High-Frequency Trading and the Centering of the (Financial) Periphery
Fri, 01 Feb 2013 05:19:00 GMT
This article is part of a series that was originally published in Volume: Centres Adrift at http:/
The financial world was once dominated by gossip, speculation, research and strategically-timed trades – by people, for people. With the introduction of computers and high-speed fiber optic cables however, the human grasp on trading is becoming evermore tenuous. Algorithms and bots are the new players on the stock market, engaging in high-frequency trading at nearly the speed of light, turning microscopic gains across a vast field into major profits. Matthew Tiessen explains how a new logic and geography is emerging from this unfathomably fast and complex practice.
In the world of finance – that is, in our increasingly ‘financialized’ world – any distinctions that may have once existed between (financial) centers and (financial) peripheries are being dissolved at, quite literally, the speed of light. This dissolution is being led by the effectively non-human demands of algorithmically-driven high-frequency trading (HFT). HFT is currently the most contentious and voluminous form of trading in the world. Indeed, these days over eighty percent of market volume is traded at microsecond and nanosecond-speeds by ultrafast computers across a globally interconnected meshwork of privately owned deep-sea fiber-optic cables according to the machinic – and unfeeling – logics and requirements of mathematics and (credit-) money. This high-frequency trading apparatus is driven by speed and feeds upon information. We can think of it as a sort of global video game that combines econometrics, psychology, game theory, risk theory, power asymmetries, and competing temporalities and spatialities into a sort of trans-national – or post-national – ‘poker-bot’ designed to siphon maximum micro-profits from exchanges in the name of injecting ‘liquidity’ into increasingly volatile financial markets. (Recall, however, that market volatility is a boon to high-frequency traders for whom the greater the ‘spread’ between prices the greater the potential profits).
In light of HFT’s appetite for unlimited speeds and unlimited financial-arbitrage opportunities, the central nodes of the global finance network – London, New York, Chicago, Tokyo, etc. – are becoming its peripheries insofar as these days it’s the spaces in between the exchanges where the real action occurs – or has the potential to occur. In other words, in today’s bot-driven world it’s not the prices of securities on the exchanges that matter, so much as the prices in between the exchanges or across different exchanges that create the conditions according to which HFT works its magic and data-mines the markets. The dissolution of the role of financial centers and the rise of the zones in between – whether those zones are literally desert, ocean, tundra, forest, or farmland – is resulting, as I will describe, in the emergence of a potentially more geographically distributed network of (financial) centers that derive their potency not so much from their being at the intersection of trade routes or from their being located where capital and commerce is concentrated, but by being geospatially located according to a sort of mathematics-of-in-between-ness – that is, by being optimally located in geographical space in a way that minimizes the impediments of time. Moreover, this emerging financial network will – as it evolves – become so profoundly integrated that no part of the network will be able to act or exist in isolation or outside of a relationship of absolute (arbitrage-gaming) interdependence. In other words, in the near future financial markets will only exist – and financial prices will only be able to be ‘discovered’ – relative to their relationship to other markets, and these relationships will only be able to be ‘gamed’ by its nonhuman agents having access to geo-spatially and geo-strategically distributed sites that are optimally positioned to minimize info-lag between formerly central and increasingly peripheral financial exchanges. According to these emerging financial realities financial markets of the future will find numbers and algorithms jousting with other numbers and algorithms in a decentralized virtual info-war, the objective of which is to produce bigger numbers (with us or without us).
In this essay, then, I want to describe this accelerating dissolution of (financial) centers and peripheries by: 1) describing how today’s HFT works and the ways it’s taking advantage of price gaps across exchanges, not to mention speed- and power-asymmetries across market participants and 2) speculating on the HFT of the future, one that – I argue – even has the potential to precipitate geo-spatial conflict insofar as powerful market actors (governments, ‘too-big-to-fail’ banks, hedge funds, etc.) will seek to occupy or colonize increasingly strategic and financially-significant sites while they all, at the same time, attempt to crowd their trades into ever smaller slivers of time. Such a financial future will necessitate that the earth itself become, as authors Wissner-Gross and Freer have recently speculated, an “econo-physical mechanism,” or an enabler of globalized financial connections, communication, and processes. This financialization of the planet’s peripheries, in turn, has the potential to lead to a heightened awareness of newly strategic geographical, political, and financial sites. In such a scenario global conflict, like financial trading, risks moving from the centers of power to power’s peripheries.
High-Frequency Trading’s Hyper-Fast Present
“And what about the effects of money that grows, money that produces more money? […] For it is a matter of flows, of stocks, of breaks in and fluctuations of flows; desire is present wherever something flows and runs, carrying along with it interested subjects – but also drunken or slumbering subjects – toward lethal destinations.”
