Quantitative Portfolio ManagementThe Art and Science of Statistical Arbitrage
<p><b>Discover foundational and advanced techniques in quantitative equity trading from a veteran insider </b></p> <p>In <i>Quantitative Portfolio Management: The Art and Science of Statistical Arbitrage</i>, distinguished physicist-turned-quant Dr. Michael Isichenko delivers a systematic review of the quantitative trading of equities, or statistical arbitrage. The book teaches you how to source financial data, learn patterns of asset returns from historical data, generate and combine multiple forecasts, manage risk, build a stock portfolio optimized for risk and trading costs, and execute trades. </p> <p>In this important book, you’ll discover: </p> <ul> <li>Machine learning methods of forecasting stock returns in efficient financial markets </li> <li>How to combine multiple forecasts into a single model by using secondary machine learning, dimensionality reduction, and other methods</li> <li>Ways of avoiding the pitfalls of overfitting and the curse of dimensionality, including topics of active research such as “benign overfitting” in machine learning </li> <li>The theoretical and practical aspects of portfolio construction, including multi-factor risk models, multi-period trading costs, and optimal leverage </li> </ul> <p>Perfect for investment professionals, like quantitative traders and portfolio managers, <i>Quantitative Portfolio Management</i> will also earn a place in the libraries of data scientists and students in a variety of statistical and quantitative disciplines. It is an indispensable guide for anyone who hopes to improve their understanding of how to apply data science, machine learning, and optimization to the stock market. </p> <br />
<p>Preface 1</p> <p>Introduction 3</p> <p><b>1 Market Data 9</b></p> <p>1.1 Tick and bar data 9</p> <p>1.2 Corporate actions and adjustment factor 10</p> <p>1.3 Linear vs log returns 11</p> <p><b>2 Forecasting 13</b></p> <p>2.1 Data for forecasts 14</p> <p>2.1.1 Point-in-time and lookahead 15</p> <p>2.1.2 Security master and survival bias 16</p> <p>2.1.3 Fundamental and accounting data 16</p> <p>2.1.4 Analyst estimates 17</p> <p>2.1.5 Supply chain and competition 18</p> <p>2.1.6 M&A and risk arbitrage 18</p> <p>2.1.7 Event-based predictors 18</p> <p>2.1.8 Holdings and flows 19</p> <p>2.1.9 News and social media 20</p> <p>2.1.10 Macroeconomic data 21</p> <p>2.1.11 Alternative data 21</p> <p>2.1.12 Alpha capture 21</p> <p>2.2 Technical forecasts 22</p> <p>2.2.1 Mean reversion 22</p> <p>2.2.2 Momentum 24</p> <p>2.2.3 Trading volume 24</p> <p>2.2.4 Statistical predictors 25</p> <p>2.2.5 Data from other asset classes 25</p> <p>2.3 Basic concepts of statistical learning 27</p> <p>2.3.1 Mutual information and Shannon entropy 28</p> <p>2.3.2 Likelihood and Bayesian inference 32</p> <p>2.3.3 Mean square error and correlation 33</p> <p>2.3.4 Bias-variance tradeoff 35</p> <p>2.3.5 PAC learnability, VC dimension, and generalization error bounds 36</p> <p>2.4 Machine learning 40</p> <p>2.4.1 Types of machine learning 41</p> <p>2.4.2 Overfitting 43</p> <p>2.4.3 Ordinary and generalized least squares 44</p> <p>2.4.4 Deep learning 46</p> <p>2.4.5 Types of neural networks 48</p> <p>2.4.6 Nonparametric methods 51</p> <p>2.4.7 Cross-validation 54</p> <p>2.4.8 Curse of dimensionality, eigenvalue cleaning, and shrinkage 56</p> <p>2.4.9 Smoothing and regularization 61</p> <p>18.104.22.168 Smoothing spline 62</p> <p>22.214.171.124 Total variation denoising 62</p> <p>126.96.36.199 Nadaraya-Watson kernel smoother 63</p> <p>188.8.131.52 Local linear regression 64</p> <p>184.108.40.206 Gaussian process 64</p> <p>220.127.116.11 Ridge and kernel ridge regression 67</p> <p>18.104.22.168 Bandwidth and hypertuning of kernel smoothers 68</p> <p>22.214.171.124 Lasso regression 69</p> <p>2.4.10 Generalization puzzle of deep and overparameterized learning 69</p> <p>2.4.11 Online machine learning 74</p> <p>2.4.12 Boosting 75</p> <p>2.4.13 Randomized learning 79</p> <p>2.4.14 Latent structure 80</p> <p>2.4.15 No free lunch and AutoML 81</p> <p>2.4.16 Computer power and machine learning 83</p> <p>2.5 Dynamical modeling 87</p> <p>2.6 Alternative reality 89</p> <p>2.7 Timeliness-significance tradeoff 90</p> <p>2.8 Grouping 91</p> <p>2.9 Conditioning 92</p> <p>2.10 Pairwise predictors 92</p> <p>2.11 Forecast for securities from their linear combinations 93</p> <p>2.12 Forecast research vs simulation 95</p> <p><b>3 Forecast Combining 97</b></p> <p>3.1 Correlation and diversification 98</p> <p>3.2 Portfolio combining 99</p> <p>3.3 Mean-variance combination of forecasts 102</p> <p>3.4 Combining features vs combining forecasts 103</p> <p>3.5 Dimensionality reduction 104</p> <p>3.5.1 PCA, PCR, CCA, ICA, LCA, and PLS 105</p> <p>3.5.2 Clustering 107</p> <p>3.5.3 Hierarchical combining 108</p> <p>3.6 Synthetic security view 108</p> <p>3.7 Collaborative filtering 109</p> <p>3.8 Alpha pool management 110</p> <p>3.8.1 Forecast development guidelines 111</p> <p>126.96.36.199 Point-in-time data 111</p> <p>188.8.131.52 Horizon and scaling 111</p> <p>184.108.40.206 Type of target return 112</p> <p>220.127.116.11 Performance metrics 112</p> <p>18.104.22.168 Measure of forecast uncertainty 112</p> <p>22.214.171.124 Correlation with existing forecasts 112</p> <p>126.96.36.199 Raw feature library 113</p> <p>188.8.131.52 Overfit handling 113</p> <p>3.8.2 Pnl attribution 114</p> <p>184.108.40.206 Marginal attribution 114</p> <p>220.127.116.11 Regression-based attribution 114</p> <p><b>4 Risk 117</b></p> <p>4.1 Value at risk and expected shortfall 117</p> <p>4.2 Factor models 119</p> <p>4.3 Types of risk factors 120</p> <p>4.4 Return and risk decomposition 121</p> <p>4.5 Weighted PCA 122</p> <p>4.6 PCA transformation 123</p> <p>4.7 Crowding and liquidation 124</p> <p>4.8 Liquidity risk and short squeeze 126</p> <p>4.9 Forecast uncertainty and alpha risk 127</p> <p><b>5 Trading Costs 129</b></p> <p>5.1 Slippage 130</p> <p>5.2 Impact 130</p> <p>5.2.1 Empirical observations 132</p> <p>5.2.2 Linear impact model 133</p> <p>5.2.3 Impact arbitrage 135</p> <p>5.3 Cost of carry 135</p> <p><b>6 Portfolio Construction 137</b></p> <p>6.1 Hedged allocation 137</p> <p>6.2 Single-period vs multi-period mean-variance utility 139</p> <p>6.3 Single-name multi-period optimization 140</p> <p>6.3.1 Optimization with fast impact decay 141</p> <p>6.3.2 Optimization with exponentially decaying impact 142</p> <p>6.3.3 Optimization conditional on a future position 143</p> <p>6.3.4 Position value and utility leak 145</p> <p>6.3.5 Optimization with slippage 146</p> <p>6.4 Multi-period portfolio optimization 148</p> <p>6.4.1 Unconstrained portfolio optimization with linear impact costs 149</p> <p>6.4.2 Iterative handling of factor risk 150</p> <p>6.4.3 Optimizing future EMA positions 151</p> <p>6.4.4 Portfolio optimization using utility leak rate 151</p> <p>6.4.5 Notes on portfolio optimization with slippage 152</p> <p>6.5 Portfolio capacity 152</p> <p>6.6 Portfolio optimization with forecast revision 153</p> <p>6.7 Portfolio optimization with forecast uncertainty 156</p> <p>6.8 Kelly criterion and optimal leverage 157</p> <p>6.9 Intraday optimization and execution 160</p> <p>6.9.1 Trade curve 160</p> <p>6.9.2 Forecast-timed execution 161</p> <p>6.9.3 Algorithmic trading and HFT 162</p> <p>6.9.4 HFT controversy 166</p> <p><b>7 Simulation 169</b></p> <p>7.1 Simulation vs production 170</p> <p>7.2 Simulation and overfitting 171</p> <p>7.3 Research and simulation efficiency 172</p> <p>7.4 Paper trading 173</p> <p>7.5 Bugs 173</p> <p>Afterword: Economic and Social Aspects of Quant Trading 179</p> <p><b>Appendix 183</b></p> <p>A1 Secmaster mappings 183</p> <p>A2 Woodbury matrix identities 184</p> <p>A3 Toeplitz matrix 185</p> <p><b>Index 187</b></p> <p>Questions index 195</p> <p>Quotes index 197</p> <p>Stories index 199</p>
<P><B>MICHAEL ISICHENKO, PhD,</B> is a theoretical physicist and a quantitative portfolio manager who worked at Kurchatov Institute, University of Texas, University of California, SAC Capital Advisors, Société Générale, and Jefferies. He received his doctorate in physics and mathematics from the Moscow Institute of Physics and Technology and is an expert in plasma physics, nonlinear dynamics, and statistical and chaos theory.
<p>Quantitative trading has become a multi-billion-dollar industry employing thousands of portfolio managers and quantitative analysts (quants) trained in mathematics, physics, and other “hard” sciences. Quants trade securities by quickly finding and exploiting mispricing in the market, creating liquidity, and maintaining the efficiency of financial markets.</P> <P>In <i>Quantitative Portfolio Management: The Art and Science of Statistical Arbitrage,</i> theoretical physicist and accomplished quantitative portfolio manager Dr. Michael Isichenko delivers a systematic review of the quant equity trading process, also known as statistical arbitrage. <P>Covering every major component of the quantitative trading process, the author discusses how to source financial data, learn future asset returns from historical data, generate and combine multiple forecasts, manage risk, build optimal portfolios mindful of risk preferences and trading costs, and execute trades <P>The book balances practical financial insights with mathematical ideas of statistical and machine learning, computational strategies, and examples gleaned from the author’s years of experience as a quant portfolio manager. You’ll also find a collection of insightful and perplexing questions asked at quant interviews. <P>Quantitative Portfolio Management includes discussions of complex topics that remain the subject of active research, like double descent of generalization error in regression and deep learning, forecast combination and its diversification limits, and market-wide elasticity. <P>Throughout, the book focuses on the application of machine learning and forecasting techniques to real-world portfolio optimization problems. It offers special closed-form solutions with impact and slippage costs and approximations for efficient algorithmic approaches. <P>Perfect for investment professionals, including quants and portfolio managers, <i>Quantitative Portfolio Management</i> will also earn a place in the libraries of traders, data scientists, and students of finance, data science, and machine learning seeking a one-stop resource from a recognized expert in quantitative finance.
<P><B>Praise for QUANTITATIVE PORTFOLIO MANAGEMENT</B></P> <P>“This is a wonderful book: deep, original, witty, and provocative. It is a survey of the most important ideas and methods of modern quantitative investment that should enthrall both seasoned and junior quants. A must-read that will no doubt become a classic.” <P><B>—Jean-Philippe Bouchaud,</B> Chairman and Chief Scientist, Capital Fund Management; member of the French Academy of Sciences <P>“In his lively and clever style, Isichenko shares from his decades of experience at some of the top quantitative trading shops. Even seasoned veterans will find unfamiliar ideas, as he includes many concepts and models nowhere else in print.” <P><B>—Colin Rust,</b> Quantitative Portfolio Manager, Cubist Systematic Strategies <P>“I encouraged Michael Isichenko not to seek publication of this book, a comprehensive and accurate survey of market structure and data and mathematical and computational approaches and results for systematic trading. I am grateful that he enlarged and extended it beyond a first draft. I now hope that competitors have so much to absorb that they'll misapply much and not eliminate all remaining avenues to profit for my firm.” <P><B>—Aaron Sosnick,</B> Founder, Analytics, Research & Trading Advisors <P><B>An in-depth and telling handbook for quant portfolio management from a leading industry expert</b> <P><I>Quantitative Portfolio Management</I> is a complete and up-to-date exploration of the quantitative analysis process. You’ll find information about sourcing financial data, alpha generation approaches, dealing with risk, portfolio construction, and trade execution. <P>The book covers both theoretical and algorithmic machine learning subjects in the context of competition-based market efficiency that imposes limits on complexity and performance of quantitative trading models. In addition to foundational subjects that form the basis of quantitative finance, you’ll also learn about lesser-known machine learning algorithms and rarely discussed topics, like forecast combining and multi-period portfolio optimization. The author expertly balances practical observations drawn from his years as a practicing portfolio manager with financial and mathematical insights in statistics and machine learning.
NeuheitenIntroduction to Energy and Sustaina... 88,99 €
Making Sense of AI 14,99 €
Black is the Journey, Africana the ... 12,99 €
Structural Anthropology Zero 17,99 €
Skin 18,99 €