Online Portfolio Selection: Principles and Algorithms

227 Pages·2016·6.06 MB·\227

by  Bin Li (Author), Steven Chu Hong Hoi (Author)

Finance & Investing / Machine Learning & Pattern Recognition L
Online Portfolio
i

a
With the aim to sequentially determine optimal allocations across a set of
n
assets, Online Portfolio Selection (OLPS) has significantly reshaped the d

financial investment landscape. Online Portfolio Selection: Principles H Selection
and Algorithms supplies a comprehensive survey of existing OLPS o
principles and presents a collection of innovative strategies that leverage i
machine learning techniques for financial investment.
Principles and Algorithms
The book presents four new algorithms based on machine learning
techniques that were designed by the authors, as well as a new back- O
test system they developed for evaluating trading strategy effectiveness.
The book uses simulations with real market data to illustrate the trading n
strategies in action and to provide readers with the confidence to deploy l
i
the strategies themselves. The book is presented in five sections that: n
e
I. Introduce and formulate OLPS as a sequential decision task

II. Present key OLPS principles, including benchmarks, follow the P
winner, follow the loser, pattern matching, and meta-learning o
III. Detail four innovative OLPS algorithms based on cutting-edge r
machine learning techniques t
f
IV. Provide a toolbox for evaluating the OLPS algorithms and present o
empirical studies comparing the proposed algorithms with the
l
state of the art i
o
V. Investigate possible future directions

S
Complete with a back-test system that uses historical data to evaluate
e
the performance of trading strategies, as well as MATLAB® code for the
l
back-test systems, this book is an ideal resource for graduate students in e
finance, computer science, and statistics. It is also suitable for researchers c
and engineers interested in computational investment.
t
i
o
Readers are encouraged to visit the authors’ website for updates:
http://olps.stevenhoi.org. n
K23731
6000 Broken Sound Parkway, NW Bin Li and Steven C.H. Hoi
Suite 300, Boca Raton, FL 33487 ISBN: 978-1-4822-4963-7
711 Third Avenue 90000
an informa business New York, NY 10017
2 Park Square, Milton Park
www.crcpress.com Abingdon, Oxon OX14 4RN, UK
9 781482 249637
w w w. c r c p r e s s . c o m
K23731 mech rev.indd 1 10/5/15 10:31 AM

Online Portfolio
Selection
Principles and Algorithms
T&FCat#K23731—K23731_C000—pagei—10/13/2015—16:23

Online Portfolio
Selection
Principles and Algorithms
Bin Li and Steven C.H. Hoi
T&FCat#K23731—K23731_C000—pageiii—10/13/2015—16:23
MATLAB® is a trademark of The MathWorks, Inc. and is used with permission. The MathWorks does
not warrant the accuracy of the text or exercises in this book. This book’s use or discussion of MAT-
LAB® software or related products does not constitute endorsement or sponsorship by The MathWorks
of a particular pedagogical approach or particular use of the MATLAB® software.
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2016 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Version Date: 20151001
International Standard Book Number-13: 978-1-4822-4964-4 (eBook - PDF)
This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication
and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information stor-
age or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copy-
right.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222
Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that pro-
vides licenses and registration for a variety of users. For organizations that have been granted a photo-
copy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Contents
ListofFigures ix
ListofTables xi
ListofNotations xiii
Preface xv
Acknowledgments xvii
Authors xix
I Introduction 1
1 Introduction 3
1.1 Background 4
1.1.1 Challenge1:VoluminousFinancialInstruments 4
1.1.2 Challenge2:HumanBehavioralBiases 4
1.1.3 Challenge3:High-FrequencyTrading 4
1.1.4 AlgorithmicTradingandMachineLearning 4
1.2 WhatIsOnlinePortfolioSelection? 5
1.3 Methodology 7
1.4 BookOverview 7
2 Problemformulation 11
2.1 ProblemSettings 11
2.2 TransactionCostsandMarginBuyingModels 13
2.3 Evaluation 14
2.4 Summary 16
II Principles 17
3 Benchmarks 21
3.1 Buy-and-HoldStrategy 21
3.2 BestStockStrategy 21
3.3 ConstantRebalancedPortfolios 22
v
T&FCat#K23731—K23731_C000—pagev—10/13/2015—16:23
vi CONTENTS
4 FollowtheWinner 23
4.1 UniversalPortfolios 23
4.2 ExponentialGradient 25
4.3 FollowtheLeader 26
4.4 FollowtheRegularizedLeader 27
4.5 Summary 29
5 FollowtheLoser 31
5.1 MeanReversion 31
5.2 Anticorrelation 32
5.3 Summary 33
6 PatternMatching 35
6.1 SampleSelectionTechniques 36
6.2 PortfolioOptimizationTechniques 37
6.3 Combinations 38
6.4 Summary 39
7 Meta-Learning 41
7.1 AggregatingAlgorithms 41
7.2 FastUniversalization 42
7.3 OnlineGradientandNewtonUpdates 43
7.4 FollowtheLeadingHistory 43
7.5 Summary 43
III Algorithms 45
8 Correlation-DrivenNonparametricLearning 47
8.1 Preliminaries 48
8.1.1 Motivation 48
8.2 Formulations 50
8.3 Algorithms 51
8.4 Analysis 56
8.5 Summary 57
9 Passive–AggressiveMeanReversion 59
9.1 Preliminaries 59
9.1.1 RelatedWork 59
9.1.2 Motivation 60
9.2 Formulations 62
9.3 Algorithms 65
9.4 Analysis 67
9.5 Summary 69
T&FCat#K23731—K23731_C000—pagevi—10/13/2015—16:23
CONTENTS vii
10 Confidence-WeightedMeanReversion 71
10.1 Preliminaries 71
10.1.1 Motivation 71
10.2 Formulations 73
10.3 Algorithms 76
10.4 Analysis 78
10.5 Summary 81
11 OnlineMovingAverageReversion 83
11.1 Preliminaries 83
11.1.1 RelatedWork 83
11.1.2 Motivation 84
11.2 Formulations 88
11.3 Algorithms 90
11.4 Analysis 91
11.5 Summary 92
IV EmpiricalStudies 93
12 Implementations 95
12.1 TheOLPSPlatform 95
12.1.1 Preprocess 96
12.1.2 AlgorithmicTrading 96
12.1.3 Postprocess 97
12.2 Data 97
12.3 Setups 99
12.3.1 ComparisonApproachesandTheirSetups 100
12.4 PerformanceMetrics 100
12.5 Summary 101
13 EmpiricalResults 103
13.1 Experiment1:EvaluationofCumulativeWealth 103
13.2 Experiment2:EvaluationofRiskandRisk-AdjustedReturn 105
13.3 Experiment3:EvaluationofParameterSensitivity 109
13.3.1 CORN’sParameterSensitivity 109
13.3.2 PAMR’sParameterSensitivity 109
13.3.3 CWMR’sParameterSensitivity 114
13.3.4 OLMAR’sParameterSensitivity 114
13.4 Experiment4:EvaluationofPracticalIssues 116
13.5 Experiment5:EvaluationofComputationalTime 120
13.6 Experiment6:DescriptiveAnalysisofAssetsandPortfolios 122
13.7 Summary 126
T&FCat#K23731—K23731_C000—pagevii—10/13/2015—16:23
viii CONTENTS
14 ThreatstoValidity 129
14.1 OnModelAssumptions 129
14.2 OnMeanReversionAssumptions 130
14.3 OnTheoreticalAnalysis 131
14.4 OnBack-Tests 131
14.5 Summary 133
V Conclusion 135
15 Conclusions 137
15.1 Conclusions 137
15.2 FutureDirections 138
15.2.1 OnExistingWork 138
15.2.2 OnPracticalIssues 140
15.2.3 LearningforIndexTracking 140
AppendixA OLPS:AToolboxforOnlinePortfolioSelection 143
AppendixB ProofsandDerivations 171
AppendixC SupplementaryDataandPortfolioStatistics 187
Bibliography 193
Index 205
T&FCat#K23731—K23731_C000—pageviii—10/13/2015—16:23