Performance Evaluation and Attribution of Security Portfolios Introduction to the Series ii The aim of the Handbooks in Economics series is to produce Handbooks for various branches of economics, each of which is a definitive source, reference, and teaching supplement for use by professional researchers and advanced graduate students Each Handbook provides self-contained surveys of the current state of a branch of economics in the form of chapters prepared by leading specialists on various aspects of this branch of economics These surveys summarize not only received results but also newer developments, from recent journal articles and discussion papers Some original material is also included, but the main goal is to provide comprehensive and accessible surveys The Handbooks are intended to provide not only useful reference volumes for professional collections but also possible supplementary readings for advanced courses for graduate students in economics KENNETH J ARROW and MICHAEL D INTRILIGATOR Performance Evaluation and Attribution of Security Portfolios by Bernd Fischer and Russell Wermers Academic Press is an imprint of Elsevier The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK 225 Wyman Street, Waltham, MA 02451, USA First edition 2013 Copyright © 2013 Elsevier Inc All rights reserved SOLNIK, BRUNO, McLEAVEY, DENNIS, GLOBAL INVESTMENTS, 6th Edition, © 2009, Reprinted by permission of Pearson Education, Inc., Upper Saddle River, NJ No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: permissions@elsevier.com Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is availabe from the Library of Congress ISBN–13: 978-0-12-744483-3 For information on all Academic Press publications visit our web site at store.elsevier.com Printed and bound in the US 12 13 14 15 16 10 9 8 7 6 5 4 3 2 1 Preface v This book is intended to be the scientific state-of-the-art in performance evaluation—the measurement of manager skills—and performance attribution—the measurement of all of the sources of manager returns, including skill-based We have attempted to include the best and most promising scientific approaches to these topics, drawn from a voluminous and quickly expanding literature Our objective in this book is to distill hundreds of both classic and the best cuttingedge academic and practitioner research papers into a unified framework Our goal is to present the most important concepts in the literature in order to provide a directed study and/or authoritative reference that saves time for the practitioner or academic researcher Sufficient detail is provided, in most cases, such that the investment practitioner can implement the approaches with data immediately, without consulting the underlying literature For the academic, we have provided enough detail to allow an easy further study of the literature, as desired We have contributed in two dimensions in this volume—both of which, we believe, are missing in currently available textbooks Firstly, we provide a timely overview of the most important performance evaluation techniques, which allow an accurate assessment of the skills of a portfolio manager Secondly, we provide an equally timely overview of the most important and widely used performance attribution techniques, which allow an accurate measure of all of the sources of investment returns, and which are necessary for precise performance reporting by fund managers We believe that our text is timely An estimated $71.3 trillion was invested in managed portfolios worldwide, as of 2009 (source: www.thecityuk.com) Managing this money, thus, is a business that draws perhaps $700 billion per year in management fees and other expenses for asset managers, in addition to a perhaps similar magnitude in annual trading costs accruing to brokers, market makers, and other liquidity providers (i.e., Wall Street and other financial centers) Our book is the first comprehensive text covering the latest science of measuring the main output of portfolio managers: their benchmarkrelative performance (alpha) Our hope is that investors use these techniques to improve the allocation of their money, and that portfolio management firms use them to better understand the quality of their funds’ output for investors We intend this book to be used in at least two ways: First, as a useful reference source for investment practitioners—who may wish to read only one or a few chapters We have attempted to make chapters self-contained to meet this demand We have also included chapter-end questions that both test the reader’s vi Preface understanding and provide examples of applications of each chapter’s concepts The audience for this use includes (at least) those studying for the CFA exams; performance analysts; mutual fund and pension fund trustees; portfolio managers of mutual funds, pension funds, hedge funds, and fund-of-funds; asset management ratings companies (e.g., Lipper and Morningstar); quantitative portfolio strategists, regulators, financial planners, and sophisticated individual investors Second, the book serves as an efficient way for mathematically advanced undergraduate, masters, or Ph.D students to undertake a thorough foundation in the science of performance evaluation and attribution After reading this book, students will be prepared to handle new developments in these fields We have attempted to design each chapter of this book to contain enough detail to bring the reader to a point of being able to apply the concepts therein, including the chapter-end problems In cases where further detail may be needed, we have cited the most relevant source papers to allow further reading We have divided our book into two sections: Part of the book covers the area of performance evaluation Chapter