automated trading system

Quantitative Trading: How to Build Your Own Algorithmic Trading Business
by Ernie Chan
Published 17 Nov 2008

P1: JYS c05 JWBK321-Chan September 24, 2008 13:55 Printer: Yet to come CHAPTER 5 Execution Systems t this point, you should have backtested a good strategy (maybe something like the pair-trading strategy in Example 3.6), picked a brokerage (e.g., Interactive Brokers), and have set up a good operating environment (at first, nothing more than a computer, a high-speed Internet connection, and a real-time newsfeed). You are almost ready to execute your trading strategy—after you have implemented an automated trading system (ATS) to generate and transmit your orders to your brokerage for execution. This chapter is about building such an automated trading system and ways to minimize trading costs and divergence with your expected performance based on your backtests. A WHAT AN AUTOMATED TRADING SYSTEM CAN DO FOR YOU An automated trading system will retrieve up-to-date market data from your brokerage or other data vendors, run a trading algorithm to generate orders, and submit those orders to your brokerage for execution.

…

Even if your brokerage’s API provides an order submission function for your use in an Excel Visual Basic macro, its speed is usually too slow if you have to run this program frequently in order to capture the latest data and generate wave after wave of orders. In this case, one must build a fully automated trading system. Building a Fully Automated Trading System A fully automated trading system (see Figure 5.2) can run the trading algorithm in a loop again and again, constantly scanning the latest prices and generating new waves of orders throughout the trading day. The submission of orders through an API to your brokerage account is automatic, so you would not need to load the trades to a basket trader or spread trader, or even manually run a macro on Real-time data feed Your proprietary desktop C++ program Desktop API from your brokerage Your brokerage account FIGURE 5.2 Fully Automated Trading System P1: JYS c05 JWBK321-Chan September 24, 2008 Execution Systems 13:55 Printer: Yet to come 85 your Excel spreadsheet.

…

See Seasonal trading strategies Capacity, 27, 158 Capital availability, effect on choices, 15 Capital allocation, optimal, 95–103 Capital IQ, 136 Chicago Mercantile Exchange (CME), 16 Clarifi, 35 CNBC Plus, 76 Cointegrating augmented Dickey-Fuller test, 128 Cointegration, 126–133 forming a good cointegrating pair of stocks, 128–130 Compustat, 136 Contagion, financial, 104–105 Correlation, 131 Covariance matrix, 97 CSIdata.com, 37 CRSP.com, 37 D Dark-pool liquidity, 71, 73, 88 Data mining, 121 Databases, historical, 37 Data-snooping bias, 25–27, 52–60, 91 out-of-sample testing, 53–55 sample size, 53 sensitivity analysis, 60 and underperformance of live trading, 91 Decimalization of stock prices, 91, 120 Printer: Yet to come INDEX Deleveraging, 152 Despair, 110 Disasters, physical or natural, 108 Discovery (Alphacet), 35, 36, 55, 85, 122–126 charting application, 125 Dollar-neutral portfolio, 43–44 Dow Jones, 36, 75 Drawdown, 20, 21–22, 43, 95 maximum, 21 calculating, 48–50 maximum duration, 21 calculating, 48–50 DTN.com, 37 Dynamic data exchange (DDE) link, 80, 81–82, 83, 84, 85 E ECHOtrade, 70 Econometrics toolbox, 168 The Economist, 10 Elite Trader, 10, 74 Elliott wave theory, 116 E-mini S&P 500 future, 16 Endowment effect, 108–109 Equity curve, 20 Excel, 3, 21, 32, 51, 163 dynamic data exchange (DDE) link to, 80, 81–82, 83, 84, 85 using in automated trading systems, 80, 81, 83, 84, 85 using to avoid look-ahead bias, 51 using to calculate maximum drawdown and maximum drawdown duration, 48 using to calculate Sharpe ratio for long-only strategies, 45–46, 47 P1: JYS ind JWBK321-Chan October 2, 2008 14:7 Printer: Yet to come 177 Index Execution systems, 79–94 automated trading system, advantages of, 79–87 fully automated trading system, building a, 84–87 semiautomated trading system, building a, 81–84 paper trading, testing your system by, 89–90 performance, divergence from expectations, 90–92 transaction costs, minimizing, 87–88 Exit strategy, 140–143 F Factor exposure, 134 Factor models, 133–139 principal component analysis as an example of, 136–139 Factor return, 134 FactSet, 35, 36 Fama-French Three-Factor model, 134–135, 153 Financial web sites and blogs, 10 G GainCapital.com, 37 GARCH toolbox, 168 Gasoline futures, seasonal trade in, 148–151 Gaussian probability distributions, 96, 105 derivation of Kelly formula in, 112–113 Generalized autoregressive conditional heteroskedasticity (GARCH) model, 120 Genesis Securities, 70, 73, 82 Global Alpha fund (Goldman Sachs), 104 Greed, 110–111 H “Half-Kelly” betting, 98, 105–106 High-frequency trading strategies, 151–153 transaction costs, importance of in testing, 152 High-leverage versus high-beta portfolio, 153–154 High watermark, 21, 48 Historical databases errors in, 117 finding and using, 36–43 high and low data, use of, 42–43 split and dividend-adjusted data, 36–40 survivorship bias, 40–42 HQuotes.com, 37, 81 Hulbert, Mark (New York Times), 10 I Information ratio.

Trading and Exchanges: Market Microstructure for Practitioners
by Larry Harris
Published 2 Jan 2003

Electronic trading systems are cheaper to operate because they do not require floor brokers and sales brokers. 27.8 SUMMARY Floor-based trading systems and automated trading systems have different strengths and weaknesses. Consequently, they appeal to different clienteles. It is unlikely that one market structure will dominate all trading. Table 27-1 summarizes strengths and weaknesses of floor-based and automated trading systems. Fully automated systems are very fast and generally cheap to use and operate. These characteristics ensure that active markets and markets that serve small traders will use automated trading systems extensively. In the U.S. equities markets, Bernard L. Madoff Investment Securities, Knight Capital Markets, and other dealers who offer automated execution systems provide excellent service to high volumes of small traders.

…

• The proponents of CISC and RISC microprocessors compete with each other in the marketplace. How does their competition differ from the competition among markets for order flow? 27 Floor Versus Automated Trading Systems Advances in communications and computing technologies now allow exchanges to completely automate their trading systems. Many exchanges have done so, and many brokers, ECNs, and dealers have created automated trading systems. Despite these developments, many of the most liquid exchanges in the world still employ floor-based trading systems. The New York Stock Exchange, the Chicago Board of Trade, the Chicago Mercantile Exchange, the New York Mercantile Exchange, and almost all U.S. options exchanges primarily use floor-based trading systems.

…

The most successful electronic competitors of the NYSE have been third market dealers, like Bernard L. Madoff Investment Securities and Knight Capital Markets. Their automated trading systems provide very quick service primarily to retail traders represented by discount brokers. * * * ▶ The Bangladeshi Stock Exchange and the New York Stock Exchange In 1999, the Bangladeshi Stock Exchange replaced its trading floor with an automated trading system. At the same time, the New York Stock Exchange considered where to build a new trading floor. The continued commitment of the New York Stock Exchange to its trading floor may be its most important decision at the turn of the millennium.

Handbook of Modeling High-Frequency Data in Finance
by Frederi G. Viens , Maria C. Mariani and Ionut Florescu
Published 20 Dec 2011

The capacity to improve the forecast of earnings surprises and abnormal return using a mixture of well-known economic indicators with social network variables also enriches the debate between the modern finance theory and behavioral finance to show how behavioral patterns can be recognized with a rigorous method of analysis and forecast. 3.4.2 ALGORITHMIC TRADING The transformation of the major stock exchanges into electronic financial markets has encouraged the development of automated trading systems in order to process large amounts of information and make instantaneous investment decisions. Automated trading systems have a long tradition on classical artificial intelligence approaches such as expert systems, fuzzy logic, neural networks, and genetic algorithms. Trippi and Turban (1990, 1993), Trippi and Lee (2000), Deboeck (1994), and Chorafas (1994) have reviewed these early systems.

…

Trippi and Turban (1990, 1993), Trippi and Lee (2000), Deboeck (1994), and Chorafas (1994) have reviewed these early systems. Goonatilake and Treleaven (1995) survey an application of the above methods to automated trading and several other business problems such as credit risk, direct marketing, fraud detection, and price forecasting. Automated trading systems include a backtest or simulation module. In this respect, agent-based models could be useful to explore new ideas without risking any money.11 The Santa Fe stock market model (Arthur et al., l997; LeBarone et al., l998; LeBaron, 2001) has inspired many other agent-based financial market models such as Ehrentreich (2002)’s, which is based on the Grossman and Stiglitz (1980) model.

…

LeBaron (1998) applied bootstrapping to capture arbitrage opportunities in the foreign exchange market and then used a neural network where its network architecture was determined through an evolutionary process. Finally, Towers and Burgess (2000) used principal components to capture arbitrage opportunities. Creamer and Freund (2007, 2010a) follow the tradition of the papers in this section that use machine learning algorithms to find profitable trading strategies and also build completely automated trading systems. The authors use very well-known technical indicators such as moving averages or Bollinger bands. Therefore, the capacity to anticipate unexpected market movements is reduced because many other traders are expected to be trying to profit from the same indicators. However, the authors reduce this effect because the algorithms try to discover new trading rules using Logitboost instead of following the trading rules suggested by each indicator.

The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution
by Gregory Zuckerman
Published 5 Nov 2019

Fellow professors were polite enough not to share their criticism and skepticism, at least within earshot. But Berlekamp knew what they were thinking. “Colleagues avoided or evaded commenting,” he says. Simons didn’t care about the doubters; the gains reinforced his conviction that an automated trading system could beat the market. “There’s a real opportunity here,” he told Berlekamp, his enthusiasm growing. Medallion scored a gain of 55.9 percent in 1990, a dramatic improvement on its 4 percent loss the previous year. The profits were especially impressive because they were over and above the hefty fees charged by the fund, which amounted to 5 percent* of all assets managed and 20 percent of all gains generated by the fund.

…

Eventually, Brown and Mercer developed an elaborate stock-trading system that featured a half million lines of code, compared to tens of thousands of lines in Frey’s old system. The new system incorporated all necessary restrictions and requirements; in many ways, it was just the kind of automated trading system Simons had dreamed of years earlier. Because the Nova fund’s stock trades were now less sensitive to the market’s fluctuations, it began holding on to shares a bit longer, two days or so, on average. Crucially, Brown and Mercer retained the prediction model Frey had developed from his Morgan Stanley experience.

…

RenTec has built a machine to model this interconnectedness, track its behavior over time, and bet on when prices seem out of whack according to these models.” Outsiders didn’t quite get it, but the real key was the firm’s engineering—how it put all those factors and forces together in an automated trading system. The firm bought a certain number of stocks with positive signals, often a combination of more granular individual signals, and shorted, or bet against, stocks with negative signals, moves determined by thousands of lines of source code. “There is no individual bet we make that we can explain by saying we think one stock is going to go up or another down,” a senior staffer says.

Traders at Work: How the World's Most Successful Traders Make Their Living in the Markets
by Tim Bourquin and Nicholas Mango
Published 26 Dec 2012

I plan on being the guy that walks around the ottoman next time. CHAPTER 13 Charles German While his career began more than twenty years ago on the floor of the Chicago Board of Trade, Charles German explains that his adoption of a trend-following strategy in 2005 was what ultimately led him to reach new heights in trading. He now uses an automated trading system that combs the markets for setups, places entry and stop orders, and executes with little or no human interaction required. While fully automated trading may be a dream for many traders, it was only through thousands of hours of meticulous backtesting and some costly, difficult lessons learned along the way that German made it a reality.

…

Menaker: That’s a really good question, and I’ve personally gone back and forth on that myself over the years—paying attention to the news or not paying attention to the news—and I think people can succeed going both ways. What I will tell you, though, is that there are hedge funds that are scraping social media, like Twitter and even Facebook, and they are scraping CNBC and Bloomberg for keywords and sentiment. Then they will apply that to an automated trading system. That’s happening more and more. I was in Japan in June for a hedge fund conference, and I met a Tokyo University professor there who’s a financial engineer. He showed me how he and some other guys were starting up a hedge fund using this model that I just explained, and they thought they were the first to do this.

…

I Index A Average directional index (ADX) Average true range (ATR) B Baiynd, Anne-Marie course offered daily and weekly time frames double-top action Elliott wave Fibonacci future market longer-term investment market internal mathematician moving average multiple time frames new trader paper trading paper trading account real money trading recruiting role retail trading Simple moving average (SMA) stochastic momentum indicator SUCCESS Magazine seminar, technical trading support and resistance swing trade technical analysis technical indicator technical trader trading day Berger, Serge Advance/Decline (A/D) line Apple Bloomberg terminal currency futures daily chart day-in and day-out economic data in equities equities, equity options, and futures extreme candles favorable intraday financial analyst full gap gap trades half gap liquidity injections longer-term positions macro view mean-reversion trade mind off the markets momentum oscillator opening gap positions reversal candle S&P 500 E-mini futures seasonal factors slow stochastics swing positions swing trade technical analysis technical tools time frame trade duration trade futures trading environment trading methodology trading opportunity trend reversal US equity indices watch list Booker, Rob Brandt, Peter algorithmic trading bid/offer spreads candlestick charts chart trader classical charting definition patterns principles closing price charts commodities floor trader cookie cutter approach corn spread corn trader currency futures Diary of a Professional Commodity Trader electronic trading entry point futures trader head-and-shoulders patterns high bar charts intraday charts longer-term time frames longer-term trades long-term charts margin-to-capital ratio margin-to-equity ratio market signaling niche—money management pit trading positions risk management Russell trade scale out short-term charts standard stop-loss swing trader Technical Analysis of Stock Trends trend line volume/open interest profile weekly chart C, D California Institute of Technology Carter, John best trade big volume Bollinger Bands breakeven career day trader dot-com crash down payment financial analyst first trading day five-year learning curve fundamental factors garden-variety trade great traders overnight guaranteed income industry trends Keltner Channels loan documentation Mastering the Trade money build up money management OEX trade options professional trader resistance point retail traders shorter-term trading SimplerOptions.com six-figure income smaller traders success measurement swing trades teaching technical analysis technical approach tech stocks The Disciplined Trader the squeeze time frames time spent TradeTheMarkets.com trading industry trading lifestyle trading philosophy trading place trading slumps volatile markets Certified Risk Manager (CRM) Chicago Board of Trade Chicago Mercantile Exchange (CME) City Slickers movie Commodity Trading Advisor (CTA) Currency Strength Index E E-mini S&P futures (ES) European Central Bank (ECB) Exchange-traded funds (ETFs) Exponential moving average (EMA) F Floyd, Gordon & Partners (FGP) Foster, Alex arbitrary profit targets assignment bear market best indicator buy-and-hold approach client vs. own account contract size economic reports ideal trade long-term trend Monsanto and JPMorgan Chase moving averages news following open positions position size price point profitable trades profit targets S&P 500 shorter-term crossover shorter-term moving averages technical indicators time frame trading options trend follower Williams %R G GAIN Capital Asset Management Gartman, Dennis German, Charles ATR automated trading system backtesting daily chart future market green trade independent trader market portfolio mentors money management moving average price action risk management rule-based approach scaling out screen-based trading software stand-alone type of program system rule trend following definition strategies tools Gordon, Todd Aspen Trading Group Australian Dollar average winner and average loser bank research Blue Chips movie chat room CNBC correlation analysis currency markets currency trade day trading decision making Elliott Wave analysis count methodology E-Trade account Fast Money Floyd, Gordon & Partners (FGP) FOREX.com, senior technical strategist Forex trading GAIN Capital Asset Management, senior trader global market analysis hedge fund initial amount, full time trading investment banks leverage magic methodology market information market makers market volatility Money in Motion moving averages NASDAQ new trader NYSE stocks NYSE trading strategy personal account research reports research time S&P 500 futures schooling share size shortcuts short-term momentum trading short-term traders SOES bandit sports analogies stock-picking service stop loss Strategy of the Day technical analysis technical charts trader quality trading jobs trading style trend lines H, I, J Hemminger, Patrick agricultural futures trading agricultural pairs trading Brent curve calendar spread commodities core position crude curve economic releases E-Mini S&P vs.

Trading at the Speed of Light: How Ultrafast Algorithms Are Transforming Financial Markets
by Donald MacKenzie
Published 24 May 2021

To human eyes, trading on Island appeared instantaneous. FIGURE 1.1. 50 Broad Street. Author’s photograph. Just as consequential as Island’s speed was that machines started to trade on it. There had been previous efforts to automate trading, but often they had not gone smoothly. It could be difficult for an automated trading system to interact seamlessly with exchanges’ systems, which in the 1980s and 1990s were usually designed on the assumption that traders were human beings, not machines. Indeed, those who ran exchanges’ early electronic trading systems often protected their human users from “unfair” automated competition by prohibiting the direct connection of computers to them.

…

Whitcomb, who continued to teach at Rutgers and live in New York, designed the relatively simple mathematical model described in the next four paragraphs, faxing instructions and formulas to Charleston to be turned into code by Hawkes’s programmers. Around the equation that was the core of the model, Whitcomb designed and the programmers coded the components of what was eventually to become a full automated-trading system. It consisted of a module to process incoming market data; a pricing module that implemented Whitcomb’s equation; a module that tracked the system’s accumulated trading position in each stock and adjusted its trading strategy accordingly; a decision module that calculated how best to trade based on the existing trading position and the pricing module’s predictions; a module that dispatched the resultant orders and if necessary canceled existing orders; a module that calculated in real time the profits or losses being made; and so on—eventually requiring in total some 80,000 lines of code (interviewee BT).

…

One, apparently the more common, is simply human error (sometimes called “fat finger” error) in entering into execution algorithms the necessary parameters of price and, especially, total size of order. Rarer, but more relevant to this book, is the second cause: serious technical malfunctions of an HFT or other automated trading system. In chapter 3, we saw how a problem of this kind—the result of a mistake in a single character in a program—caused a financial loss to Automated Trading Desk that could have been catastrophic had the human trader involved not noticed it quickly and switched off the automated system involved.

Wall Street Meat
by Andy Kessler
Published 17 Mar 2003

Some of it had to do with those over-the-counter traders not answering their phones. Wall Street got sued for not answering phones and the SEC insisted the Street put in a system known as Small Order Execution System, SOES. This automated execution of small orders would lead to day traders and would eventually lead to automated trading systems known as ECNs. These ECN trading systems would represent over half of overthe-counter trades by 2001, with commissions a hundredth of what they had been in 1987. It changed the way Wall Street gets paid. Commissions were toast. Banking fees would replace commissions, and eventually kill research in the process. · · · For the rest of the year, everyone was in shock.

…

I’m not so sure. 246 Index Activision, 165 Alexander, Margo, 12, 22, 27–28, 38, 60, 81, 85, 86 offsite analyst meetings, 51, 77–78 Alexander, Pam, 184, 218 Alexander Ogilvy, 184 “All American Research Analyst poll,” 25, 48 Alliance Capital, 224 Alliance Management, 43, 47 allocating capital, 90 Ally, Steve, 30, 67, 141 alternate pay phone companies, 37–38 Amazon.com, 174–75, 181, 186–87 AMD, 45, 144 American Electronics Association, 62 American Superconductor, 137 America Online, 105, 156, 178, 218 Amerindo, 168 analyst(s), 8, 234–40 banking, 107–8 basics of, 24–26 boutique, 109 conference calls and, 37–38 after crash of October 1987, 72 “dialing for dollars” and, 47 industry immersion and, 27 institutional clients and, 23 investment bankers and, 241–42 press coverage and, 48 ranking, by Institutional Investor, 25, 46–48 reputation and, 231, 237–41, 244 small cap, 148 types of, 32 visits to accounts by, 47 written reports and, 47–48 Apple Computer, 16–17, 158 Armstrong, Michael, 216 Arrowwood, 51, 77 Ashton-Tate, 82–83 athletes, on Wall Street, 67–68 Index Atlantic Crossing, 213 AT&T, 7, 33, 59, 212, 216–17 automated trading systems, 72 Avid, 162–63 ax in a stock, 35, 112 ax syndrome, 209–18 bankers, technology, 137 banking analysts, 107–8 banking fees, 90 Barlage, Jim, 17 Barnes and Noble Booksellers, 174 Be, Inc., 206 Beard, Anson, 89, 109 Bell Labs, 7 Berens, Rod, 84–87 Berkowitz, Jeff, 202–3 Bezos, Jeff, 174 Biggs, Barton, 24–25, 92, 123–24, 125, 126, 127, 129, 145, 152–53 Blodget, Henry, 181–85, 214–16, 218, 225–26, 231 Blum, Scott, 207 Boesky, Ivan, 56 Bogle, John, 172 bonus pool, 90 Boston Company, 103 Boucher, David, 141–42, 243 boutique analyst, 109 Boutros, George, 170 Brady, Bill, 139–40, 157, 170, 223 Bright Lights, Big City (McInerney), 39–40 248 Broadcast.com, 177–78 Brooke, Paul, 129, 143 bulge bracket firms, 108, 221 bull market(s), 50–69 takeovers, buyouts and, 53 Burroughs Corporation, 56 Business Week, 216 Buy.com, 180, 207–8 “buying it off the box,” 196–97 buy-side firms, 25 Callahan, Dennis, 73, 150, 215 Cantor Fitzgerald, 67 capital, allocating, 90 Carroll, Jim, 72 Carroll, Paul, 67 Case, Steve, 156 Cashin, Art, 42 C-Cube Microsystem, 165 CDMA, 204–5 Chinese Wall, 94, 216, 221 chip industry, 26–27 CIBC Oppenheimer, 182 Cirrus Logic, 93 Cisco, 102, 105, 126 Citigroup, 216 Citron, Jeff, 197–98, 199 Clark, Jim, 156, 165–67 Clark, Mayree, 226 CMGI, 168 CNBC, 64, 181 Colonna, Jerry, 203–4 commissions, 72, 90 compensation, on Wall Street, 90 Index Compuserve, 36 conference calls, 37 conferences, 119 Contel, 39 convertible bonds, 91 Cordial, Steve, 126 Cornell, Robert (Bob), 5–16, 20, 26–30, 72, 79, 85, 155, 167 Cowan, Ollie, 31, 53 Cramer, Jim, 182, 183, 200, 202–3 crash of October 1987, 71 creative accounting, 219 CS First Boston, 1, 179–80, 190, 223 Cuban, Mark, 178 Cuhney, Adam, 79–80 Curley, Jack, 129–30, 140, 156 Cushman, Jay, 129–30, 132 Dale, Peter, 85, 92, 96, 108, 110 Data Resource Inc.

Learn Algorithmic Trading
by Sebastien Donadio
Published 7 Nov 2019

He has taught various computer science courses for the past ten years in the University of Chicago, NYU and Columbia University. His main passion is technology but he is also a scuba diving instructor and an experienced rock-climber. Sourav Ghosh has worked in several proprietary high-frequency algorithmic trading firms over the last decade. He has built and deployed extremely low latency, high throughput automated trading systems for trading exchanges around the world, across multiple asset classes. He specializes in statistical arbitrage market-making, and pairs trading strategies for the most liquid global futures contracts. He works as a Senior Quantitative Developer at a trading firm in Chicago. He holds a Masters in Computer Science from the University of Southern California.

…

They are extremely fast at reacting to market data, and they don't get distracted or make mistakes (unless they were programmed incorrectly, which is a software bug and not a drawback of computers themselves). They don't have emotions, so don't deviate from what they are programmed to do. All of these advantages make computerized automated trading systems extremely profitable when done right, which is where algorithmic trading starts. Evolution of algorithmic trading – from rule-based to AI Let's take a simple example of a trend-following strategy and see how that has evolved from a manual approach all the way to a fully automated algorithmic trading strategy.

High-Frequency Trading: A Practical Guide to Algorithmic Strategies and Trading Systems
by Irene Aldridge
Published 1 Dec 2009

Properly programmed computer systems typically outperform human traders in these “mission-critical” trading tasks, particularly under treacherous market conditions—see Aldridge (2009), for example. As a result, computer trading systems are rapidly replacing traditional human traders on trading desks around the world. The development of a fully automated trading system follows a path similar to that of the standard software development process. The typical life cycle of a development process is illustrated in Figure 16.1. A sound development process normally consists of the following five phases: 1. Planning 2. Analysis 3. Design 4. Implementation 5. Maintenance Planning Analysis Maintenance Design Implementation FIGURE 16.1 Typical development cycle of a trading system. 

…

I 325 Index Accounting services, importance of, 26 Accuracy curves, back-testing, 228–229 Admati, A., 277 Administrative orders, 70 Aggarwal, V., 181 Ahn, H., 67 Äijö, J., 183 Aite Group, 18–19 Ajayi, R.A., 181 Alam, Zinat Shaila, 132, 274, 277–278 Aldridge, Irene, 13–14, 19, 214–215, 222–231 Alexakis, Panayotis, 88 Alexander, Carol, 89, 216–217 Algorithmic trading, 15, 16–19, 22, 23–24 distinguished from high-frequency trading, 16 execution strategies, 16–17, 273–274 portfolio optimization, 213–217 trading signals, 16–17 All or none (AON) orders, 69 Almeida, Alvaro, 168 Almgren, R., 274, 275, 295 AMEX, 9 Amihud, Y., 37–38, 134, 192, 195, 264 Amoaku-Adu, Ben, 192 Analysis stage, of automated system development, 234–235 Anand, Amber, 158–159 Andersen, T.G., 106, 109, 176–178 Andritzky, J.R., 183 Ang, A., 208–209 Angel, J., 133 Anonymous orders, 69–70 Apergis, Nicholas, 88 Arca Options, 9 ARCH specification, 88 Asset allocation, portfolio optimization, 213–217 Asymmetric correlation, portfolio optimization, 208–209 Asymmetric information, measures of, 146–148 Augmented Dickey Fuller (ADF) test, 98 Autocorrelation, distribution of returns and, 94–96 Automated liquidity provision, 4 Automated Trading Desk, LLC (ATD), 12 Automated trading systems, implementation, 233–249 model development life cycle, 234–236 pitfalls, 243–246 steps, 236–243 testing, 246–249 Autoregression-based tests, 86 Autoregressive (AR) estimation models, 98–99 Autoregressive analysis, event arbitrage, 167–168 Autoregressive moving average (ARMA) models, 98, 101, 106 Avellaneda, Marco, 138–139 Average annual return, 49–51 327 328 Bachelier, Louis, 80 Back-testing, 28, 219–231 of automated systems, 233 directional forecasts, 220, 222–231 point forecasts, 220–222 risk measurement and, 255, 268 Bae, Kee-Hong, 67, 68 Bagehot, W., 151 Bailey, W., 183 Balduzzi, P., 182 Bangia, A., 263 Bank for International Settlements (BIS), 43–44 BIS Triennial Surveys, 44 Bannister, G.J., 183 Barclay, M.J., 277 Basel Committee on Banking Supervision, 251, 253, 265 Bayesian approach, estimation errors, 209–211 Bayesian error-correction framework, portfolio optimization, 213–214 Bayesian learning, 152–155 Becker, Kent G., 183 Benchmarking, 57–58 post-trade performance analysis, 296–298 Berber, A., 142 Bernanke, Ben S., 180 Bertsimas, D., 274 Bervas, Arnaud, 38, 263, 264 Best, M.J., 209 Bhaduri, R., 270 Biais, Bruno, 12, 67, 160, 163 Bid-ask bounce, tick data and, 120–121 Bid-ask spread: interest rate futures, 40–41 inventory trading, 133, 134–139 limit orders, 67–68 market microstructure trading, information models, 146–147, 149–157 post-trade analysis of, 288 tick data and, 118–120 Bigan, I., 183 Bisiere, Christophe, 12 INDEX BIS Triennial Surveys, 44 Black, Fisher, 193, 212 Bloomfield, R., 133 Bollerslev T., 106, 176–178 Bollinger Bands, 185 Bond markets, 40–42 Boscaljon, Brian L., 174 Boston Options Exchange (BOX), 9 Bowman, R., 174 Boyd, John H., 180 Bredin, Don, 184 Brennan, M.J., 147, 192, 195 Brock, W.A., 13 Broker commissions, post-trade analysis of, 285, 287 Broker-dealers, 10–13, 25 Brooks, C., 55 Brown, Stephen J., 59 Burke, G., 56 Burke ratio, 53t, 56 Business cycle, of high-frequency trading business, 26–27 Caglio, C., 142 Calmar ratio, 53t, 56 Cancel orders, 70 Cao, C., 131, 139, 142 Capital asset pricing model (CAPM), market-neutral arbitrage, 192–195 Capitalization, of high-frequency trading business, 34–35 Capital markets, twentieth-century structure of, 10–13 Capital turnover, 21 Carpenter, J., 253 Carry rate, avoiding overnight, 2, 16, 21–22 Cash interest rates, 40 Caudill, M., 113 Causal modeling, for risk measurement, 254 Chaboud, Alain P., 191 Chakravarty, Sugato, 158–159, 277 Challe, Edouard, 189 Chan, K., 67 Chan, L.K.C., 180, 289, 295 Index Chen, J., 208–209 Chicago Board Options Exchange (CBOE), 9 Chicago Mercantile Exchange (CME), 9, 198 Choi, B.S., 98 Chordia, T., 192, 195, 279 Chriss, N., 274, 275, 295 Chung, K., 67–68 Citadel, 13 Clearing, broker-dealers and, 25 CME Group, 41 Cohen, K., 130 Co-integration, 101–102 Co-integration-based tests, 89 Coleman, M., 89 Collateralized debt obligations (CDOs), 263 Commercial clients, 10 Commodities.

Who Owns the Future?
by Jaron Lanier
Published 6 May 2013

The networking of finance occurred independently and in advance of the rise of the familiar Internet. There were different technical protocols over different infrastructure, though similar principles applied. Some of the early, dimly remembered steps toward digitally networked finance included: 1987’s Black Monday (a market anomaly caused by automated trading systems), Long-Term Capital, and Enron. I will not recount these stories here, but those readers who are not familiar with them would do well to read up on these rehearsals of our current global troubles. In all these cases there was a high-tech network scheme at play that seemed to concentrate wealth while at the same time causing volatility and trauma for ordinary people, particularly taxpayers who often ended up paying for a bailout.

…

Brian, 169n artificial hearts, 157–58 artificial intelligence (AI), 23, 61, 94, 95, 114, 116, 136, 138n, 147, 155, 157, 178, 191, 192–93, 325, 330, 354, 359n artificial memory, 35 art market, 108 Art of the Long View, The (Schwartz), 214 ashrams, 213 assets, 31, 60 “As We May Think” (Bush), 221n asymmetry, 54–55, 61–66, 118, 188, 203, 246–48, 285–88, 291–92, 310 Athens, 22–25 atomic bomb, 127 “attractor nightmare,” 48 auctions, 170, 286 aulos, 23n austerity, 96, 115, 125, 151, 152, 204, 208 authenticity, 128–32, 137 authors, 62n automata, 11, 12, 17, 23, 42, 55, 85–86, 90–92, 97–100, 111, 129, 135–36, 155, 157, 162, 260, 261, 269, 296n, 342, 359–60 automated services, 62, 63, 64, 147–48 automated trading systems, 74–78, 115 automation, 7, 85, 123–24, 192, 234, 259, 261, 343 automobiles, 43, 86, 90–92, 98, 118–19, 125n, 302, 311, 314, 343, 367 avatar cameras, 265 avatars, 89n, 265, 283–85 baby boomers, 97–100, 339, 346 bailouts, financial, 45, 52, 60, 74–75, 82 Baird-Murray, Kathleen, 200n “Ballad of John Henry, The,” 134–35 bandwidth, 171–72 banking, 32–33, 42, 43, 69, 76–78, 151–52, 251, 269n, 289, 345–46 bankruptcy, 2, 89, 251 bargains, 64–65, 95–96 Barlow, John Perry, 353 Barnes & Noble, 62n, 182 barter system, 20, 57 Battlestar Galactica, 137, 138n “beach fantasy,” 12–13, 18, 236–37, 331, 366–67 Beatles, 211, 212, 213 behavior models, 32, 121, 131, 173–74, 286–87 behavior modification, 173–74 Belarus, 136 belief systems, 139–40 Bell, Gordon, 313 bell curve distribution, 39, 39–45, 204, 208, 262, 291–93 Bell Labs, 94 Bentham, Jeremy, 308n Berners-Lee, Tim, 230 Bezos, Jeff, 352 big business, 265–67, 297–98 big data, 107–40, 150, 151–52, 155, 179, 189, 191–92, 202–4, 265–66, 297–98, 305, 346, 366, 367 big money, 202–4, 265–67 billboards, 170, 267, 310 billing, 171–72, 184–85 Bing, 181–82 biodiversity, 146–47 biological realism, 253–54 biotechnology, 11–13, 17, 18, 109–10, 162, 330–31 Bitcoin, 34n BitTorrent, 223 blackmail, 61, 172–73, 207, 273, 314, 316, 322 Black Monday, 74 blogs, 118n, 120, 225, 245, 259, 349, 350 books, 1–2, 62, 63, 65, 113, 182, 192, 193, 246–47, 277–78, 281, 347, 352–60 bots, 62, 63, 64, 147–48 brain function, 195–96, 260, 328 brain scans, 111–12, 218, 367 Brand, Stewart, 214 brand advertising, 267 Brandeis, Louis, 25, 208 Brazil, 54 Brooks, David, 326 Burma, 200n Burning Man, 132 Bush, George H.

One Click: Jeff Bezos and the Rise of Amazon.com
by Richard L. Brandt
Published 27 Oct 2011

One of the headhunters called him and said, “I know you said you would kill me if I even proposed the finance thing, but there’s this opportunity that’s actually a very unusual financial company.” That company was D. E. Shaw. It was founded in 1988 by David Shaw to create a newfangled computer-automated trading system for Wall Street. Shaw was a computer science professor at Columbia University who was lured to Wall Street to help computerize stock trading systems, and then started his own company. At the time, computer software was already being used to track small differences in stock prices around the world, allowing arbitrageurs to buy the stock at one price and immediately sell it for a profit at the higher price.

Army of None: Autonomous Weapons and the Future of War
by Paul Scharre
Published 23 Apr 2018

Like many high-frequency trading firms, their business was lucrative. On the morning of July 31, 2012, Knight had $365 million in assets. Within 45 minutes, they would be bankrupt. At 9:30 a.m. Eastern Time on July 31, U.S. markets opened and Knight deployed a new automated trading system. Instantly, it was apparent that something was wrong. One of the functions of the automated trading system was to break up large orders into smaller ones, which then would be executed individually. Knight’s trading system wasn’t registering that these smaller trades were actually completed, however, so it kept tasking them again. This created an endless loop of trades.

Futureproof: 9 Rules for Humans in the Age of Automation
by Kevin Roose
Published 9 Mar 2021

But this cautionary note apparently hasn’t made it to the decision makers of corporate America, because they still seem to be placing excessive trust in the wisdom of AI, often with consequences much more severe than accidentally advertising some offensive T-shirts or coming up with a disgusting cocktail recipe. A trading firm called Knight Capital, for example, lost $440 million in a forty-five-minute span on August 1, 2012, after an improperly installed automated trading system rapidly bought and sold millions of shares, pushing their prices up and creating massive losses when they had to be resold. The losses nearly put Knight Capital out of business, and the firm had to receive hundreds of millions of dollars in emergency financing in order to stay afloat. Or take Watson, the IBM-owned AI that famously defeated a Jeopardy!

The Asylum: The Renegades Who Hijacked the World's Oil Market
by Leah McGrath Goodman
Published 15 Feb 2011

There also was a David-and-Goliath effect: the scrawnier traders with the intellectual know-how, though perhaps not the athletic prowess that had once dominated the pits, were finally able to match wits against the brawnier traders, without fear of physical retribution. On Wall Street, the move to the screen was inciting a technological arms race among the banks and hedge funds, which began sinking millions of dollars into state-of-the-art automated trading systems that could move at near-lightning speeds, outpacing the floor traders, many of whom still traded with no more than paper and pencil. The New York floor traders, whose brutality had once crushed the Marine Corps officers, were now getting a dose of their own medicine on their home turf—and they did not like it.

…

Once on the screen, the energy traders in arcades around the world became so intent on hitting their financial targets, they lost sight of the potential human repercussions. The screens had desensitized them to what was happening on the ground. Any harm done to energy consumers in the crossfire of their trading games was blithely categorized as collateral damage. Banks and hedge funds, whose automated trading systems bought and sold millions upon millions of energy contracts in just milliseconds, followed the same credo. Computer programmers were being hailed as the new gunslingers of Wall Street, fetching breathtaking salaries. Veteran energy traders began reporting that some of the computerized trading systems were letting banks and funds trade on the screen in ways that would have been illegal in the pits.

Electronic and Algorithmic Trading Technology: The Complete Guide
by Kendall Kim
Published 31 May 2007

If the intention is to increase staff and trading volume, as well as venture into different asset classes, a scalable in-house software solution may be the answer. 13.6 Conclusion The sell side will continue to undertake the difficult task of maintaining strong relationships with the buy side, which will allow them to grasp a foothold on market share. Major broker-dealers will enhance their market data infrastructure in order to translate large quantities of real-time data demanded by algorithmic and other automated trading systems for best execution. This will eliminate as much latency as possible. Direct market access companies, OMSs, and ECN aggregators will continue to be acquired by broker-dealers. Individual investors will put further pressure on their brokers and mutual fund managers for more transparency and to better understand management and operation fees.

How We Got Here: A Slightly Irreverent History of Technology and Markets
by Andy Kessler
Published 13 Jun 2005

These new fund managers take risks, with assurance that the companies they invest in provide accurate information, have liquid shares, trade cheaply and quickly and exist free of stock manipulation. These funds rarely own Russian gas refiners, as they fail all of the above assurances. But funds do own weird companies that make components for optical wave division multiplexing or some new biopharma company or the latest in supply chain management software, but they require automated trading systems to stay quick of foot. The New York Stock Exchange is still a people intensive exchange. Its specialist system was created in 1871. And we are still stuck with the NYSE monopoly on listed shares. Lots of reasons are offered, such as centralized pools of liquidity or orderly markets, etc.

More Money Than God: Hedge Funds and the Making of a New Elite
by Sebastian Mallaby
Published 9 Jun 2010

Before leaving Nabisco, Vannerson had spent a year working on the Dunn & Hargitt data, analyzing daily prices for fifteen commodities; and by the time Commodities Corporation opened its doors in March 1970, he had satisfied himself that price trends really did exist, no matter what academics might assert to the contrary.25 Moreover, Vannerson had devised a computer program that could trade on that finding. He called his brainchild the Technical Computer System, or TCS. It was one of the first in a long line of automated trading systems spawned by the hedge-fund industry.26 Weymar was initially skeptical of Vannerson’s project.27 His trend-following concept seemed disarmingly simple: Buy things that have just gone up on the theory that they will continue to go up; short things that have just gone down on the theory that they will continue to go down.

…

Even though Vannerson’s program took a step beyond that—it tried to distinguish upticks that might signify a lasting trend from upticks that signified nothing—Weymar still doubted that anyone could make serious money from something apparently so trivial. But by the summer of 1971, Weymar had reversed himself. The humiliation of the corn episode was one reason: The great virtue of an automated trading system was that risk controls had to be programmed into the computer from the start, and there was no danger of overconfident traders exceeding their allowed limits. But the TCS had proved itself to be superior at calling the market too. Weymar’s cocoa model, which had worked so well at Nabisco, had misjudged the direction of the market expensively during Commodities Corporation’s first year.

Crapshoot Investing: How Tech-Savvy Traders and Clueless Regulators Turned the Stock Market Into a Casino
by Jim McTague
Published 1 Mar 2011

The market was so tilted in their favor that in March 2011, Credit Suisse saw a big business opportunity in launching a new kind of trading platform—an Electronic Communication Network (ECN) that had rules favoring institutional investors and disadvantaging high-frequency traders. The SEC was so out of touch with the markets, it actually believed that the consolidated tape was the best source of information for the best prices in a listed security, precluding the need for investors to subscribe to data feeds from each of the 10 exchanges and 70-plus automated trading systems (ATS).4 In fact, right under the SEC’s nose, a duel market had developed: a high-end market for the moneyed traders and a low-end, less efficient market for retail investors. The SEC never noticed until Saluzzi and Arnuk got its attention in 2009. Regulators also failed to grasp a more fundamental fact.

Flash Boys: A Wall Street Revolt
by Michael Lewis
Published 30 Mar 2014

A fourth investor was told, by three different banks, that they didn’t want to connect to IEX because they didn’t want to pay the $300-a-month connection fee. Of all the banks that dragged their feet after their customers asked them to send their stock market orders to IEX, Goldman Sachs had offered the best excuse: They were afraid to tell their computer system to do anything it hadn’t done before. In August 2013, the Goldman automated trading system generated a bunch of crazy and embarrassing trades that lost Goldman hundreds of millions of dollars (until the public exchanges agreed, amazingly, to cancel them). Goldman wanted to avoid giving new instructions to its trading machines until it figured out why they had ceased to follow the old ones.

Finding Alphas: A Quantitative Approach to Building Trading Strategies
by Igor Tulchinsky
Published 30 Sep 2019

How do I start the search?” Even experienced quants working with many alphas can miss key components required to build a robust, diversified portfolio. For instance, one of the most difficult aspects of alpha portfolio construction is the need to optimize the level of diversification of the portfolio. In automated trading systems, decisions on diversification make up a major area of human intervention. It’s not easy to visualize the many pieces of the portfolio, which contain hundreds or thousands of alphas, and their interactions. TAP emerged from those concerns. When new quants start conducting alpha research, many begin by searching the internet for articles about developing alphas.

Broken Markets: How High Frequency Trading and Predatory Practices on Wall Street Are Destroying Investor Confidence and Your Portfolio
by Sal Arnuk and Joseph Saluzzi
Published 21 May 2012

NASDAQ was an acronym for National Association of Securities Dealers Automated Quotation system and was member-owned by the National Association of Securities Dealers (NASD). What started as a bulletin board quotation system grew into a full stock market, as the NASDAQ added volume reporting and automated trade systems. Due to less stringent listing requirements, small startups would raise money via initial public offerings (IPOs) on NASDAQ as opposed to the NYSE. It is hard to believe, but Intel Corp. started out as a $6.8 million IPO and Microsoft as a $60 million IPO. Through NASDAQ, broker dealers competed with each other by providing two-sided quotes in each stock listed.

Python for Algorithmic Trading: From Idea to Cloud Deployment
by Yves Hilpisch
Published 8 Dec 2020

Whenever it seems appropriate, references are given that point the reader to sources that address issues left open during the exposition. All in all, this book is written for readers who have some experience with both Python and (algorithmic) trading. For such a reader, the book is a practical guide to the creation of automated trading systems using Python and additional packages. This book uses a number of Python programming approaches (for example, object oriented programming) and packages (for example, scikit-learn) that cannot be explained in detail. The focus is on applying these approaches and packages to different steps in an algorithmic trading process.

Going Infinite: The Rise and Fall of a New Tycoon
by Michael Lewis
Published 2 Oct 2023

Sam saw his job as listening to what others had to say, and yet the members of the management team, along with pretty much everyone else in the new company, felt that he wouldn’t listen to a word they said. Amid the turmoil, Sam inhabited what amounted to his own reality. His attitude toward the missing money was, Eh, it’ll probably turn up somewhere. So let’s fucking trade! His first stab at an automated trading system was losing money at an alarming clip, but he’d created another, supposedly better one. Modelbot, it was called. Modelbot had been programmed to scour the world’s crypto exchanges for inefficiencies to exploit. If for even a few seconds it was possible to buy bitcoin on some Singaporean exchange for $7,900 and sell it for $7,920 on an exchange in Japan, Modelbot would do it over and over again, thousands of times per second.

Systematic Trading: A Unique New Method for Designing Trading and Investing Systems
by Robert Carver
Published 13 Sep 2015

www.ivolatility.com Source of option prices. www.oanda.com/currency/historical-rates Source of historic FX rates. stats.oecd.org Official source of macroeconomic data. mba.tuck.dartmouth.edu/pages/faculty/ken. french/data_library.html Academic source of equity value data. Brokers and platforms If you are not running a fully automated trading system, I hesitate to recommend a specific broker. A key selection criteria is price. Estimate the expected annual total fee from how often you’re trading, in what size, and with what total account value. Then use this for comparison purposes. Equally important is whether you can trade the products you want and with leverage if needed.

Dark Pools: The Rise of the Machine Traders and the Rigging of the U.S. Stock Market
by Scott Patterson
Published 11 Jun 2012

WHILE Maschler’s Datek traders were good at SOES, they had little edge over skilled competitors such as Houtkin. That quickly changed after Josh Levine began tinkering with Datek’s SOES trading system. Levine was still working at Russo and freelancing, but he started to poke around Datek’s operation at 50 Broad Street soon after he’d heard they’d been using an automated trading system. While he didn’t know the ins and outs of SOES at first, he quickly caught on—and liked what he saw. The greedy fat-cat insiders were getting their lunches eaten by Datek’s hit-and-run traders. It was a thing of beauty, a pristine example of how technology could shift the ground beneath the entrenched elite and transfer the power to their smarter, faster rivals.

The Permanent Portfolio
by Craig Rowland and J. M. Lawson
Published 27 Aug 2012

A problem in the trading systems of one firm can cause a cascade effect in the world markets. In spring of 2010 the U.S. stock markets experienced a “Flash Crash” during which the Dow Jones stock index sunk by 1,000 points within five minutes before quickly recovering. During this time, automated trading systems piled on sell orders, making the problem escalate quickly before finally coming back under control (some investors who had set stop losses in their accounts took large losses as they were automatically traded out of positions that recovered almost immediately and other bad trades were later backed out and canceled).

Fortunes of Change: The Rise of the Liberal Rich and the Remaking of America
by David Callahan
Published 9 Aug 2010

Back in the fabled 1980s, the flood of MBAs onto Wall Street was just beginning, and many top earners made do with BAs. Things changed as c01.indd 24 5/11/10 6:17:19 AM educated, rich, and liberal 25 finance became more complicated—as “quants” found new ways to make money using statistical techniques and algorithms, as automated trading systems came online, and as new financial products for managing risk or structuring debt, such as derivatives, came into being. Top universities spat out PhDs in fields such as mathematical finance and computational finance. Newly created master’s programs sprang up in financial engineering. MBA programs bolstered their quantitative offerings.

The Quants
by Scott Patterson
Published 2 Feb 2010

But since he didn’t have any computer programming skills, it limited his ability to design and implement models. Instead, he became PDT’s “human trader.” At the time, there were still certain markets, such as stock index futures, that weren’t fully automatic. Trades spat out by PDT’s models had to be called in over the telephone to other desks at Morgan. That was Tuttle’s job. The automated trading system didn’t always go smoothly. Once PDT mistakenly sold roughly $80 million worth of stock in about fifteen minutes due to a bug in the system. Another time Reed, who was running the Japanese stock system at the time, asked another trader to cover for him. “Just hit Y every time it signals a trade,” he said.

Unknown Market Wizards: The Best Traders You've Never Heard Of
by Jack D. Schwager
Published 2 Nov 2020

In essence, the fundamentalist seeks to uncover trading opportunities by identifying potential transitions to significantly more ample or tighter supply-demand balances. For financial futures, fundamental inputs will include items such as central bank policy, inflation statistics, employment data, etc. Technical and fundamental analysis are not mutually exclusive approaches. Many traders use both in the decision-making process or as components of automated trading systems. Delivery Shorts who maintain their positions in deliverable future contracts after the last trading day are obligated to deliver the given commodity or financial instrument against the contract. Similarly, longs who maintain their positions after the last trading day must accept delivery.

The Fund: Ray Dalio, Bridgewater Associates, and the Unraveling of a Wall Street Legend
by Rob Copeland
Published 7 Nov 2023

In these meetings and others, Dalio portrayed himself as a kindly supplicant offering a public good. There was merit to this; in sharing Bridgewater’s research with those in a position to make policy decisions, he was offering for free the type of insight for which clients around the world paid Bridgewater top dollar. Of course, Bridgewater also stood to benefit. While Bridgewater’s automated trading systems were positioned relatively conservatively, in 2007 and 2008 Dalio himself ordered a number of manual adjustments so the fund would profit more heavily from an overall decline, investment staffers there then recall. Throughout this period, he had placed a series of bets that would pay off if central bankers printed money to revive the economy—the exact move he had predicted.

Machine Learning Design Patterns: Solutions to Common Challenges in Data Preparation, Model Building, and MLOps
by Valliappa Lakshmanan , Sara Robinson and Michael Munn
Published 31 Oct 2020

However, such a model will miss all the information held in the destinations that the consolidator or travel agent was interested in at various times—this will affect things like how airports and hotels are embedded. One way to retain that information while not affecting the pricing decision is to use a neutral class for these transactions. Reframing with neutral class Suppose we are training an automated trading system that makes trades based on whether it expects a security to go up or down in price. Because of stock market volatility and the speed with which new information is reflected in stock prices, trying to trade on small predicted ups and downs is likely to lead to high trading costs and poor profits over time.

Future Politics: Living Together in a World Transformed by Tech
by Jamie Susskind
Published 3 Sep 2018

In time, fewer trades will be executed by screaming stockbrokers on trading floors, and more by intelligent trading algorithms that respond to market events at lightning speed. There used to be 600 US cash equities traders at Goldman Sachs’ headquarters in New York. Now there are just two. One-third of the bank’s staff are computer engineers.6 The introduction of automated trading systems has made financial regulation a nightmare for human officials. The algorithms are often too fast, too complex, and too adaptable to be subject to human oversight. OUP CORRECTED PROOF – FINAL, 26/05/18, SPi РЕЛИЗ ПОДГОТОВИЛА ГРУППА "What's News" VK.COM/WSNWS Force 103 Increasingly, the most practical method of enforcing the law is through the use of other algorithms programmed to detect and prohibit errant behaviour, or at least flag it for human attention.7 One of the digital systems being used to keep tabs on financial behaviour is a version of IBM’s Watson.8 The idea of digital law enforcement may seem weird but it doesn’t really disrupt our notion of what the law is.

The Crisis of Crowding: Quant Copycats, Ugly Models, and the New Crash Normal
by Ludwig B. Chincarini
Published 29 Jul 2012

In 2006, the NYSE became an electronic exchange, saying goodbye to the traditional floor brokers and specialists who ran around on those old wooden floors, screaming prices as they went.2 Stock purchases and sales can be completed through a variety of networks: a national exchange such as the NYSE, an electronic communication network (ECN), a large broker-dealer, or a dark pool. Most trading—64%—takes place on the exchanges. Broker-dealers account for another 18%, about 10% occurs through ECNs, and the remaining 8% occurs in dark pools.3 The exchanges consist of highly automated trading systems that respond to stock orders in less than one millisecond. ECNs are alternative trading systems that offer services that are very similar to those offered by the exchanges. They try to match buyers and sellers at the best bid and offer prices. All the trades that occur on the exchanges and the ECNs must be posted on the consolidated tape, including price and quantity traded.4 The consolidated tape is a real-time history of every trade in the U.S. markets.