Book Summary and Analysis |

Economics, Business and  Finance | Money, Personal Finance and Investments | Science and Technology


A Guide to High-Frequency Trading

with a Summary and Critique of

Michael Lewis’s Flash Boys –

What Is Wall Street Hiding from Investors?

by I.R. Mullin

Copyright©2014 I.R. Mullin. All Rights Reserved. No part of this book may be reproduced or retransmitted in any form or by any means without the written permission of the author.

Should you have any questions, please contact us at


While Michael Lewis’s book Flash Boys: A Wall Street Revolt draws public attention to the issues created by high-frequency trading (HFT), familiarity with HFT-related terminology is essential for professionals and the general public who want to develop a better understanding of the subject. Part III of this book explains this terminology in layman’s language.




Algorithm refers to a mathematical formula used in a computer program to execute trades. Algorithms used by HFT firms are called “high-frequency traders (HFTs)” or “algos.”

Ask price (or offer) refers to the price that a seller is willing to accept for an individual security.




Best execution refers to the formal obligation of a broker-dealer to complete the best execution for a customer.

Bid refers to the price at which a buyer is willing to bid in order to buy a security.




Co-location refers to the practice of an exchange to provide space for the servers of traders, such as HFT firms, in the same data center where the exchange’s servers are located. HFT firms seek co-location because it allows them to access stock prices a fraction of a second before other investors receive the stock price information. HFT firms pay exchanges millions of dollars for the opportunity of co-location. Michael Lewis writes in Flash Boys: A Wall Street Revolt that some stock exchanges have expanded their data centers considerably in order to accommodate co-location. For example, the NYSE Euronext data center in Mahwah, New Jersey is almost nine times larger than the old NYSE building.

Consolidated tape refers to the system that gathers and distributes reports of trades executed on exchanges and over the counter.




Dark pools refer to closed securities trading venues that stay away from the large exchanges. Dark pools are privately owned and operated. Dark pools are inaccessible to the public, and they allow large volumes of securities to be traded anonymously.

Dark pool trading refers to trading of large volumes of securities by institutional investors who choose to remain anonymous throughout the trade. That is, dark pools do not display the trading orders publicly. They inform the market about a trade only after the trade is completed. This is done in order to reduce the risk of price movement that would put the trader at a disadvantage. Presently, dark pool trading makes up a significant percentage of U.S. stock trading volume. 

Decimalization refers to the switch from quoting stock prices in 1/16ths or 1/8ths of a dollar to quoting in pennies or decimals. This transition occurred in the year 2000.




ECN refers to an electronic communications network.

Electronic front-running refers to a practice by HFT firms designed to: (1) offer kickback payments to exchanges in order to get early notice of initial bids and offers placed on exchanges and other trading venues by investors; (2) race those investors to the other exchanges; (3) purchase the looked-for securities at better prices; (4) sell those securities to the initial investors at a higher price.




Flash orders refer to a procedure defined by a trading venue, such as an exchange or ECN, to show buy orders, which are priced at the national best offer, or customer sell orders, which are priced at the national best bid. The order is exposed for a few milliseconds. The trading venue’s subscribers with fast connections can execute the orders at the flash price. If the order is not executed within the specified period of time, then it is cancelled (according to the order-entry firm’s instructions) or the order is routed to other exchanges.

Flash trading refers to a controversial computerized trading practice adopted by some stock exchanges wherein an exchange releases “flash” information about buy and sell orders to subscribers (e.g., HFT firms) for a few fractions of a second before the information about the order is made available to the public. This practice enables flash traders to estimate supply and demand and identify movements in the market before other traders do. HFT firms use this flash information in order to trade ahead of pending orders, and such activity can be considered as front-running. In June 2009, U.S. Senator Charles Schumer urged the SEC to ban flash trading. He argued that flash trading created a two-tiered system that offered preferential treatment to privileged traders and deprived institutional and retail investors of a fair price for their transactions.




High-frequency trading (HFT) refers to a computerized trading practice wherein powerful computers with super-fast connections run complex algorithms (HFTs) to analyze the market and execute a large number of orders at extremely high speeds, within fractions of a second. HFT takes advantage of small price movements and it produces small percentage point gains for each trade. However, HFT firms collect huge profits in the long run because of the high volume of trades executed by their HFTs.




Latency refers to the time that passes from the moment a signal is sent to its receipt. The time that it takes the signal to travel equals the distance travelled divided by the speed of the signal. Lower latency can be achieved with faster speed. HFT firms invest heavily to acquire the fastest computers, software and connections in order to take advantage of low latency. Latency also depends on the distance that the signal has to travel. Light travels at a consistent 186,000 miles per second (186 miles a millisecond). The shorter the distance, the lower latency. In order to reduce latency, HFT firms have their servers placed (co-located) within an exchange’s data center.


Layering refers to waves of large false, multiple orders that are placed by HFT firms with the intention to create the impression that the market for shares of a particular financial product is deep in order to take advantage of the market’s reaction to such orders. Another name for layering is “spoofing.” The SEC used the term “spoofing” when it charged the owner of a New Jersey-based trading firm for manipulating the market through an illegal practice of spoofing.

Liquidity refers to the availability of a security to a market. It measures how quickly a trader can get into and out of a security at the same price level.

Liquidity rebate refers to one of the tactics of a “maker-taker model” adopted by most exchanges in order to promote stock liquidity. This model offers a small rebate paid by the exchange to investors and traders who put in limit orders.  From the point of view of the exchange, such investors and traders contribute to liquidity in the stock. The exchange refers to them as liquidity makers. The amount of each rebate is usually fractions of a cent per share. Liquidity rebates add up to significant amounts for HFT firms that trade millions of shares every day.




Market maker refers to a trader or investor who provides liquidity to the market by buying at the bid and selling at the offer. Market makers, which are registered with trading venues, have certain obligations to the market. Some HFT firms are not registered as market makers; however, they use a market maker trading tactic in order to profit from the spread between the bid and the offer.

Market order refers to an order to buy or sell a financial product, such as shares in a stock, at the current market price.

Matching engine refers to the software algorithm used by an exchange’s trading system to continuously match buy and sell orders. Before the computerization of the market, specialists on the trading floor performed this function. The matching engine runs on the exchange’s computers, which is why HFT firms try to place their own computers as close to the exchange servers as possible. 

Momentum ignition refers to HFT strategy wherein HFTs manipulate the market by identifying and taking advantage of other market participants’ intention to trade. HFTs do so by bidding in front of (raising the price) or offering in front of (reducing the price) slower market participants. Their strategy attempts to prompt a number of other market participants to trade quickly and cause short-term price volatility (i.e., a rapid price move).




Offer refers to an offer to sell, or the price of a security at which a seller is willing to sell the security.





Pinging refers to the HFT tactic of entering small orders (about 100 shares) in order to find out about large hidden orders in exchanges or dark pools. Institutional investors use computer algorithms to break up large orders into much smaller ones in order to reduce the market impact (i.e., price volatility) of large orders. In order to identify the large orders, HFT firms submit bids and offers in 100-share stacks for every listed stock. What happens to these orders will tell the HFT if any large buy-side order is entering the market, and HFTs then front-run the large order and make a nearly risk-free profit at the expense of the institutional investor (e.g., a mutual fund or pension fund). As a result, the investor will get an unfavorable price for its large order.


Point of presence refers to the point at which a trader connects to an exchange.


Predatory trading refers to trading strategies utilized by HFT in order to make practically risk-free profits at the expense of investors. In his book Flash Boys: A Wall Street Revolt, Michael Lewis describes three strategies for predatory trading:

  1. Slow market arbitrage or latency arbitrage in which HFT firms make money on price differences of stocks between different exchanges
  2. Electronic front-running, which involves an HFT firm jumping in front of a large order on an exchange, buying all the shares that the large order intends to buy at various other exchanges (if it is a buy order) or hitting all the bids (if it is a sell order), and then selling them to (or buying them from) the large investor and making money on the price difference
  3. Rebate arbitrage in which HFT is used to collect liquidity rebates paid by exchanges without adding any real liquidity





Quote refers to the current price of a security.  According to regulations, most quotes must be displayed publicly as a part of the consolidated quote stream.

Quote stuffing refers to the practice adopted by HFT firms wherein they place a huge number of buy or sell orders on a security and then immediately cancel the orders. The purpose of quote stuffing is to cause confusion in the market.




Scalping refers to a strategy that allows a trader to make profits on extremely small price changes by entering and exiting a stock within fractions of a second, seconds or minutes.

Security refers to any tradable financial asset. Securities are broadly categorized into debt securities, such as bonds and banknotes; equity securities, such as common stocks; derivative securities, such as swaps, futures and options.


Securities Information Processor (SIP) refers to the technology used to collect quote and trade data from various exchanges, consolidate the data, and constantly broadcast real-time price quotes and trades for all stocks. The SIP calculates the National Best Bid and Offer (NBBO) for all stocks. Because of the huge amount of data it has to process, the SIP has a certain latency period, which is greater than that of the direct feeds supplying market data to HFT firms. The SIP sluggishness makes much of HFT activity possible.

Slow-market arbitrage (or latency arbitrage) refers to a strategy adopted by HFT firms wherein their HFTs detect changes in the price of a stock on one exchange and execute orders on other exchanges before those exchanges update their own bid/offer quotes.

Smart routers refer to technology that is able to determine to which exchanges orders have to be sent.


Spoofing refers to a tactic adopted by HFT firms wherein they place large orders without any intention to execute them. By doing so, HFTs trick investors into buying or selling stock at artificial prices. Another name for spoofing is “layering.” The SEC used the term spoofing when it charged the owner of a New Jersey-based trading firm for manipulating the market through an illegal practice of spoofing. Sanjay Wadhwa, a senior associate director of the SEC’s New York regional office, said, “The fair and efficient functioning of the markets requires that prices of securities reflect genuine supply and demand. Traders who pervert these natural forces by engaging in layering or some other form of manipulative trading invite close scrutiny from the SEC.”

Spread refers to the difference in price between an offer to sell and a bid to buy.

Sub-pennying refers to a tactic used by HFTs wherein they step in front of displayed limits orders, having stocks being propped up by passive limit orders.


Go Back to Part II A Summary and Critique of Lewis’s Flash Boys

Go Back to Table of Contents   


Related content

Dead Wake by Eric Larson

Clinton Cash

Promoted links from around the web

Dead Wake by Eric Larson

Clinton Cash