What Caused the “Flash Crash”?

By Ankush KhardoriOctober 15, 2020

What Caused the “Flash Crash”?

Flash Crash by Liam Vaughan

A CURIOUS EPISODE in the annals of modern finance came to an anticlimactic conclusion in January, when a federal judge in Chicago sentenced Navinder Sarao, once a UK-based financial trader, to a year of home confinement after he pled guilty to a crime called “spoofing.” The event would have been unremarkable except that prosecutors at the Justice Department had suggested that Sarao was partly to blame for an event that occurred a decade ago, on May 6, 2010, that became known as the “Flash Crash” — when markets in the United States suffered a 30-minute crash and rebound that briefly resulted in the loss of roughly a trillion dollars in value.

“Spoofing,” in the context of finance, is a term that describes a tactic used to manipulate prices — the rapid placement and cancellation of orders on one side of a market in order to benefit an order placed on the other side of the market, by moving the price toward the second order. Suppose, for some reason, you are in the market to buy some gold futures on a financial exchange but you do not want to pay the prevailing market price. You can move the market price down by briefly placing a large sell order for gold futures — your spoof order — which will lead other market participants to believe that supply has increased, which, in turn, will have the effect of pushing the market price down. Once you are able to buy your gold futures at the (previously below-market) price you had wanted, you then cancel the large sell order, and there you have it — you have successfully engaged in spoofing. Sarao’s spoofing took place in the market for futures based on the S&P 500.

Flash Crash by Liam Vaughan — a UK-based financial journalist — presents the story of Sarao’s professional life and his eventual entanglement with the US government and its response to the Flash Crash. (Full disclosure: I used to work at the Justice Department unit that charged Sarao, including on a spoofing prosecution and corporate settlement, but I had no personal involvement with Sarao’s case. I also know several of the people involved with the investigation and prosecution.)

Vaughan’s book presents a story that draws most of its appeal from the enduring mystery of the crash and from Sarao’s strange rise and fall. Sarao amassed a fortune worth tens of millions of dollars and eventually lost it all by becoming a victim of a Ponzi scheme himself. But Flash Crash also poses difficult questions about the priorities of the Justice Department in deciding which white-collar crimes it will pursue. How may an apparent interest in generating publicity lead to investigations of questionable value? Further, was Sarao in fact the victim of a years-long lobbying effort by powerful high-frequency trading firms to have the government pursue market misconduct that affects their profitability, but that is otherwise practically — if not entirely — inconsequential to the public?

¤


Let’s get the easy part out of the way first: if the subject matter sounds remotely interesting, you should read this book. It is briskly written and well paced. It explains complex financial concepts in lucid prose that should be accessible to newcomers, while also providing details that will engage more knowledgeable readers. Vaughan identifies, for instance, perhaps the first-ever reference to the act of spoofing (in a 1719 essay), and traces the origin of the term to a card game from the 1880s (one that “revolved around trickery”).

Remarkably, Vaughan did not have direct access to Sarao, even as it is clear that he became close to his subject — whom he typically refers to by his nickname, “Nav.” (Prosecutors do not like cooperators to speak with third parties while they are completing their work with the government.) Given the extensive details about Sarao’s life that Vaughan describes, Flash Crash is an impressive example of an obscure journalistic genre — the “write-around” profile, in which an author writes about someone without their cooperation. Beyond that, it is a worthy and engaging addition to the genre of book-length narrative financial journalism.

¤


The question that animates the book is a deceptively simple one: what role did Sarao have in causing the Flash Crash? Vaughan is meticulous and fair, but the weight of the evidence he provides points to Sarao having little, if anything, to do with it.

The most authoritative account remains a 104-page report jointly issued by the Commodity Futures Trading Commission and the Securities and Exchange Commission about six months after the event. They concluded, as Vaughan summarizes, that the crash was inadvertently caused by “a huge, clumsy, one-way sell order from an old-school fund [that] arrived at exactly the wrong time, sending an already highly volatile market into meltdown.” Further, “the trading behavior of HFTs” — high-frequency trading firms — “appears to have exacerbated the downward move in prices.”

Things changed, however, in 2012, when a day trader in Chicago — described anonymously in the book as “Mr. X” — noticed a heavy trading imbalance in the period leading up to the Flash Crash that he suspected had been the result of manipulation in the market in which Sarao had been trading. Mr. X concluded that the Flash Crash was (in Vaughan’s words) “at least partially the result” of that manipulation. Mr. X relayed his findings in a whistleblower submission to the CFTC, which eventually linked the trading to Sarao. The agency went on to enlist the involvement of Justice Department attorneys in Washington, DC.

The CFTC and Justice Department lawyers eventually concurred with Mr. X’s analysis, but little is said about how they went about that analysis. In fact, Vaughan writes that the CFTC’s own expert — a finance professor from the University of California, Berkeley — “resist[ed] pressure from the CFTC to tie the day’s events to Sarao” and never actually endorsed the proposition.

¤


Prosecutors reduce complex fact patterns to simple ones. Partly this is a professional necessity — prosecutors tell stories that juries made up of laypeople can grasp and find compelling. Partly it is because almost no white-collar prosecutors (myself included) have any serious academic or professional training in finance or economics. And partly it is because of confirmation bias.

The notion that pretty much everyone else who looked at the cause of the Flash Crash got it wrong is tempting for its plucky, dramatic flair. It may even partially be true: the claims of the government’s report and Mr. X are not necessarily mutually exclusive. A car can spill gasoline on a street only for another to come along and provide a spark that creates a fire. But none of this came through from the government’s eventual indictment of Sarao in 2015.

Sarao’s indictment alleged — in a critical line — that his trading “contributed to the order-book imbalance that the CFTC and SEC have concluded, in a published report, was a cause, among other factors, of the Flash Crash.”

Vaughan writes that this tenuously constructed allegation contained “a formulation that was hard to refute,” but that is not all that the indictment said on the matter. The indictment also alleged that Sarao was spoofing “extensively and with particular intensity” on the day of the Flash Crash, and that it “was particularly intense in the hours leading up to the Flash Crash.” It also described trades Sarao made in the minutes leading up to a 600-point fall in the Dow — most of its loss during the event. The implication was that Sarao’s trading was a meaningful contributor — perhaps the most significant contributor — to the crash.

The allegation in the indictment was all the more curious because it appears to have been unnecessary. Most of the 22 charges in the indictment — which included wire fraud, commodities fraud, and anti-spoofing charges — did not require the Justice Department to establish that Sarao had been responsible for any market losses.

The remaining 10 counts — which charged Sarao with price manipulation and attempted price manipulation — would have required the government to show that Sarao actually caused (or attempted to cause) artificial prices in the market, but including these charges provided no prosecutorial benefit. They did not provide a vehicle for any evidence that would have been unavailable through the other counts, and they would not have increased Sarao’s sentence if he had been convicted at trial. Not only is it unusual to charge crimes with heavier evidentiary burdens than necessary — particularly in complex financial schemes — but the Justice Department advises its prosecutors in multiple ways not to do so.

Taken as a whole, it is hard to avoid the conclusion that the manipulation counts were included for the purpose of drawing attention to the allegation that Sarao had caused the Flash Crash, even as a government expert was refusing to endorse that theory. This is not to say that Sarao did not engage in spoofing. He admitted to it when he pled guilty, but that does not vindicate the far more aggressive claim that Sarao’s misconduct had a role to play in the Flash Crash.

¤


The Justice Department’s claim was widely ridiculed at the time, with one journalist at The Financial Times calling it “nonsense.” Vaughan writes that the prosecutors found this “predictable but irritating” because it supposedly ignored the nuance they had included in the indictment. But in fact, the indictment reads as an attempt by the Justice Department to have it both ways — publicly tying Sarao to the Flash Crash but surrounding it with equivocal language. (Some of the suggestions recounted by Vaughan that the prosecutors were unfairly treated in the press border on the comical, like complaining about the title of the Justice Department’s own press release.)

The passage of time does not appear to have changed anyone’s mind. In 2017, two years after Sarao was indicted, a group of well-regarded academics conducted their own study that corroborated the CFTC/SEC joint analysis and politely concluded that the Justice Department’s claim regarding Sarao’s role was wrong. A review of Vaughan’s book for The Financial Times concluded that whether Sarao “was truly to blame for the 2010 crash will remain a point of debate.” The reviewer for The Wall Street Journal was less forgiving, asserting that “Sarao was clearly not the cause of the flash crash.” None of these observers sought to excuse Sarao’s conduct — to suggest that spoofing is acceptable conduct or that it should not be illegal — but there is a difference between holding someone accountable for criminal misconduct and suggesting, as the government had done, that that misconduct had far more serious consequences than it appeared to.

The book suggests the uncomfortable possibility that the Justice Department and the CFTC knew that they would make news when they claimed that Sarao caused the Flash Crash — excuse me, “contributed to” “a cause, among other factors, of the Flash Crash” — and that at least some of them understood that it might elevate their public profiles. The government attorneys often “joked about who was going to play them in the inevitable movie” — a topic described as “their perennial favorite subject.” And on the night after Sarao eventually pled guilty, the government team “went for dinner and raised a toast to the notorious Flash Crash Trader.”

Perhaps the most glaring evidence that the government enjoyed the publicity it had created is Vaughan’s book itself. It reflects extensive and highly unusual cooperation from government officials. There are reams of inside details that must have come from them — details of internal deliberations, internal meetings, internal monologues, and so on — and that would normally never become public.

One disturbing, surely inadvertent admission that emerges from the government’s cooperation with Vaughan is that the case against Sarao appears to have been flimsy when he was arrested in 2015 and only solidified after the government obtained evidence during the arrest. A prosecutor named Robert Zink who joined the case after the indictment — and who now helps to oversee the 600 lawyers in the department’s Criminal Division — “marveled at his predecessor’s foresight and guts in charging Sarao when so much crucial evidence had emerged after his arrest” (emphasis in original). This is an odd thing to celebrate, because the other way of looking at it is that the government got extremely lucky after it charged a very thin case — a case that, absent that good fortune, could have fallen apart. I knew this prosecutor while I worked at the department, and I have briefly questioned his ascension elsewhere, but the account rings very true, since he was the sort of person who was often incapable of following ideas to their furthest implications.

It is also laughably easy from the details reported in the book to determine that he and almost every government official named in the book must have spoken with Vaughan at length (their fault, not his). The book recounts the childhood backgrounds, for instance, of two of them.

The cumulative result is entertaining. But none of it reflects a desire on the part of the government to proceed cautiously or avoid publicity.

¤


Does it matter if the government unfairly presented Sarao as the cause of the Flash Crash, or if it presented the issue in a way that led the public to believe he was?

Mr. X, the person who started the government’s investigation of Sarao, ridicules the notion — telling Vaughan that it is “comical” to place such emphasis on the question of causation because “markets are very complex systems.” As Vaughan notes, this trader filed a formal whistleblower complaint that provided him with a potential financial stake in any recovery by the CFTC. It is likely that connecting Sarao’s trading to the Flash Crash was a critical feature of this complaint, since most whistleblower complaints go nowhere while this one spurred the government to action. (I do not take issue with the trader’s decision — whistleblower programs exist to incentivize disclosures — but as someone who used to review such complaints to determine whether they merited criminal investigations, they are to be treated cautiously because, all things equal, the prospect of a monetary reward leads people to overstate their factual claims.)

Questions of causation are among the most complex moral and legal questions we encounter. We draw — or at least should draw — a distinction in our minds between the arsonist who starts a fire and the person who inadvertently helps it spread. The law provides myriad ways of thinking about these issues — distinguishing “but-for” causation from “proximate” causation, for instance — but they routinely trip up even the most experienced practitioners.

How different would the world be — would Sarao’s life be — if the government had been silent on the matter of whether or to what extent he “contributed to” “a cause, among other factors, of the Flash Crash”? It is impossible to know for sure, but Sarao’s name would probably not have become synonymous with the Flash Crash, and he likely would not have become the subject of a book (and potential movie). Zink, the prosecutor who patted the government on the back for commencing a questionable prosecution — by his own admission, albeit surely unintentional — would probably not have seen his career advance as it has.

On a matter of greater public concern, it is also debatable whether the Sarao prosecution would have become the precursor for the government’s multi-million-dollar, years-long crackdown on spoofing in financial markets — an effort whose social value, as I have noted elsewhere, is highly debatable. Alleged contributions to Flash Crashes aside, spoofing generally causes losses that are small and diffuse. And as Vaughan’s book makes clear, the “victims” are principally high-frequency trading firms that are not well known to the public. In fact, at the most recent spoofing trial prosecuted by the government, which concluded near the end of September, the government offered testimony on behalf of just two “victims” of the alleged spoofing by the two defendants on trial: one was a representative from Citadel Securities, which made over $2 billion in profit in just the first six months of this year, and the other was a company called Quantlab, which, as The Wall Street Journal has noted, is “one of the world’s most secretive and highly profitable high-frequency trading firms.”

The defendants in that trial were acquitted on more than two-thirds of the charges against them, including the charge that was by far the most serious. The amount of harm caused by the conduct for which they were convicted may literally be in the single-digit thousands of dollars. The department tried to spin this as a victory — and no doubt the defendants would have preferred being acquitted on all of the charges — but it was an embarrassing loss. In fact, although about half a dozen people have pleaded guilty, the government has lost all three of the trials resulting from its anti-spoofing initiative.

Even setting aside this trial record, many people would probably be irate if they knew that the government had been spending millions of dollars and precious man-hours protecting the market positions of extraordinarily profitable firms. Yet even today, the government has persisted in a highly unusual (and highly questionable) years-long public relations campaign to try to generate support for its anti-spoofing campaign — one that has included press availabilities, on-the-record interviews with senior Justice Department officials, and embarrassingly obvious government-friendly leaks, notwithstanding the fact that, as you may have heard, comments about ongoing investigations are supposedly frowned upon.

Somehow, nearly half a decade into this questionable initiative, even those senior government officials struggle to explain why the average person should care about these cases. The head of the Justice Department’s criminal division told The Wall Street Journal early this year that the cases “protect the integrity” of “markets” — a cliché that could be invoked to justify almost every financial fraud prosecution. The head of enforcement at the CFTC has said that spoofing undermines market “liquidity” — another cliché but one that also happens to parrot the line used by HFTs to justify their existence even though “the data [has] suggested that the most profitable of these firms mostly removed liquidity” (emphasis in original).

The ongoing enforcement effort is all the more dubious at a time when financial fraud prosecutions are at their lowest number ever (according to data maintained by Syracuse University). Which also just so happens to be a time when the public is being victimized by internet-based fraud at its highest volume ever, to the tune of over $3.5 billion last year (according to data maintained by the FBI), and when both private and public institutions are providing evidence that corporate misconduct has been on the rise for years.

None of this is to say that spoofing should be legal, or that misconduct that causes small and diffuse losses should be ignored. In fact, while working at the department, I defended the soundness of the legal theory that was used in the most recent case. Stealing a little bit of money from a lot of people is not acceptable. To the contrary, as I have explained elsewhere, some of the most effective fraud schemes rely on just this structure in order to avoid serious scrutiny. But not all harms are equal: a $500 loss to the average American is not the same as a $500 loss to the billionaire who owns Citadel Securities.

Setting criminal enforcement priorities is a complicated affair, and doing so intelligently requires identifying an appropriate equilibrium that properly calibrates the resources that the government can devote to the harm that is being addressed, as well as estimating how significant a criminal enforcement presence is necessary in order to achieve the desired effect. This may seem unsatisfying — and it is — but it is sadly unavoidable, the consequence of the fact that we have extraordinarily limited law enforcement resources relative to the unprecedented level of financial fraud that now exists across all sectors of the global economy. Every million dollars that is spent on a spoofing prosecution is a million dollars that is not being spent on a different financial fraud — the vast, vast majority of which go unprosecuted and unpunished.

¤


All of this brings us, finally, to the questions of how and why it is the government started caring about spoofing in the first place. The provision that specifically outlawed it was part of the Dodd-Frank Act, passed in 2010, but no one seems to know quite how it got put in the bill or why it made it through the typical gauntlet of interested parties to passage.

Vaughan traces it to a list of “Disruptive Trading Practices” that had been prepared by members of the CFTC’s Enforcement Division, but I am skeptical about why it is that spoofing got onto that list, and Vaughan provides plenty of reason to be skeptical about how it came to stay in the bill. He writes that the provision was one that “pleased the HFT community,” and that although market participants typically resist new rules “regardless of how sensible they might be, […] the HFT community got behind the spoofing ban straight away.” For most of these firms — which comprise a lobby that, as Vaughan notes, is “slick, well funded and organized” — “spoofing is a hindrance” because it disrupts the work of their algorithms. The book quotes one trader telling Vaughan, “Imagine a system where the biggest, most powerful players get to tell the regulators and exchanges who to go after based on who is taking money off them. Welcome to the futures market.”

Given the way these dominoes fell — from the HFT-supported ban on spoofing, to the prosecution of Sarao, to the years-long, multimillion-dollar crackdown on spoofing by the Justice Department — some questions arise. For instance, did the lobbying of the HFTs ultimately “cause” Sarao’s arrest and disproportionate ignominy? Did the efforts of the HFTs eventually “cause” multiple government agencies to become unwitting tools in the service of those firms’ parochial, financial self-interests — interests that serve their profitability first and foremost rather than any significant public good?

Perhaps, in the final analysis, it would be more accurate to say that these companies — highly lucrative, private enterprises whose contributions to the public good are questionable — “contributed to” “a cause, among other factors,” of these events.

¤


Ankush Khardori is an attorney and former federal prosecutor who specialized in financial fraud.

LARB Contributor

Ankush Khardori is an attorney and former federal prosecutor who specialized in financial fraud. His writing has been published by The Washington Post, The Wall Street Journal, The New Republic, The New York Review of Books, Slate, and The American Prospect. His website is khardori.com.

Share

LARB Staff Recommendations

Did you know LARB is a reader-supported nonprofit?


LARB publishes daily without a paywall as part of our mission to make rigorous, incisive, and engaging writing on every aspect of literature, culture, and the arts freely accessible to the public. Help us continue this work with your tax-deductible donation today!