Sociology of Market Microstructure Workshop (The New York Stock Exchange, August 9th, 2013)

Organizers: Daniel Beunza (LSE) and Yuval Millo (University of Leicester )

Bringing sociology to the NYSE

This workshop put together a coherent body of sociological research focused on financial exchanges and related issues —high frequency trading, Regulation NMS, the Flash Crash, etc. The goal was to develop a sociological literature that can offer an intellectual counterpart to the economics of “market microstructure”. To do so, we invited fifty sociologists to join us in the old Luncheon Club of the NYSE.

Understanding automated financial exchanges

Why this workshop now? Three decades ago, sociologists like Wayne Baker and Mitch Abolafia developed a distinct understanding of financial exchanges, emphasizing the importance of social relations on trading floors. The existence of networks of intermediaries on the floor, they argued, helped investors grapple with uncertainty and limit opportunism. Nowadays, however, the current trend towards automation calls for revisiting this sociological framework in order to explicitly consider how automation reshapes the social aspect of the market. Similarly, the ongoing regulatory debate over high frequency trading calls for renewed sociological research. Appropriately, the workshop took place at the New York Stock Exchange, and included a panel of industry experts as well as a visit to the floor.

SCHEDULE OF PRESENTATIONS

(Click on the hyperlinks for access to the papers)

9.05 First session: sociological approaches to exchange automation

Daniel Beunza (Management Department, London School of Economics): Folding: Integrating algorithms into the floor of the New York Stock Exchange   

Abstract: How does automation reshape markets? Existing sociological studies have argued that market automation need not entail a dilution of social relations, but the empirical evidence is inconclusive. Our multi-period ethnography of the New York Stock Exchange addresses this gap by exploring how the NYSE automated trading while preserving its floor intermediaries. Our study reports on observations in 2003 before automation was introduced, outlining the functions traditionally played by specialists and floor brokers. It then analyzes how the Exchange preserved these intermediary roles in 2006-08 as it introduced algorithmic order matching, and proposes the notion of folding to denote the process of automating a market while preserving its social structure. Finally, our analysis of the Flash Crash in 2010 suggests that folding allowed the NYSE to outperform its floorless rivals.

Audio recording of the talk

Donald MacKenzie (Sociology Department, University of Edinburgh): Insurgent capitalism: Island, bricolage and the re-making of finance

Abstract: Drawing on recent discussions of the material cultures of markets and of financial innovation as bricolage, this article explores the development of Island, a new share- trading venue set up in 1995. We examine Island’s roots in a very specific conflict in the US financial markets and in the information libertarianism of ‘hacker culture’, and examine the material bricolage involved in Island’s construction. The article also outlines the processes that led to a dramatic ‘Latourian’ change of scale: Island was originally a ‘micro’ development on the fringes of US markets, but within little more than a decade key features of Island became close to compulsory, as the nature of North American and Western European share trading changed utterly.

Audio recording of the talk

Discussant:  Karin Knorr Cetina (Sociology Department, University of Chicago)

Audio recording of Knorr Cetina’s comments

10.00 Second session: historical sociology of financial exchanges

Yuval Millo (School of Management, University of Leicester): Automating away the intermediary: The techno-politics of market shaping 

Abstract: The automation of financial exchanges is one of the most prominent changes in global capitalism in the last decades, affecting markets around the world. In spite of the importance of the phenomenon, there is no agreement about the nature of financial market automation process. On one hand, economists claim that automation was enacted because it lowered costs and quickened processes. On the other hand, economic sociologists argue that automation reflects different ideologies.  Neither of the approaches, however, account for nature of the automation process itself and, in particular, neither addresses the question of what were the changes that financial markets underwent as automation unfolded. In this paper we analyze the history of market automation in the US, as we follow the Securities and Exchange Commission’s National Market System initiative from the late 1960s to 2007. Using an extensive documentary analysis and interviews with key figures, we find that market automation was ignited by the SEC, that promoted a vision of inclusive, democratic markets, where the principals control the trading process, and intermediaries are replaced by machines. Automation, however, unfolded as a ‘three-legged stool‘. The SEC’s regulatory vision was legitimized by economic theory and was realised by information technology.  The initiative, however, ended up enriching  ‘big players’ who control automated systems and contributed to more fragmented less transparent markets. Our work contributes empirically to the economic sociology literature and develops a theory about the techno-political construction of the economic actor.

