booklet, downloadable at https://research-repository.st-andrews.ac.uk/handle/10023/11688
 Brown interview
 Karin Knorr Cetina and Urs Bruegger, “Global Microstructures: The Virtual Societies of Financial Markets,” American Journal of Sociology 107, no. 4 (2002).
 This is my claim, but it’s supported by Posner’s account of strategic rivalry among exchanges. Elliot Posner, The Origins of Europe’s New Stock Markets (Cambridge, Mass.: Harvard University Press, 2009).