Why Michael Lewis got it wrong.

May 2, 2014

“There is systematic corruption in the market. A rigging, a rigging in the market. And it’s the provision to high frequency traders of information that ordinary investors do not have [which is at the core of this form of corruption]”.

This is, in a nutshell, the central claim of Michael Lewis’ new book, Flash Boys. Built around the almost heroic journey of Royal Bank of Canada’s Brad Katsuyama to understand how the interconnected American stock market works, Lewis’ story is now a centerpiece of the debate on HFT. In the smallish world of finance and technology, it is definitively the talk of the town.

As much as it has been praised (particularly by detractors of high frequency trading), Lewis’ book has also become the subject of intense, acerbic and at times quite emotional critiques. There is, indeed, something both provocative and awkward in Lewis’ account. Perhaps it is its frankness. Perhaps it is the heroic narrative of moral entrepreneurship upon which his story is predicated. Perhaps such awkwardness simply derives from factual inconsistencies in Lewis’ account. Is Lewis right, I wonder? Is the stock market rigged, as he argues in Flash Boys and countless interviews thereafter?

Custom indicates that, in addressing these questions, I should be as terse as possible, that I should offer something close to a standard, balanced, boilerplate analysis stating something of the sort: “Lewis got some bits of the story right, others less so.” But let me be provocative and offer an answer that, like Spread Networks’ cable, is a bit more direct: Michael Lewis is wrong, and there are important reasons why his being wrong matters to the social studies of finance.

As far as I can tell, there are three elements of Lewis’ story that misrepresent HFT, markets, and more importantly the problems facing contemporary finance.

The first element is factual. The basic structural setup of Lewis’ book is of a story of ignorance, discovery and moral transformation. In the spring of 2006, Brad Katsuyama observed that the US market was “behaving oddly,” when “suddenly” RBC’s recently purchased electronic trading system didn’t work as he expected it to: “when he pushed a button to complete a trade, the offers would vanish,” writes Lewis. As Katsuyama sought to understand his impossibility to complete trades, he contacted others in the industry, including “big-time hedge funds” that were facing similar difficulties. As Lewis insists, no one knew what was going on. Neither Katsuyama, nor his developers, nor the heads of sophisticated hedge funds understood what had changed in the market. They were ignorant of the transformed structure of the system in which they operated. It was only after working with telecommunications expert Ron Ryan that Katsuyama realized that physical distance mattered once again in finance; he now understood why his orders weren’t being filled (and came to see, in the process, that the market was ‘rigged’ by exchanges in cahoots with high frequency traders).

Note that, central to this narrative, is the claim that no one understood the microstructure of American financial markets in early 2007. This is what separates Lewis’ story from a story of negligence.  Indeed, in a number of interviews, Katsuyama is presented as he who singlehandedly discovered the terrible secret of American finance.  However, this claim of wholesale ignorance seems to contradict my experience and, quite probably, that of many others both within and outwith the industry.

I can think, in particular, of one of the early field experiences that informed my research on the history of the London Stock Exchange. In summer 2007, I was invited to attend a trade conference in London on the future of stock exchanges. At the time, I was a PhD student at the University of Edinburgh; the event mattered because it provided me with access to a key gatekeeper of the UK’s finance and technology community who would become a central informant of my doctoral project. Remember that, at the time, my head was in the past, in how markets changed in 1980s Britain—everything else was archived for the future. One aspect of the meeting, however, was especially salient and caught my attention immediately: more than half of the conference dealt with latency, network configurations, co-location, and competition in global finance. Indeed, a memorable handout provided to conference participants was a copy of a Financial Times article that earlier in the year had hailed the arrival of millisecond trading. For most in the room, the types of questions that Katsuyama was trying to answer were perfectly intelligible. There was, in this sense, no ignorance.

The conference was in London so, naturally, one could imagine that news of a change in market microstructures, and discussions about latency, order types and speed, took some time to reach the US. I am being ironic, of course: among conference participants were Benn Steil, a renowned financial expert and since 1999 member of the Council on Foreign Relations, and Mary Shapiro, then head of the Securities and Exchange Commission. Indeed, the claim that in 2007 no one knew what was going on in American finance seems woefully incorrect: at least one major (and relatively accessible: I was there) conference dealt with the topic. Perhaps there is an agnatological argument to be made, but the point is that, if the FT and SEC knew about millisecond trading and the re-configured structure of global finance, then it was probably old news to the market. Perhaps rather than being shocked at how some firms were moving markets towards faster speeds, we should be a bit more concerned by what seems to have been a grievous lack of information about market structure within investment banks and hedge funds. Then again, the shortsightedness of incumbents tends to be a classical feature of technical change (and, in a different manifestation, had global consequences a few months after the London conference).

But Flash Boys is also erred in its epistemology. Beyond Katsuyama’s story, Lewis pursues the broader argument that key to the problems of the American stock market is the fact that innovation resulted in an unnecessarily complex and opaque system that creates opportunities for catastrophe and malfeasance. Complexity, for Lewis, is dangerous and widespread.

Complexity has long been a scapegoat when dealing with the regulation of technological systems: to say that something is ‘complex’ removes it rhetorically from the domain of what is knowable and casts doubt on our collective ability of control. For Lewis, the hallmark of the market’s complexity is his observation that, in today’s America, when investors submit their orders they know not where they go. Such lack of knowledge permits, for Lewis, the type of front-running-like practices that, as he argues, are derivatives of a rigged market architecture.

What Lewis perhaps forgets is that, in many ways, the architectures of financial markets are far more domesticated, transparent and tractable than those of ostensibly more commonplace and less controversial technological systems. Like the proprietary systems in finance, the internet is also a tangle of cables, routers, intermediaries, semiformal agreements, standards and regulations. Yet unlike the systems deployed in financial markets, and as one telling article notes, something as simple as determining what affects a customer’s broadband speed is largely an impossible task. If anything, developments such as Spread Network’s direct communication between Chicago and New York reduce, rather than increase, complexity: the users of such systems will be better able to manage and know how the system behaves since they will rely on one, rather than a multiplicity of service providers (though at a hefty cost).

The source of this rhetoric of complexity in Lewis’ account may well have two origins. The first, which is admittedly very hypothetical, is the fact that there may too much Liar’s Poker in Flash Boys: when Lewis invokes the image of investors not knowing where their orders went, he invokes, too, a materialistic conception of order flow in markets. There was indeed a time when physical objects would ‘flow’ in financial centers: tickets on the trading floor, and certificates in clearinghouses and banks.  Thirty years of automation later, it would be naïve to think of markets as flows of concrete material packages. Markets today are all about data and information and stock exchanges are closer to IT providers than members clubs. And for these markets, for these spaces of transaction, the traditional metaphor of a specialist on the trading floor no longer holds valid. Think, rather, of computers, algorithms and network topologies.

A second source of this rhetoric may be a deeper, more historically entrenched cultural division within the financial sector. Despite the rise of technology in financial markets, Lewis seems to hint at a conflict between traders and technologists, manifested in the absence of a sort of interactional technological expertise (to paraphrase Harry Collins and Robert Evans) within the more traditional, trading-minded managerial spheres of banks and investment funds. Take the example of Brad Katsuyama. He was not a quant in any way, nor was he trained in computer science or telecommunications engineering. Rather, he followed a very ‘traditional’ career trajectory within Wall Street, having risen through the trading room and into managerial ranks. This is why, perhaps, technology seemed so unintelligible to both him and most of the financial characters in Lewis’ book: their origins were in trading, not engineering. “Thank God,” said a ‘big investor’ when Katsuyama explained market structure to him, “finally there is someone who knows something about high-frequency trading who isn’t an Area 51 guy”. (Back in the 1980s, technologists were ‘those back office people.’ They’ve been upgraded: now they work for Area 51.) Indeed, maybe Brad Katsuyama’s success itself was built on this difference and on an acquired ability to translate between two cultures.

Finally, Flash Boys falls short in history. To say that markets are rigged is almost a claim of conspiracy: actors (in this case, high frequency traders and stock exchanges) colluded to create a system advantageous to them. Yet the recent history of American financial infrastructures would suggest that the origins of today’s market architecture are temporally and institutionally more distributed than what Lewis suggests. Much of the rise of algorithmic and high frequency trading, for instance, traces back to decisions taken in the 1960s and 1970s by US Congress to establish a National Market System and to innovations associated to a new-found technoliberalism in 1990s American finance. Ironically, the same type of moral imperatives that motivate Lewis’ book framed such decisions, namely, that powerful intermediaries (i.e. the NYSE) maintained an opaque and expensive marketplace that was an uneven playing field for ‘ordinary investors’. In one way or another, the quest for equality and fairness led to the forms interconnection and fragmentation upon which HFT grew. Responsibility is distributed, as much upon some putatively opportunistic actors in the present as on the failure of the (weak) American state and its regulator to act in the past.

So why does this matter, particularly to social studies of finance?

At one level, Flash Boys highlights the fact that there remains much scope for researching market change in recent years. There are, for instance, open and exciting questions on the organizational cultures, historical trajectories and overall epistemologies imbricated in market evolution: What are the consequences of automation to knowability? Have market ontologies changed? What type of creatures are contemporary financial markets, and why does this matter? And given that technologies always fail, what does this mean for the future of finance and regulation? Indeed, one of the key contributions of SSF may well be to provide intelligent answers to the type of questions posed by Lewis and the characters in his book: there is clear scope for SSF to be part of a stronger form of public sociology, scrutinizing the mechanisms of markets and their public controversies (as evidenced, for instance, in Daniel and Yuval’s work on the NYSE).

At another level, Lewis’ story reminds us of the importance of dealing with morality in markets. Lewis’ is a moral tale, one in which the questionable practices of some actors (HFTs) are contrasted with the virtuous principles of his story’s hero. This is, perhaps, where most of the awkwardness in Lewis’ tale lies, that is, in having told the story through a dualistic moral narrative of the virtuous and the immoral. What we know as sociologists, however, is that markets are always and ever moral projects: what matters is understanding the dynamics and consequences of the struggles of authority and power played on moral grounds. In effect, the controversy surrounding Flash Boys may not only have unveiled diverging interests within the market: its public reception may, indeed, be more productively seen as reflecting moralized disputes in evaluating the legitimacy of finance in contemporary societies.

More critically, however, is being careful to not transform the debate on HFT into a form of escapism, which I feel is what Lewis has done. Yes: the architecture of American financial markets may well need reconsidering. Yes: equality of access is and will always be an issue. And, yes: Spread Network’s cable seems like the folly of baroque innovation. But let us not forget that this set of practices, which are structurally important in providing liquidity, represents a mere £2 billion in global revenues per year, according to most meaningful estimates. This is the magnitude of HFT, of the tax it levies on trading in what are probably the cheapest and most efficient financial markets in history. In 2014 alone, the bonuses in Barclays—a large, yet single bank—amounted to £2.4 billion. And in 2013, the NY Times reported, disappointingly, that Wall Street bonuses were down to $91.44 billion from a previous high $92.49 billion. Surely, high frequency trading is the least of our collective problems.


12 Responses to “Why Michael Lewis got it wrong.”

    • Cam Grierson Says:

      I read your article and it lead me to believe that Michael Lewis got it right and that JP got it wrong.

      I found that JP’s article address a lot of fringe topics but did not address the material issues in Lewis’ book.

      Lewis is not attacking High Frequency Trading (HFT) or the complexity related to HFT so why talk about something that is not part of the material issue.

      Lewis did say that some participants in the HFT area are using methods that allow them to profit without taking any risks – they are gaming the system as opposed to profiting due to superior skill.