-Gilles Deleuze and Félix Guattari
On May 6, 2010, the Dow Jones Industrial Average sustained its biggest intraday point decline in history. This now notorious so-called ‘Flash Crash’ revealed what can happen when automated digital decision-making exceeds human capacities to intervene. Although all the causes of the Flash Crash have not been identified conclusively, HFT was identified as the primary contributor. What was also made clear that day was that interconnected and automated computerized algorithms communicating and interacting at near light-speed on an effectively virtual battle field carry very real economic, not to mention social, risks. This unprecedented socio-economic episode, as Executive Director of the Bank of England Andrew Haldane observed at the time, exposed the precarious frontiers of algorithmically- and digitally-driven financial ‘innovation’: “Trading in securities [stocks, bonds, derivatives] generated trading insecurities,” he remarked, adding that “the impatient world was found, under stress, to be an uncertain and fragile one.”
HFT, just one instantiation of what Michel Serres has called the “algorithmic revolution” is a cultural and financial phenomenon brought to the public’s attention in 2009 by the New York Times in an article entitled “Stock Traders Find Speed Pays, in Milliseconds.” The continual growth of HFT has resulted in human traders, human communication, and human decision-making being rendered effectively obsolete by digital tools that assess risk and crunch economic data at near-instantaneous speeds. This outsourcing of investing and ‘value’-assessment to high-speed computers risks purely financial logics taking precedence over, or entirely ignoring, other more human-focused or ecologically motivated priorities. Indeed, it is the non-human or post-human dimension of these trading strategies that is perhaps most challenging to those of us who inhabit the ‘flesh-scape’, not only because of the financial risks and asymmetrical distribution of (technological and financial) power, but also because of the way HFT positions humans in relation to technology – i.e. as too slow to act relative to the technologies and tools we’ve devised.
HFT, then, is a relatively recent addition to the financial trading landscape made possible by an arguably post-capitalist, post-nationalist, and fiber-optically connected world of online digital media and communication. It is a financial strategy whereby powerful financial actors and firms (hedge funds, ‘too-big-to-fail’ banks, private equity firms, etc.) attempt to extract incremental profits at near lightning speed by ‘front-running’ other traders and by taking advantage of price gaps and arbitrage opportunities across globally-interconnected financial markets. Indeed, these days high-frequency traders are leveraging not only proprietary automated digital algorithms, high-speed computers, and privately owned networks of fiber optic cables, but also news-analyzing and headline-reading algorithms provided, for example, by news agency Thomson Reuters, whose subsidiary Thomson Reuters News Analytics sifts through fifty thousand news websites and over four million social media sites – owned, no doubt, by themselves as well as others – in order to sell bespoke market-making information to today’s high-speed traders.
Perfect HFT algorithms would, of course, not only operate at the speed of light, but would also preemptively be able to modulate future market events. These objectives are well on their way to being achieved. Indeed, today major stock exchanges cater to traders’ need to gain incremental speed and knowledge advantages by selling ‘co-location’ and ‘direct market access’ (DRM) arrangements that allow traders’ servers to reside as physically close to exchanges’ computers as possible (i.e. attached to the wall of the exchange’s data center) (Garvey and Wu 2010). These co-location agreements, of course, result in those who can pay to play – the ‘haves’ – gaining access to incoming and outgoing market information ahead of those without the fiscal resources to do so – the ‘have-nots’. The speed gains achieved by server locations designed to optimize high-frequency traders’ ability to game the system will only further disempower those unable to take advantage of tomorrow’s trading technologies. But as speed theorist Paul Virilio notes: “Speed and wealth go hand in hand.”
The lengths (both spatial and financial) that high-frequency traders will go to optimize their micro-temporal advantage is demonstrated by the recent announcement that a 300 million dollars of fiber optic cable destined purely for financial transactions is being laid across the floor of the Atlantic ocean from New York to London. By 2013 the 6,021 kilometer cable – known as the Hibernian Express – will plot the shortest possible route between two of the world’s most potent financial centers in order to shave five or six milliseconds off of transaction times (Williams 2011). In fact, the privately-owned cables being installed between exchanges by these market actors are so speed-centric that they will never carry audio, video, or internet media, or data at all. In other words, the world’s fastest and most powerful digital media networks and data-distribution backbones are not catering to the tweeting, googling, poking, app-using, or youtubing practices of the internet-using public, but rather, to the demands of private equity to be able to ‘front run’ the market activities of less well-heeled – and less speedy – competitors (i.e. the rest of us).
HFT, in so far as it is an attempt to breach both spatial and temporal limitations, can be understood as a ‘limit case’ of digital media itself. Indeed, the techno-enhanced profit-making strategies of HFT are so new and fluid that government legislation is struggling to keep up. On Dec. 13, 2011, for example, the Commodity Futures Trading Association’s Technology Advisory Group in the US held a year-end hearing, the objective of which was to begin to develop merely a working definition of what exactly HFT is in order to begin to develop legislation to deal with it. This, we can presume, is the limit of financial literacy, legality, and – perhaps – lunacy (considering that HFT already makes up the majority of trades on today’s markets).
High-Frequency Trading’s De-Centered Future (Light and the Speed of Space)
“A path is always between two points, but the in-between has taken on all the consistency and enjoys both an autonomy and a direction of its own.”
-Gilles Deleuze and Félix Guattari
In this final section I’d like to shift away from interrogating the HFT of today, to imagine that of tomorrow. We might assume that HFT, in its present state, would already have ‘smoothed’ financial space enough to appease the demands of contemporary capital. But its practitioners are not ones to be caught standing still. HFT might continue to evolve in the future, a future where, as I mentioned at the beginning, even the smoothest spaces can never be smooth enough for facilitating the demands of capital flows and where the curvature of the earth itself comes to be regarded as a speed bump – an information-inhibiting impediment to the ability of financial data to flow at high speed and in straight lines.
Virilio reminds us that speed – whether in the world of communication, commerce, or war – is power and that with great speed and power come potentially catastrophic risks. Virilio’s analysis of the risks of speed leads him, in a prescient reflection in an interview in 1998, to comment on the use of computers on financial markets. Virilio describes the increasingly concentrated dangers of ever-faster flows of financial risk-taking as follows:
“Program trading […] is the image of the general accident, no longer the particular accident like the derailment or the shipwreck. In old technologies, the accident is “local”; with information technologies it is “global.” […] We are faced [here] with a new type of accident for which the only reference is the analogy to the stock market crash, but this is not sufficient. […] I think that the infosphere – the sphere of information – is going to impose itself on the geosphere. …. The capacity of interactivity is going to reduce the world, real space to nearly nothing.”
The compression of space and time described by Virilio is nowhere more palpably manifested than in the machinations of the HFT of the fast-approaching future which, having reduced temporal impediments to a technological minimum – insofar as market information now travels at ninety percent the speed of light – is left seeking to smooth physical space in the never-ending quest for ever-smaller trading advantages.
One of the ways this smoothing of trade-space is being accomplished is by the smoothing of geographical space. Looking into the future, some researchers have suggested that this will be accomplished through the optimization of the geo-spatial localizations of the high-frequency trader’s computer servers by actually optimizing the location of their servers across the surface of the globe – whether that surface is on land or on water. After all, once your data is moving through space at the speed of light, you need to optimize light’s speed by adjusting the paths it travels through space.
In a recent article entitled “Relativistic Statistical Arbitrage” Alex Wissner-Gross of the MIT Media Laboratory, and Cameron Freer of the University of Hawaii suggest that in HFT’s inevitable future the earth itself will be tasked with becoming what they term an “econophysical mechanism” charged with smoothing inter-exchange communication and transaction times to traders’ advantage. According to their work, the speed with which contemporary financial transactions are being carried out – and the increasingly important role being played by the peripheral spaces in between exchanges – necessitates that “optimal intermediate locations between trading centers” be located so that “coordination of arbitrage trading from those intermediate points [is able to maximize] profit potential”. In other words, in order to maximize trading efficiencies by reducing latencies between exchanges, intermediate locations must be mathematically and geo-spatially determined so that trading servers can be positioned to minimize the distance trades must travel. They describe how what they call “relativistic statistical arbitrage trading nodes” can be optimally distributed “across the Earth’s surface” to minimize distances between globally significant stock markets. As they explain:
“Recent advances in high-frequency financial trading have made light propagation delays between geographically separated exchanges relevant. [We] show that there exist optimal locations [on our planet] from which to coordinate the statistical arbitrage of pairs of space-like separated securities, and calculate a representative map of such locations on Earth.”
More often than not, these “optimal locations” are located in the world’s peripheries: in, for example, the middle of oceans and deserts – spaces Deleuze and Guattari identified as ‘smooth spaces’, that will – in our algorithmically-driven future – be driven or striated by fiber optic striations designed to maximize the movement and smoothed flows of money-data to achieve maximum arbitrage-seeking advantage. These optimal locations, of course, would far outnumber the world’s financial exchanges since each exchange (New York, for example) would require as many intermediate nodes as there are other exchanges in order to maximally take advantage of the arbitrage opportunities.
The speed imperatives inherent to HFT, then, can be understood to have made the earth itself a decentralized form of digital media, a communication technology, a tool for digital mediation. As Wissner-Gross explains it:
“Eventually, this may lead to the development of a truly global computing infrastructure, covering even the most remote locales. I see this work” he adds, “as one possible justification for making the entire surface of the planet more computationally capable… and in effect, making the whole planet smarter.”
Wissner-Gross and Freer’s proposal for geo-optimization in service of achieving maximum speed demonstrates that finance’s demand for speed and today’s algorithm’s capacity to parse data in this ‘new knowledge economy’ exceed all spatial and temporal limits and disrupt the conventional role once played by (financial) centers and peripheries. We can imagine, then, that the location of ideal server space will become increasingly contested information-rich terrain, as large numbers of financial and state actors seek spatial and speed advantages in their efforts to game global markets within smaller and smaller units of time. We can imagine too that as transmission times are technologically and spatially shrunk, risk will become increasingly concentrated and communication information highways increasingly congested according to the dictates, quite literally, of a zero sum game. In such potential spatial and temporal conflict scenarios, perhaps communication between real, live, individuals – rather than just between competing algorithms – might, in fact, stage a comeback. We are left, then, to imagine a future where financial geo-engineering and the immaterial- and even inhuman-labor performed by HFT machines leads increasingly to geopolitical conflict as a result of optimized trading node-territory colonization with the pursuit of smoothness by technologically-driven financial actors leading, in turn, to a concentration of both financial and military conflict. As these new sites of financial warfare emerge it will again become clear that today – and in the future – state space will continue being subsumed by financial space and that profit-led smoothing remains the way striations will continue to gain ground in the twenty-first century. The last word, then, will go to Deleuze and Guattari who – following the demise of the Bretton Woods financial system thanks to President Nixon in the 1970s and the de-linking of money from its material restraints (gold) – could not have envisioned our financialized, de-centered, deterritorialized, post-capitalist, post-humanist, and post-national future more clearly:
“When the flows reach this capitalist threshold of decoding and deterritorialization […], it seems that there is no longer a need for a State, for distinct juridical and political domination, in order to ensure appropriation, which has become directly economic. The economy constitutes a worldwide axiomatic, a “universal cosmopolitan energy which overflows every restriction and bond” [….] Today we can depict an enormous, so-called stateless, monetary mass that circulates through foreign exchange and across borders, eluding control by the States, forming a multinational […] organization, constituting a de facto supranational power untouched by governmental decisions. But whatever dimensions or quantities this may have assumed today, capitalism has from the beginning mobilized a force of deterritorialization infinitely surpassing the deterritorialization proper to the State.”
Matthew Tiessen is a Postdoctoral Research Fellow at the Infoscape Research Lab at Ryerson University in Toronto. He teaches and publishes in the area of technology studies, digital culture, and visual culture.
1 A. D. Wissner-Gross and C. E. Freer. ‘Relativistic Statistical Arbitrage’ in Physical Review 82, 2010 pp. 056104-1-056104-7.
2 Gilles Deleuze and Félix Guattari. Anti-Oedipus: Capitalism and Schizophrenia, trans. Robert Hurley, Mark Seem, and Helen R. Lane (Minneapolis: University of Minnesota Press, 1983).
3 Andrew G. Haldane ‘Patience and Finance’ Speech by Mr Andrew Haldane, Executive Director, Financial Stability, Bank of England, at the Oxford China Business Forum, Beijing, September 9 2010. Bank for International Settlements. Retrieved from http:/
4 Michel Serres ‘The Science of Relations: An Interview’ in Angelaki, 8:2, 2003. pp. 227–238.
5 Greg Elmer and Andy Opel ‘Surviving the Inevitable Future’ in Cultural Studies, 20:4, 2006 pp. 477–92. & Brian Massumi “Potential Politics and the Primacy of Preemption” in Theory & Event, 10:2, 2007.
6 Ryan Garvey and Fei Wu ‘Speed, Distance, and Electronic Trading: New evidence on why location matters’ in Journal of Financial Markets, 13:4, 2010 pp. 367–96.
7 Paul Virilio and Sylvère Lotringer, ed. Politics of the Very Worst, trans. Michael Cavaliere, (New York: Semiotext(e)1999).
8 Christopher Williams ‘The $300m Cable that will save Traders Milliseconds’ in The Telegraph. 2001 Retrieved from http:/
9 Gilles Deleuze and Félix Guattari, A Thousand Plateaus: Capitalism and Schizophrenia, trans. Brian Massumi, (Minneapolis: University of Minnesota Press 1998).
10 Paul Virilio and James Der Derian, eds.The Virilio Reader, trans. Michael Degener, James Der Derian, and Lauren Osephchuk (Oxford: Blackwell Publishers. 1998).
11 Thomas McCabe ‘When The Speed Of Light Is Too Slow: Trading at the Edge’ (2010). Retrieved from http:/
12 James Rickards. Currency Wars: The Making of the Next Global Crisis. Portfolio hardcover. (2011).
13 Gilles Deleuze and Félix Guattari. A Thousand Plateaus: Capitalism and Schizophrenia, trans. Brian Massumi (Minneapolis: University of Minnesota Press, 1987).
The opinions and writing contained in this article are of the author alone and do not necessarily represent those of HFTReview.com.