Audio recording of the talk

Michael Castelle (Sociology Department, University of Chicago): The Transaction Concept: Concurrency and Durability for Financial Markets

Abstract: A transaction is a form of exchange with a bounded beginning and end, and thus differs strongly from other manifestations of dyadic social exchange, such as a gift or loan, whose endogenous social repercussions may persist long past the moment of direct interaction. Markets for financial securities are predicated on the material mediation and optimization of flows of transactions. These environments—when coexistent with technological systems which permit said mediated flows to be reliable and durable in situations of varying volatility—caneventually facilitate the “automation” (or symbolically-structured initiation and execution) of said transactions.The underlying historical and sociotechnological question, then, is: how was the high reliability for large volumes of continuous transactions achieved, especially in what had been an era of both low reliability and limited concurrency in computing hardware and software? I suggest that the digital mechanization and automation of market processes, while demonstrably a historically gradual and socially contingent development (Pardo-Guerra (2010b), MacKenzie (2012)), is one whose possibility is also prefigured in the very concept of the transaction as a materializable, temporally bounded and socially disembedded object. A close examination of the historical ontology of the computerized transaction(as it developed in the nascent database research literature of the 1970s, and subsequently in the commercial1implementations of Tandem Computers in the 1980s) can then be used to illustrate the crucial early transitions of mechanization at the NYSE which took place between the Paperwork Crisis of the 1960s—in which over 100member brokerages departed due to “severe cases of too much business” (Rustin, 1975)—and the Black Monday crash of 1987, which instead saw then-doubled rates of 200,000 transactions a day with no major backend system failures.

Audio recording of the talk

Juan Pablo Pardo-Guerra (Sociology Department, London School of Economics): Making markets: infrastructures,engineers and the moral technologies of finance 

Abstract: How do markets change? Conventional sociological accounts answer this question by stressing the weight of social structures on the  transactional core of the marketplace. This paper provides an alternative approach. Market change is identified as an infrastructural transformation in which novel market devices and  classifications are defined as the legitimate platforms for exchange.  Rather than focusing on the traditional subjects of sociological  enquiry, this study looks at the developers of market infrastructures  in order to appraise the evolution and reinvention of markets.  Empirically, the paper focuses on four historical episodes relating to  the invention and dissemination of the electronic order book, a device  that is central to global financial capitalism. These show how infrastructural work was implicated in creating the politics and  structures of modern finance by criticising established institutions,  mounting competitive challenges against incumbent institutions,  establishing expansive projects of marketization and integrating  otherwise disconnected marketplaces.

Audio recording of the talk

Discussant: Bruce Carruthers (Department of Sociology, Northwestern University)

Audio recording of Carruthers’ comments

 11.30 Industry panel

An industry panel comprising the founder of two electronic exchanges, an automated trading pioneer, a specialist in “social trading” and the ex-Chairman of a major exchange. It was moderated by a journalist with long experience in market microstructure.  Panelists (left to right): William Harts (Harts & Co),  Andrew Upward (Rosenblatt Securities), Stephen Wunsch (Wunsch Auction Associates LLC),  Neal Wolkoff (Wolkoff Consulting Services, LLC) Moderator: Nina Mehta, financial journalist.

Audio recording of the session

12.45 Lunch: Quad Room, round tables Over lunch, courtesy of the NYSE, the participants had the chance to hear how the Exchange has been adapting to automated trading directly from its own executives, Bob Airo and Lou Pastina.

   Lou Pastina (NYSE) and Bob Airo (NYSE)

1.45 Third Session: social relations in financial exchanges

Aaron Pitluck (Sociology Department, Illinois State University): Watching Foreigners: How counterparties enable herds, crowds, and generate liquidity in financial markets

Abstract: This article provides a new view on the old problem of herding in the global south by foreign portfolio investors. It advocates a liquidity perspective that problematizes the capacity for a herd to form because of the absence of sufficient counterparties willing to trade. Drawing on ethnographic interviews with local professional investors in Malaysia (a substantively and theoretically important stock market) the findings are non-intuitive relative to the common-sense expectations of the information asymmetry and identity-based herding literatures. Although locals watch their foreign competitors closely, and therefore could imitate their trades, these small, local finance firms find few reasons to imitate these powerful international actors. Instead, locals enable crowds of foreigners because they are willing to be counterparties even when they perceive the foreigner’s trade as savvy, highly skilled, or informed. The conclusion explores implications for herding, global capital flows, and social structures that may generate liquidity in business-to-business markets.

Audio recording of the talk

Zsuzsanna Varga (School of Management, University of Leicester): Do financial markets work? The missing concept of work in social studies of finance.

Abstract: Currently we are only talking about financial markets as markets, the whole theoretical and conceptual apparatus of SSF and soc of finance generally, is geared to parallel economics. This is of course, very important. But in the process we shed the idea and the subjective disposition of people involved, that this is also a kind of work, job, occupation, or perhaps even profession (and a lifestyle in the case of traders). I will argue that to understand market microstructure, you have to understand the organization, experience and constitution of work as it takes place in different sites of these markets. Bringing in concepts of work such as control, management, hierarchy (heterarchy), even bureaucracy, and the sociology of accounting, can help understand some of the dynamics of stock exchanges and financial markets more broadly. One of those dynamics is how participants have reacted to different elements of “automation”. So, we can narrate the shift from face-to-face, manual, electronic, automated, high-frequency in new ways. There are some historical studies on parts of this shift, such as Juan-Pablo’s work on LSE or Alex Preda’s on occupations in NYSE in the 19th century. We can also use concepts like (career) trajectories and labor markets–Godechot was and Karen Ho is doing some of this. But there is much more to do here, such as more explicit discussions of skills. This is implicit in the studies of face-to-face trading and automated trading, as part of explaining market outcomes. A big question is whether trading is actually work in the sense of labor. If it is, how can we account for it within theories of capitalist production and exchange. Informally, insider accounts like Michael Lewis’ hinted that face-to-face brokers were more similar to blue collar workers. But if we think it is not work, then what is it? The talk is more provocative rather than assertive, so I will not be able to answer all of these questions. To begin thinking about trading as work, I will briefly show an example from my own work on the selling/service aspects of banking, how concepts of control can be used for financial work, and how these shape the content of finance. Now a complication is that after STS, we can’t really say that this work is just what is done by people. This is really clear in the case of financial markets. This kind of “market work” is done in the distributed cognitive networks that many authors have already described (including you two). So there is a really fruitful area of not only the sociology of work contributing to finance, but also SSF contributing to the sociology of work.

Audio recording of the talk

Robert Wardrop (The Judge Business School, Cambridge University): From Bank Financing to Bond Financing: Financial Exchanges and the Small and Medium Enterprise Micro Bond Market.

Abstract: The financial crisis has reduced the lending capacity of banks, particularly for lending to small and medium-­‐sized companies (SMEs). Since 2009, SMEs have raised debt financing via small ‘micro bond’ issues sold to individuals and small institutional investors. Initial results from a study of 200 micro bond issuers and 16,000 investors in Germany and the UK suggests that the relationship between these issuers and their investors is much closer than the relationship between large corporate bond issuers and investors in the debt capital markets. This has implications for the comparative advantage of incumbent financial exchanges relative to local entrants launching platforms for the listing and trading of SME bonds.

Audio recording of the talk

Discussant: Bill Maurer (Department of Anthropology, University of California, Irvine)

Audio recording of Maurer’s comments

3.30 – 4.15 Visit to the trading floor (Members’ Gallery)

4:30 – 6:00 After the workshop the discussions continued in Bobby Van’s Wall Street bar that is nearby, at 25 Broad Street.

Press coverage

“What do Traders in Emerging Markets Want? Just Ask Them.” Brendan Greeley, Businessweek, August 13, 2013.

Social media

TWEETER HASHTAG: #nysesociology

Media contact

If you wish to report on this workshop, feel free to email/ call us at:

Daniel Beunza: d.beunza@lse.ac.uk +44 20 7106 1146

Yuval Millo: ym95@le.ac.uk +44 116 229 7385

Photo credit

First image was taken by Daniel Beunza. The images from the workshop were taken by Martin Giraudeau