Flash Boys: Cracking the Money Code

High speed trading was controversial in financial circles before Michael Lewis began writing this book. But now that he has, and it is a bestseller, it has reached the popular consciousness. It is worth describing a bit what is high speed trading, and why it is so controversial.

High speed trading began in earnest when the SEC implemented the ‘Regulation National Market System,’ known as REG NMS, in 2007. Among other things, this allowed for more competition among stock exchanges, and led to the proliferation of stock exchanges within the US. Let us describe the situation before the REG NMS. In essence, there were two types of market orders: a ‘market order’ and a ‘limit order.’ Market orders are popular with ordinary traders trading small amounts of stock, since the important thing is to make the trade and one is not worried about large changes in price in the small amount of time it takes to make a trade, between the time one places the order and the time it is executed. For large institutional traders, such as insurance companies, mutual funds, pension funds, university endowment funds, charities, and sovereign wealth funds, typically large blocks of shares were traded (usually broken up into small pieces), and the risk of moving the price substantially with a large buy or sell order (even when it is broken up into smaller pieces) is larger, so limit orders were (and still are) preferred. The large institutional traders provided liquidity to the market, by providing both demand and supply, depending on whether they were buying or selling, enabling the small traders readily to find counter parties to their desired trades. As such, they were able to earn liquidity profits at the (small) expense of the ordinary traders. These liquidity profits greased the system, giving profits both to the large institutional traders as well as to the exchanges, which cleared the trades. (See for example Jarrow and Protter (forthcoming) for a more detailed explanation and mathematical analysis.)

This system changed dramatically after REG NMS, and this is explained at length, in an entertaining style, in the book under review. The book reads almost like a novel with too many characters, but by giving delightful descriptions of individuals players in the financial markets, Michael Lewis is able to rend fascinating a subject that might otherwise seem quite dry. Before I read this book, I had often idly wondered why so many new stock exchanges had suddenly burst onto the scene, changing the NYSE from a dominant major player to a non dominant although still significant player. At the time there was a bit of a scandal over the executive compensation of Richard Grasso, the Chairman and Chief Executive of the NYSE, when it was revealed that he received $140 million in 2003 and was forced to resign. I realized there was serious money to be made in stock exchanges, and I naïvely assumed that the mushrooming of new exchanges was motivated by these profits. When I became aware of the ‘dark pools’ established in the major banks and investment houses, private exchanges that operated in secret for their more important clients, I assumed this was a profits grab for part of the money made by the exchanges, as well as an attempt to shield these clients from the HFTs. But Michael Lewis gives a more sinister explanation of these exchanges and also of the dark pools, and how they serve the ends of the high frequency traders (hereafter, the HFTs), and the interests of the investment banks, and do not at all serve the interests of the users of the stock exchange, even not the big and important clients of the banks and investment houses. This surprised me, but Lewis makes a convincing case.

The HFTs have computers which operate at enormously fast speeds, on the order of 10⁻⁷ second. The NYSE has located its processing computers not at its traditional headquarters in downtown NY, but in a small town named Mahwah, in New Jersey. Mahwah is 27 miles from Wall Street. If one were to place an order for a trade from Broad Street (the traditional and still touristic site of the NYSE), and the trade instruction were to travel with fibre optic cables in a straight line, at the speed of light, the delay would be 0.000145 seconds. It is frightening to think about this, but it is much too slow for the HFTs. Their strategies, nicely explained in Lewis’ book, depend not only on extraordinary speed, but the most profits go to the fastest trader. As a consequence, the HFTs ‘co-located’ in the same building that houses the NYSE processors, with carefully measured fibre optic cable connections arranged so that no one co-locator has a shorter fibre optic cable than any other. This reduces any speed advantage of any given HFT either to better state of the art hardware, or to better (faster) software, giving lucrative programming opportunities to talented programmers. The question becomes: Why is speed so important?

It is hard to summarize a complicated situation, but in essence, Lewis explains how the new exchanges have created a panoply of new types of orders that are used only by the HFTs. These orders, for example an IOC (immediate or cancel) order, coupled with the great speed of the HFTs, allows the HFTs (or more properly their computers) to view the structure of the limit order book for a very small time into the future, and thus ‘know’ (a probabilistic knowing, not an actual knowing) the direction of the market in selected stocks, infinitesimally into the future. This speed advantage and the strategies Lewis explains, have allowed companies such as the high frequency trading company Virtu Financial, to have a record of only one losing day in five years of activity, as is documented in its IPO filing (Levine 2014). This has led to accusations of wrongdoing by the NY State Attorney General (Alden 2014), and in work with Younes Kchia of Goldman Sachs, we have mathematically modelled this as, in effect, a type of insider trading (Kchia and Protter forthcoming).

The profits of the HFT industry are impressive. Billions of dollars are made each year. Where does this money come from? It comes from the institutional traders, including pension funds and mutual funds, who are slow-moving traders who trade in big blocks of stocks (often broken up in algorithmic trading, which can be very fast, but not in comparison to the HFTs), and essentially skims the cream off the liquidity profits the institutional traders used to make. A critic of the book of Lewis, Mathew Phillips (2014), claims HFTs do not affect ordinary people (who are market traders), he neglects to consider the impact of HFT activities on (for example) the pension funds of ordinary people (or more properly those of us ordinary people who still have pension funds). It also increases costs to insurance companies, for example, and inevitably these costs are passed on to its customers, in large part ordinary people. Within academia, there is no agreement over the value or harm of HFT activity, and one can consult (for example) the research papers in O’Hara and Lopez de Prado (2013) for an alternative view of how HFT traders provide liquidity to the market, considered to be a good thing. This is in direct conflict with the message both of Lewis and Arnuk and Saluzzi (2009), Jarrow and Protter (conditionally accepted for publication), where it is the institutional traders taking positions in the market who provide liquidity to the market. They do this by arriving with shares to sell or demand to buy, rather than the HFTs who end the day with a net zero position in all stocks. The HFTs trade not to take positions, or bring buy or sell orders to the market, but to profit in many somewhat sinister ways, such as ‘slow market arbitrage,’ one of the major themes of the book of Lewis. In slow market arbitrage, the HFTs use the proliferation of exchanges and dark pools, coupled with their extraordinary speed to, in essence, ‘front run’ orders of the large, institutional players.

The book Flash Boys of Michael Lewis is a page turner that is fun to read, where one roots for Brad Katsuyama and his heroic crusade; the book ends with his success not yet certain, but seemingly with positive derivatives. Lewis’ ability to describe people is superb: his description of 9/11 and its effect on Zoran Perkov (an employee of Katsuyama) takes only three pages, but ambushes the reader with its emotional power. An interesting feature is that Lewis resists the cheap temptation to vilify market players, taking the more abstract standpoint (and in my view, correct) that people are simply doing what they must, given the system in its current incarnation. While controversial, the book nevertheless makes an important, and I would argue significant, contribution to our ongoing attempts to understand the US (and increasingly the world) financial system, and to make sense of its more esoteric but potentially catastrophic features, such as – yes – high frequency trading.

Philip Protter
Columbia University, New York, USA
© 2015, Philip Protter

Additional information

Funding

Supported in part by NSF grant DMS-1308483.