Afficher l’image
Crédits : Princeton university press

1Trading at the Speed of Light, Donald MacKenzie’s latest book, is a sociological and historical study of high-frequency trading, a class of ultra-fast and automated trading where profits depend largely on the speed at which trading is executed. These algorithms automatically issue (and very often cancel), buy and sell orders by exploiting basic market signals. Since these signals are basic, speed of execution is essential. This is what distinguishes high-frequency trading from simple automated trading. A high-frequency trading is able to execute millions of orders in only few minutes. Developed in the late 1990s, this form of trading accounts for around half of trades in stock markets and even more for buying and selling orders issued. This book aims at unveiling this significant and yet little-known aspect of contemporary finance.

2As MacKenzie explains in the introductory chapter, the book belongs to the research program of the Social Studies of Finance. These studies do not focus on the economic aspects of financial markets, but rather on the social and political dimensions that govern the practice of market participants. One of the main strengths of the book is the richness of interviews that support the analysis, 337, with various actors involved in high-frequency trading industry: traders, brokers, analysts, members of trading venues, regulators, technicians, and researchers (21). This allows MacKenzie to analyse the emergence of high-frequency trading through a variety of situated perspectives, each reflecting a particular group of interests. While MacKenzie focuses only on western markets, particularly the US market, he covers the most important and pioneering exchanges regarding high frequency trading, such as the Chicago Board of Trade (CBOE), the Chicago Mercantile Exchange (CME), the New York Stock Exchange (NYSE) or the NASDAQ.

3Another virtue of the book is its focus on the material aspects of high-frequency traders’ practices. The author adopts an approach he calls “the mundane political economy of finance,” namely the “undramatic, everyday ways in which money is made, often individually fairly small amounts of money, but time and time again” (17). The book does not merely aim to understand the overall impact of high-frequency trading, but rather to analyse the daily and material constraints faced by industry players. This approach is particularly beneficial for understanding these new trading practices, which are highly technical, and whose inner workings remain largely unknown to the public and even to economists.

4After the introductory chapter, the book is structured into five main chapters. Chapters two, three, and four analyse the rise of electronic and algorithmic trading within major financial markets. Each of these three chapters focuses on different markets: derivatives, stocks, and, for the fourth chapter, both debt and currencies. The chapters offer a multi-dimensional history, covering the economic, social, technical and regulatory aspects central to the emergence of high-frequency trading. In this respect, the book is a brilliant demonstration of what sociologists call the social construction of markets, and the cultural and political dimensions that permeate them.

5Chapter two, for instance, recounts the steady introduction of electronic trading at the Chicago Mercantile Exchange. Historically, trades at Chicago were conducted through an open outcry system. Traders gathered in a pit and shouted or made codified hand signals to trade with each other. MacKenzie explains how, from the 1970s onwards, the introduction of computers and electronic trading gradually changed the organisation of the market. The arrival of computers was primarily driven by economic factors. In particular, the liberalization of capital flows increased the competition with other financial markets. Chicago Mercantile Exchange officials adopted electronic trading because they feared of being overtaken by European exchanges. However, the development of electronic trading at Chicago was slow and faced significant cultural resistance from brokers (the class of traders responsible for ensuring product liquidity on an exchange). This resistance was primarily cultural: open outcry trading required physical attributes to be able to stand out in the pit. Such skills became obsolete in front of a monitor. Moreover, brokers held numerous superstitions associated with the physical aspects of trading, like a special location in the trading pit or a specific pen used to finalize a trade. All these aspects seem incidental compared to the promise of profit associated with the introduction of digital technologies. Still, these rituals were an integral part of a way of working and a way of conceiving the profession. Open outcry only disappeared in the 2000s at Chicago. This is just one example of the many illustrations in the book of the cultural aspects that go hand in hand with, and sometimes clash with, the purely economic logic of arbitrage.

6In these three chapters, MacKenzie examines how high-frequency trading has transformed the structures of financial markets, namely the set of rules and mechanisms that organize a market and facilitate trading. To an economist, trading exchanges are often uniformly grouped under the label of “financial markets.” From the sociologist’s point of view, each exchange is different. For historical, sociological, geographical, technological, or even regulatory reasons, exchanges have brought together different types of traders, and the latter have created different organizational structures for each exchange. As a result, MacKenzie shows that the impact of high-frequency trading has not been uniformed across markets. For instance, high-frequency traders have become increasingly popular on the equity and derivatives markets, but they faced strong resistance in other markets such as those of sovereign debts.

7The fifth chapter delves into the central issue of the speed of information transmission. The chapter recounts the relentless race for speed that the industry embarked on during the 2000s. MacKenzie illustrates how this race for exploiting arbitrages is embodied in material constraints, both on macro and micro levels. On the macro level, the central issue is to transmit as quickly as possible information between trading exchanges. MacKenzie offers a comprehensive analysis of the “gold line” (139), the internet connection between Chicago (where derivatives are traded) and New Jersey (which houses the data centre of the main capital markets). Hundreds of millions of dollars were spent by the industry to improve the speed of information transmission by a few microseconds. The chapter also analyses how the material challenges lie at the micro level, in particular, in the processing of information. The race for speed is also embodied in the materiality of computers: the chip storing information, the languages used for programming, and the format in which messages (mostly buy and sell offers) should be encoded. For instance, MacKenzie tells the story of how high-frequency traders use the Internet data transmission protocols to cancel orders faster than other high frequency traders (168). Based on a particular anticipation, an algorithm can execute a series of buy (or sell) orders and cancel the order if the anticipation is not fulfilled. To save time, some traders have exploited the way information is transmitted across the internet. Internet protocols break down the sender’s message into packets, which are then transmitted individually to the receiver’s computer, where they are reconstituted. Instead of sending a cancellation message, the algorithm sends an incomplete number of packets to the exchange. If the anticipation is fulfilled, it transmitted the rest of the packets, and the order can be executed. If, on the contrary, the anticipation is not fulfilled, it stopped transmitting the rest of the packets. The exchange’s system receives the incomplete file and treats it as an error, which de facto, from the trader’s viewpoint, cancels the order.

8The sixth chapter examines the two major classes of high-frequency traders, the “makers” and the “takers” of liquidity. In high-frequency trading, “makers” are traders who place limit orders that add liquidity to the market, typically hoping others will match their orders, while “takers” are traders who place orders that remove liquidity by matching against the existing limit orders on the order book. Essentially, makers provide potential trades and takers accept them. The chapter explains how these algorithms work and how they interact with each other. The chapter also shows how these two forms of high-frequency trading entail the development of two distinct perspectives on trading, with different narratives justifying and legitimizing their practice (174-175). The “makers” highlight the role of high-frequency trading in ensuring market liquidity. Since they fill the order book, they allow other traders to trade more easily. Meanwhile, the “takers” argue that they improve the price signal. Based on a particular piece of information, these traders clear the order book, which modifies the price, and allows that piece of information to be integrated into the price. These narratives are not new. In fact, they have been present somehow since the emergence of modern finance in the late 19^th century, where they were already used to warrant the social utility of speculation. However, MacKenzie also shows how these discourses primarily serve the traders themselves. Indeed, the millions spent to gain few milliseconds created legitimacy crises among high-frequency traders. For instance, a interviewee (a “maker”) recounted how, in the aftermath of the 2007-2008 crisis, he “couldn’t look his grandmother in the face.” His manager called a meeting of all the firm's employees to explain to them the value of their business and in particular how they helped to ensure the liquidity of the financial markets. (174). This insight sheds light on the moral and psychological aspects of trading practices which coexist with the search of arbitrages.

9The depth of detail in MacKenzie’s analysis may also cause its only flaw. The reading is slowed down by the high level of detail provided and the technical aspects discussed. This is particularly evident in the excessive use of acronyms and neologisms that sometimes made the writing cumbersome. But perhaps more importantly, it is sometimes hard to follow the overall narrative of the book, which seems to get lost in the many anecdotes told by the author. However, MacKenzie’s goal of unveiling the mundane daily life of high-frequency trading is achieved, and Trading at the Speed of Light goes well beyond this objective. Beyond shedding light on high-frequency trading, this book offers more broadly a sociology of arbitrage by explaining how this pivotal mechanism of modern financial economics plays out in the everyday lives of economic agents.