      The idea that it is OK for some participants to have a free ride at the expense of all the other participants is a very corrosive proposition. These few participants are effectively doing “insider trading” – the fact that it is “nanosecond” insider trading does not change the end result.

      Yes Lewis is a good story teller – we would not even be having this discussion if his book had been about the facts only – nobody would have read it or would care about the message.

      Even if Lewis is wrong about some of the background – other people did know about HFT when he said they didn’t and complexity is good not bad – who cares – these points have almost nothing to do with the end goal of his story. Insider trading or cheating should not be allowed no matter how little time it takes or how it is accomplished.

      Cam Grierson

  1. Steve Wunsch Says:

    Excellent analysis! It is amazing how the narrative of discovery, as if no one knew how modern trading works before Katsuyama explained it to us all, was able to keep us transfixed by this tale of supposed moral malfeasance. But as you knew from at least 2007, it was actually old news all along. I said something similar in a three part series of articles on tabbforum (http://tabbforum.com/opinions/the-tradeoff-between-fairness-and-liquidity-is-an-old-story-part-3). One additional point, relevant to the formation of markets as social structures, is that it is not necessary to disclaim the rigging part of their history to understand how regulation has transformed them for the worse. They are now ridiculously cheap and easy to use, as long as you don’t do dumb things like Katsuyama did in his tests. But the flip side of cheap is that they don’t raise much capital anymore. Both results are, as you point out, thanks to the SEC’s National Market System.

    • Juan Pablo Pardo-Guerra Says:

      Thank you very much, Steve! Narratives of discovery and reinvention are always best-sellers (and they have been so for many years). I guess there is something comforting to the reader in knowing that someone ‘out there’ is fixing the ‘problem’. Yet this type of stories tend to be far too simplistic (and, at the end of the day, quite politically charged). Alas, Lewis has chosen to write a very partial and simplistic story of how markets changed.

      Thank you, also, for the link to your articles: I definitively agree with the points you make, as with your observation on regulation. Things would have been much different if, forty years ago, discussions on markets had taken a different route. But markets changed, an so should our ideas of what they are and how they ‘should’ work. (It is, indeed, interesting to see that some of the debates on HFT are almost verbatim copies of debates on specialists and exchanges from the 1970s and 1980s which paved the road for the market’s current architecture).

  2. marthapoon Says:

    Hi JP,

    This post does a fantastic job of laying out the kind of research that SSF can do with regards to market structure. But, I do think it takes the book to task Flash Boys for things it isn’t trying to do.

    Lewis has written a story that made people sit up and listen. The main argument of the book is the ‘character’ of Katsuyama, a brilliant piece of narrative engineering who represents the kind of spirited curiosity Lewis thinks more people should have towards markets. Katsuyama is appealing not because he ‘discovers’ in some absolute sense of the term, but because he discovers from the subjective position a lay reader can recognize. He’s the guy who discovers even though he’s *not* an expert.

    When it comes to engaging with the public, Michael Lewis is clearly getting something really right. So the problem isn’t that he’s wrong; it’s that he’s so incredibly compelling.


    • Juan Pablo Pardo-Guerra Says:

      Thanks Martha!

      The thing is: I still think he’s wrong. There is a very clear claim of corruption at the core of Lewis’ story, which I think is at the very least incorrect and at the very worse dangerous.

      Lewis is a tremendously gifted storyteller. There is no discussing that. If he weren’t, this would be merely one of the stack of books already out there criticizing HFT in some way or another.

      But it is precisely Lewis’ compelling narrative which makes his story problematic: it travels well. And with it, it presents readers to a moralizing claim about experts making the system complex and opaque in order to profit from such complexity. The solution is, perhaps, to reduce these new intermediaries to some sort of copy of what specialists and market-makers once were.

      I just wonder, is this not a rather conservative argument? In opening the black boxes of HFT, does Lewis not simplify and even obviate key elements of contemporary financial markets (and with this, obviate the messy politics that have shaped markets for over a century)? Should stock markets be understandable to a ‘lay’ reader who might identify with Katsuyama’s quest? What matters? ‘Ordinary investors’ (which, you know as well as I do, are but fictions) knowing ‘where there orders are’, or the tight spreads that have been produced by electronic trading? Much to discuss!


  3. Thanks for sharing your review! May I suggest that you try to get an amended version of it published somewhere? (Perhaps you’re already planning to do so.) With a little more introduction, I believe journals could be interested in a review of an influential book like this one. — Manuel

    • Juan Pablo Pardo-Guerra Says:

      Dear Manuel,

      Many thanks for the suggestion–I hadn’t thought of writing a review for a journal, but that is great recommendation.

  4. Henrik Says:

    Even though, I concur with the main points of your analysis, I´d like to draw attention to two different aspects. The first one is that the relatively small amount of money traded (and we still speak of 2 billion dollar) by no means renders Lewis´s account insignificant. Relativity does not seem to be helpful when it comes financial markets – given the moral sensitivity of the topic.

    The second thing is that Michael Lewis, in general, has quite interesting a relationship to the SSF in my opinion. Although inaccurate in some details, “the big short” is way more insightful to the public than say “the credit crisis as a problem … “. So, if one wants SSF to have an impact on the public, maybe the terminology should be reconsidered. Have “market ontologies changed” or is one market at a specific time simply different from another market? Why drawing on an almost metaphysical nomenclature to say something which is common sense? SSF seems to replicate the subject matters´ complexity on the level of wording without necessarily adding any new insights in doing so. Rather, it seems to dress up what we do know anyway in different words. Merton, and following him Weick, used to speak of self-defeating prophecies … SSF comes up with counter-performativity (and falls short of elucidating the need for this coinage).

    So on a more general level, Lewis is not just a good read but embodies a serious contender for the public reception of SSF. After all, sociology should, as you do point out so clearly, correct the notion that markets are rigged in the sense Mr. Lewis implies.

    • Juan Pablo Pardo-Guerra Says:

      Dear Henrik,

      Thanks for your comments. I mainly agree with your points. The size of the industry by no means reduces the significance of its moral and political implications, or of the troubles surrounding financial markets in contemporary societies. However, I do think Lewis’ claim that the market is rigged in favour of HFT places a disproportionate weight on these practices (and is, in a sense, a bit of a moral panic). For instance, the two most public catastrophes in HFT (the flash crash and Knight Securities’ near collapse) had far less public consequences than, say, the 2007/8 crisis.

      Your second point is also great: we have a critical problem in communicating with a broader public. (Words like ontology and epistemology don’t help.) This is a theme that I’d been interested in exploring some time ago (but never had the chance to): what can we learn from public understanding of science when trying to explain financial markets and our research on them? For example, although it seems commonsensical, the claim that markets are different and change through time is not necessarily shared by everyone. You would surely be surprised at how central the idea of the trading floor is in understanding stock markets today (even though trading floors are no longer what makes markets). Metaphors can be quite resilient.


  5. Peter Says:

    The SSF article is a good one. Lewis book is a relatively meaningless
    piece which distracts the public from real issues. Someone please tell me how the old specialist system and today’s NASD market makers is much different w.r.t. to the inside advantage, if not speed of transaction.

    Also, the meaningless comment comes from the the fact that very few, but for a rare few voices in the wilderness like Ron Paul, try to get the public to focus on the massive transfer of wealth from the public to the banks et al. engineered by Fed policy. If people understood that, there would be a grey haired revolt. There should be one.

  6. […] Pardo-Guerra, J.-P. (2014), “Why Michael Lewis got it wrong”, blog post, socfinance, 3rd May [https://socfinance.wordpress.com/2014/05/02/why-michael-lewis-is-wrong]. Thacker, E. (2004), “Networks, Swarms, Multitudes”, CTHEORY, a142a 5th May [net […]

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Connecting to %s

%d bloggers like this: