Computing, communication and markets

August 19, 2013

How does the history of computing contribute to the history of finance? This issue has been on my mind for some time. More specifically, I’ve been wondering what we need to know about the evolution of computing systems to understand how markets have changed with the implementation of machines.

True computer aficionados will probably find the following totally mundane, but for anyone studying finance who hasn’t seriously considered this question these videos are worth a peek.

My first selection is a theater clip from 1948 which shows an analog computer invented by Vannevar Bush* at UCLA. This early mechanical device solved differential equations but did not yet have language, a screen interface, or internal programs.

Despite spinning gears and moving parts, the narrator cheerfully compares this early computer to a human brain! (The Thinking Machine). The machine required teams of people dedicated to maintaining smooth physical functioning.

Fast forward to 1972, into the age of digital machines. This short documentary – Heralds of Resource Sharing – explains ARPANET, the first attempt to make electronic computers transmit information from one to another. Here we see the invention of a network, in which signals will move between machines through telephone wires.

In this period, the computer scientists employ an entirely different language peppered with terms like ‘talking’ and ‘the user’. These machines do not just calculate; they pass messages, they communicate.

Wow! Until I saw these videos I had never really considered the distinction between [computation] and [communication]. The two concepts are intimately tied together in neoclassical theories of markets and trading which assume a natural relationship between calculation and communication when people are economic agents. And yet, there is a precise moment in engineering history when calculating machines morphed with (tele)communication systems!

(Thanks to Yuval for nudging me with his comment on my last post, and to Alison Powell for talking it out.)

*Footnote: Bush was a major figure in US science policy in post-war America. His writings are core reading in STS because they found the argument for the National Science Foundation, which now provides funding for American STS as well as scientific projects those scholars have tended to study.  

4 Responses to “Computing, communication and markets”

  1. Tero Karppi Says:

    Thanks for the excellent post and opening an important discussion of the relation of computational culture and financial markets. For me what is important here is that computers do not only calculate or pass messages but they also run the financial markets in particular ways as well as shape the ways in which we understand these markets. That is to say that computers and their software are not neutral operators but their operations are always designed and programmed. Especially events where the computational logic fails, such as the flash crash of Knight Capital caused by a technical glitch last year, reveal us how deeply contemporary financial markets are run by computational processes and how these processes are automatically able to generate real situations like ‘financial crises’ out of nothing.

  2. marthapoon Says:

    Hi Tero,

    Thanks for joining the discussion!

    Yes, the idea of market design (market microstructure) is well tread terrain in economics. SSF has its own term to talk about this: ‘market devices’, which does not exclusively financial markets, but to any technological arrangements – including the way stalls are set up in a physical marketplace, or how the terms of fishing licences are set up – designed to format terms of exchange.

    The automation of stock exchanges has indeed, as you point out, led to the replacement of other possible types of devices with computation processes in equities markets. What is at stake in research, and what distinguishes it perhaps from work in finance and economics, is the attempt to capture what a financial market is becoming – the generation of novel market conditions – through digital and algorithmic systems. So equities is one key place where history of computation meets the history of finance. And researchers in both fields recognize the concept of design.

    BUT, both kinds of researchers will have to move quite slowly to absorb what the other side thinks is the ‘object’ of design. From the point of view of finance, the idea that ‘computation processes’ can automatically generate unwanted or unexpected volatility in the stock market, black boxes the numerous underlying elements – What is being calculated to drive the bid? What is being transmitted to and from the market? What agents are receiving what kind of information? What laws are governing this moment? etc. – that must come together for trades to be executed. Finance comes with a lot of rich and complicated domain level detail.

    In equities for example, a number of functional process come into play such such bid ordering which precedes market automation, even if it is now expressed through a computer system. [For the core work on this topic see: Pardo-Guerra, Juan Pablo (2010) Creating flows of interpersonal bits: the automation of the London Stock Exchange, c. 1955-90 Economy and society, 39 (1). 84-109. ISSN 1469-5766] In other words, financial discussions drive the design of computational processes matter as much as ‘the fact’ of computational processes in and of themselves. The question is, how do we put these two histories together in one narrative…?

    This is why I encourage a ‘back-to-basics’ approach. As the fields currently stand, some scholars understand finance but much less about computational processes, while others understand history of media technology with very little insight into the emergence of markets. This is ‘silo busting’ we need to work on… and you chiming in here is a great first step!


  3. Juan Pablo Pardo-Guerra Says:

    Great post, Martha!

    I agree that a more consistent conversation between students of finance and students of information technologies is necessary. Indeed, I think some important steps have already been taken (think, for instance, of Fabian’s work on the Paris Bourse, the more recent work by Yuval and Daniel on the National Market System, or recent studies on second generation Electronic Communication Networks like Island). But I am sue that you will agree that much territory remains unexplored.

    One of the things that intrigued me the most about your post (and, indeed, some other discussions elsewhere on finance and technology) is the idea of ‘design’. Something that I have found particularly useful in understanding financial markets from the perspective of infrastructures is that ‘design’ seems to be a local orientation, rather than a global realization. In other words, whilst I think that many actors in finance seek to ‘design’ their informational environments, they rarely succeed, simply because their designs must both interface and compete with those of so many other agents. Finance is built upon a complex and disjointed series of gateways and interconnections, rather than upon a seamless sociotechnical web. Indeed, just like it would be odd to think of software as a designed field (it is, after all, so full of sub-optimal, inelegant solutions that design is the last thing one has in mind), so too it would be wrong to think of finance as a product of design.

    Why am I bringing up this point? First, because I think it is something that an infrastructures perspective adds to how we imagine markets. But second, and more importantly, because it attenuates the way we think about market automation in the context of a potential narrative that combines the history of finance with the history of media, computing and communication. As students of finance start new conversations with students of media and communication, there is a slight risk that what will focus our attention is the search of grand narratives, of concrete materializations of some sort of ‘cybernetic’ world in markets that show modern finance as the realization of a particular computational dream. Putting these two stories together may result, in this sense, in a narrative that is, just perhaps, too large to capture the fault lines of financial innovation. Do we need a narrative, or a series of small stories that show the difficulties of building the infrastructures of markets? Is a narrative that combines the histories of finance and media possible, viable, or even desirable?

    Perhaps in establishing the conversation, we should focus less on how these two narrative articulate and more on the disjuntures, path dependencies, failures and surprises that characterize the development of technologies in finance. If the experience of the London Stock Exchange is anything to go by, what we will probably find is certainly a number of attempts to design markets, to domesticate the marketplace and bring into being a particular cybernetic dream through technology; but we will see, too, that these attempts often fail, and that what remains is a heterogeneous series of systems shaped by serendipity and chance.

    Anyway, great conversation to have!

  4. […] experienced our own debate about design in the last post (Computing, communication and markets). Tero raised the issue that computers and software are necessarily designed by people; I replied […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: