INDEX

Action-based manipulation, 21–22 input layer nodes, 47 Advanced detection system (ADS), output layer nodes, 49 30 Autoregressive integrated moving Amsterdam markets, 1 average (ARIMA), 17 Analytical techniques, 30 Behavioural finance, 3 x2 approximation test, 62–63 Artificial neural network-genetic Bombay (BSE), 4, 39 index, 4 algorithm based composite Box’s M-test, 44, 62 mode (ANN-GA based composite model). See also C#.NET 2.0, 53, 83 Support Vector Machines , 4, 10–11 model (SVM model), 35–37, in India, 3–4 39–40, 53, 83–85 trading behaviour in, 10 applying weights to neural Capital Markets Cooperative network, 71 Research Centre (CMCRC), artificial neural network based 2 model, 47–49 Chromosome, 68 comparison of results, 53–54 Classification, 30 convergence, 70–71 Classifier equation, 37 for detecting stock price Closing price manipulation, 25 manipulation, 46–51 Co-location facility, 4–5 determining weights using Composite model, 34, 67 genetic algorithm, 68–71 Confusion matrix, 39–40, 72, development of model, 67–71 81–82 fitness function, 70 Continuous auction framework, 20 generating chromosome, 68 Convergence, 70–71 reproduction, 70 Corporate practices, 21 results, 72–73 Crossover operation, 47, 49 weight extraction, 69–70 Currency Artificial neural networks (ANN), derivatives segments, 4 31, 46, 66, 72 futures and options, 4 ANN based model, 47–49 computing weights using genetic , 8 algorithm, 49–51 Department of Company Affairs hidden layer nodes, 47–49 (DCA), 5

101 102 Index

Department of Economics Affairs Hedge Funds, 27 (DEA), 5 Helsinki Stock Exchange, 27 Depositories, 11–12 Hyperplane, 51, 75 Derivatives Market, 4 maximum , 77 Dhaka Stock Exchange (DSE), 17 Illegal price manipulation, 21 Direct Market Access (DMA), Indian Capital Market, 85 4–5 Indian Equity Exchanges, 71, 81 Discriminant analysis, 31, 35, 37, Indian Equity Market, 37, 39 39–40, 55 Indian Stock Exchange, 1, 38 Discrimination, 30 Indian , 3, 5, 33 DTREG, 53, 83 BSE, 4 Dual Decision function, 80 identifying research gap, Econometrics and network, 31 33–34 Effective Market Surveillance, 10 key developments in, 4–5 Efficient market, 16–17 limitations of scope, 35 Efficient stock market, 3 NSE, 4 Efficient-Market Hypothesis research objectives, 35–36 (EMH), 3, 16–17 scope of research, 34–35 Equal variance-covariance, 62–63 Information test of equal variance-covariance dissemination, 16 matrices, 43–44 false, 23–24 Equities based derivatives, 4 flow, 17 Equity based ETFs, 4 information-based manipulation, 21–22, 24 F-approximation insider, 23 method, 63 material, 3 test, 45 private, 25 Fitness function, 70 fl – Information ow dynamics, 17 FIX capabilities, 4 5 Informed trader, 25 Flash Orders, 15 Instruments, 2 Futures & Options segments Integrated Market Surveillance (F&O segments), 37–38 System (IMSS), 11–12 Game theory, 23 Integrated Surveillance Department Generalized Squared Distance of SEBI, 11 Function, 64 Intermediaries, 2 Genetic Algorithm (GA), 47, IPOs, 4–5, 19 50–51, 66 Istanbul Stock Exchange, 31 determining weights using, Karush–Kuhn–Tucker conditions 68–71 (KKT conditions), 79 Global Capital Markets, 6–7 Kernel function, 79–80 Gold ETF, 4 Graph clustering algorithm, 31 ‘Lead-lag’ linkages, 17 Guinness Four Business Scandal, Linear Classification Function, The, 6 41–42, 56, 59 Index 103

Linear classifier, 85 Market Linear discriminant analysis. See data dissemination, 3 also Quadratic Discriminant integrity, 17–18 analysis, 83–84 regulation, 7 Linear discriminant function (LDF). surveillance system, 7, 11, 13, See also Quadratic 15, 18, 29–30, 83 Discriminant Function transparency, 18 (QDF), 35, 39–40, 53, 55, Market manipulation, 5, 9, 13, 19, 59, 85 29 comparison of results, 53–54 action-based manipulation, 22 for detecting stock price empirical studies in, 26–29 manipulation, 41–42 information-based development of model, 55–56 manipulation, 22–24 F-approximation test, 45, 63 theoretical foundation to, 20–26 limitation of model, 58 trade-based manipulation, linear classification function, 24–26 41–42, 56 Market Quality Forum, 34–35 results, 56–58 Market structure, 2–3 test for multivariate normality, instruments, 2 42–43 market data dissemination, 3 test of equal variance-covariance market participants and matrices, 43–44 intermediaries, 2 test to check data, 60 models adopted in present work, test to check for equal variance- 13 covariance, 62–63 regulator and regulations, 2 testing assumptions governing, technology, 2 42, 45, 59, 63 MATLAB, 53, 83 x2 approximation test, 62–63 Microsoft. Net framework, 83 Linear kernel function, 79 Mini Nifty, 4 Linear Soft Margin Classifier, Misclassification tables, 37, 40 52, 75 Mobile trading, 4–5 Linearly Separable Classifier, Multi discriminant analysis 52, 75–76, 79 (MDA), 34 Liquid market, 16 Multivariate normality, test for, Livedoor Scandal, The, 6 42–43 Logistics regression, 31 Mutual Fund Service System, 4 Logit, Genetic Algorithm, 34 Stock Market, 30 Dated Options, 4 Nasdaq-Liffe (futures) stock Manipulate/manipulation, 19, 24, markets, 30 33 National Stock Exchange (NSE), 4, in stock market, 1 6, 37–39 of stock prices, 1 Neural network model, 34, 46, techniques to detect, 1, 30–31 49–50, 68 Manipulators, 1 applying weights to, 71 104 Index

Neurons, 46–47 Saddle point, 78 Securities Exchange Board of India (NYSE), 18 (SEBI), 5–7, 10, 12, 19–20, Non-linear classifier, 52, 75, 79, 81 38–39 IMSS, 11–12 Online message board database, Securities Exchange Commission, 24 23 Opportunistic individuals, 23 Securities Observation, News Over the counter (OTC), 30 Analysis, and Regulation Penny , 8 system (SONAR system), Polynomial kernel function, 79 30 Price of stock, 1, 3 Securities regulation, 18 Profit maximization, 24 Self-regulating organizations Pump-and-dump strategy, 28 (SROs), 6, 83 SENSEX, 4 – Q-Q plot, 42, 60 61 Sequential trade framework, 20 fi Quadratic Classi cation function, Share prices, 20 64 Sigmoid kernel function, 79 Quadratic discriminant analysis, Sigmoidal activation function, 71 – – 63 64, 83 84 Small cap stocks, 8 – results, 64 66 Smart Order Routing (SOR), 4–5 testing assumptions governing SQL Server 2005, 83 Linear Discriminant Stock, 75–76 – Function, 59 63 Stock market Quadratic discriminant function efficient, 3 . See also (QDF) Linear Indian Stock Market, 3–5 discriminant function, motivation for research, 12–13 – – – 35 36, 39 40, 45 46, 53, SEBI, 10–12 – – 63 65, 84 85 stock price manipulation, 5–9 – comparison of results, 53 54 surveillance, 9–12 QDF based model, 85 surveillance systems, 31 Quadratic Discriminant analysis Stock price manipulation, 5, 9, 55 for detecting stock price ANN-GA based composite – manipulation, 45 46 model for detecting stock – Radial basis function (RBF), 79 price manipulation, 46 51 kernel function, 52, 81 issues in identifying Random walk behaviour, 16 manipulation, 7–9 Redistribution method, 40 Linear Discriminant Function Regulations, 2 for detecting, 41–42 Regulator, 2 Quadratic Discriminant analysis Reproduction, 70 for detecting, 45–46 Reserve Bank of India (RBI), 5–7 SVM model for detecting, 51–53 Retail equity , 4–5 types of manipulation, 9 Retail Government Securities, 4 Structure–based methods, 31 Index 105

Substantiation, 21 Theory of Efficient Market, 16 Support vector machine model Trade data, 37–39 (SVM model). See also Trade-based manipulation, 21–22, Artificial neural network- 24, 26 genetic algorithm based Trade-based market composite mode (ANN-GA efficient market, 16–17 based composite model), 31, literature review of, 15 34–37, 39–40, 53, 75–77, market integrity, 17–18 81, 84–86 market manipulation, comparison of results, 53–54 19–29 confusion matrix, 82 market surveillance, 29–30 development of model, 75–81 techniques to detect error count estimates for stock, manipulation, 30–31 82 Trade-based price manipulation, linearly separable classifier, 20 76–79 Trading regulations, 29 model for detecting stock price , 21 manipulation, 51–53 Tunisian Stock Market (TSE), non-linear classifier, 79–81 17 results, 81–82 Surveillance system, 7, 9 Variance-covariance matrix, 34, 59, 83–84 stock market, 9, 12 , 3, 26 Technology, 2 Weight extraction, 69–70 (TSE), 31