provides a short overview of the basics of empirical asset-pricing as applied to performance assessment, including basic factor models, the CAPM, the Fama-French three-factor model and the research on momentum, and the characteristic-based stock benchmarking model of Daniel, Grinblatt, Titman, and Wermers Chapter provides an overview of returns-based factor models, and the issues involved in implementing them Chapter discusses the issue of luck vs skill in generating investment returns, and presents the fundamental performance evaluation measures, including those based on the Chapter factor models In addition, extensions of these factor models are introduced that contain factors that capture the ability of portfolio managers to time the stock market or to time securities over the business cycle Chapter presents the latest approaches to using portfolio holdings to more precisely measure the skill of a portfolio manager Chapter provides a complete system for evaluating the skills of a portfolio manager using her portfolio holdings and net returns Many managed portfolios generate non-normal returns Chapter shows how to apply bootstrap techniques to generate more precise estimates of the statistical significance of manager skills in the presence of non-normal returns and alphas Chapter covers a very new topic: how to capture the time-varying abilities of a portfolio manager (as briefly introduced in Chapter 3) Specifically, this chapter shows how to predict which managers are most likely to generate superior alphas in the current economic climate Chapter also covers a very recent topic in performance evaluation: the assessment of the proportion of a group of funds that are truly skilled using only their net returns This approach is very useful in assessing whether the highest alpha managers are truly skilled, or are simply the luckiest in a large group of managers Preface Finally, Chapter is a “capstone chapter,” in that it provides an overview of the research findings that use the principles outlined in the first chapters As such, it is a very useful summary of what works (and what does not) when looking for a superior asset manager (a “SAM”) and trying to avoid an inferior asset manager (an “IAM”) Part of the book primarily concerns performance attribution and related topics Since attribution analysis has become a crucial component within the internal control system of investment managers and institutional clients, ample space is dedicated to a thorough treatment of this field The focus in this part lies on the practical applications rather than on the discussions of the various approaches from an academic point of view This (practitioner’s) approach is accompanied by a multitude of examples derived from practical experience in the investment industry Great emphasis was also put on the underlying mathematical detail, which is required for an implementation in practice Chapter 10 provides an overview of the basic approaches for the measurement of returns In particular, the concepts of time-weighted return and internal rate of return, as well as approximation methods for these measures are discussed in detail Attribution analysis, in practice, requires a deep understanding of the benchmarks against which the portfolios are measured Chapter 11 provides an introduction to the benchmarks commonly used in practice, and their underlying concepts Chapter 12 covers fundamental models for the attribution analysis of equity portfolios developed by Gary Brinson and others Furthermore, basic approaches for the treatment of currency effects and the linkage of performance contributions over multiple periods are considered Chapter 13 contains an introduction to attribution analysis for fixed income portfolios from a practitioner’s point of view The focus lies on a methodology that is based on a full valuation of the bonds and the option-adjusted spread In addition, various other approaches are described Based on the methodologies for equity and fixed income portfolios, Chapter 14 presents different methodologies for the attribution analysis of balanced portfolios This chapter also illustrates the basic approaches for a risk-adjusted attribution analysis and covers specific aspects in the analysis of hedge funds Chapter 15 describes the various approaches for the consideration of derivatives within the common methodologies for attribution analysis The final chapter (Chapter 16) deals with Global Investment Performance Standards, a globally applied set of ethical standards for the presentation of the performance results of investment firms The authors are indebted to many dedicated academic researchers and tireless practitioners for many of the insights in this book Professor Wermers wishes to thank the many investment practitioners that have provided data or insights into the topics vii viii Preface of this book, including through their professional investment management activities: Robert Jones of Goldman Sachs Asset Management (now at System Two and Arwen), Rudy Schadt of Invesco, Scott Schoelzel and Sandy Rufenacht of Janus (now retired, and at Three Peaks Capital Management, respectively), Bill Miller and Ken Fuller of Legg-Mason, Andrew Clark, Otto Kober, Matt Lemieux, Tom Roseen, and Robin Thurston of Lipper, Don Phillips, John Rekenthaler, Annette Larson, and Paul Kaplan of Morningstar, Sean Collins and Brian Reid of the Investment Company Institute Professor Wermers also wishes to thank all of the classes taught on performance evaluation and attribution since 2001—at Chulalongkorn University (Bangkok); the European Central Bank (Frankfurt); the Swiss Finance Institute/FAME Executive Education Program (Geneva); Queensland University of Technology (Brisbane); Stockholm University; the University of Technology, Sydney; and the University of Vienna Special thanks are due to students in that first class of the SFI/FAME program during those dark days in September 2001, 10 days after the 9-11 attacks Professor Wermers is also indebted to his loving family, Johanna, Natalie, and Samantha, for the endless hours spent away from them while preparing and teaching this subject He gratefully acknowledges Thomas Copeland and Richard Roll of UCLA and Josef Lakonishok of University of Illinois (and LSV Asset Management) for early inspiration, as well as Wayne Ferson, Robert Stambaugh, Lubos Pastor, and Mark Carhart for their recent contributions to the field In addition, he owes his career to the brilliant mentoring of Mark Grinblatt and Sheridan Titman at UCLA, pioneers in the subject of performance evaluation This text would not have been possible from such humble beginnings without their selfless support and guidance Dr Fischer is indebted to his colleagues at IDS GmbH—Analysis and Reporting Services, an international provider of operational investment controlling services Over the past years he has greatly benefited from numerous discussions surrounding practical applications Thanks are also due to Dr Fischer’s former team members at Cominvest Asset Management GmbH The design and the implementation of a globally applicable attribution software from scratch, and the implementation of the Global Investment Performance Standards were exciting experiences which left their mark on the current treatise He also wishes to thank various colleagues (Markus Buchholz, Detlev Kleis, Ulrich Raber, Carsten Wittrock, and others), with whom he co-authored papers in the past Several sections in this book are greatly indebted to the views expressed there Dr Fischer is also indebted to the CFA institute and the Global Investment Performance Committee for formative discussions surrounding the draft of the GIPS in 1998/1999 and during his official membership term from 2000 to 2004 Both authors wish to thank J Scott Bentley of Elsevier, whose vision it was to create such a book, and whose patience it took to see it through To those whose contributions we have overlooked, our sincere apologies; such an ambitious undertaking as condensing a huge literature necessitates that the authors Preface choose topics that are either most familiar to us or viewed by us as most widely useful Surely, we have missed some important papers, and we hope to have a chance to create a second edition that expands on this one Finally, to the asset management practitioner: we dedicate this volume to you, and hope that it is useful in furthering your goal of providing high-quality investment management services! ix Chapter An Introduction to Asset Pricing Models ABSTRACT This chapter provides a brief overview of asset pricing models, with an emphasis on those models that are widely used to describe the returns of traded financial securities Here, we focus on various models of stock returns and fixed-income returns, and discuss the reasoning and assumptions that underlie the structure of each of these models 1.1 HISTORICAL ASSET PRICING MODELS Individuals are born with a sense of the perils of risk, and they develop mental adjustments to penalize opportunities that involve more risk.1 For example, farmers not plant corn, which requires a great deal of rainfall (which may or may not happen), unless the expected price of corn at harvest time is sufficiently high Currency traders will not take a long position in the Thai baht and short the U.S dollar unless they expect the baht to appreciate sufficiently In essence, the farmer and the currency trader are each applying a “personal discount rate” to the expected return of planting corn or investing in baht The farmer’s discount rate depends on his assessment of the risk of rainfall (which greatly affects his total corn crop output) and the risk of a price change in the crop The currency trader’s discount rate depends on the relative economic health of Thailand and the U.S., and any potential government intervention against currency Gibson and Walk (1960) performed a famous experiment that was designed to test for depth perception possessed by infants as young as six months old Infants were unwilling to crawl on a transparent glass plate that was placed over a several-foot drop, proving that they possessed depth perception at a very early age Another inference which can be drawn from this experiment is that infants already perceive physical risks and exhibit risk-averse behavior at a very early age (probably before they are environmentally taught to avoid risk) Performance Evaluation and Attribution of Security Portfolios http://dx.doi.org/10.1016/B978-0-08-092652-0.00001-7 © 2013 Elsevier Inc All rights reserved For End-of-chapter Questions: © 2012 CFA Institute, Reproduced and republished with permission from CFA Institute All rights reserved Keywords Asset Pricing Models, CAPM, Factor Models, Fama French three-factor model, Carhart four-factor model, DGTW stock characteristics model, Estimating beta, Expected return and risk Index A AAPL See Apple (AAPL) ABS See asset-backed securities (ABS) accrual accounting, 652 active management, 287, 289 active manager, 289 funds benefits, 289–290 passive alternative vs., 290 active management approach See also passive management approach currency contribution separation, 435–437 portfolio and benchmark data, 436 active management in efficient markets academic literature, 286 active vs passive management, 287–288 actively managed bond funds, 288 asset classes and vehicles, 286 average underperformance, 287 financial crisis, 288–289 four factor alphas, ISA, 287 self-selection bias, 286 study’s results, 287 problems, 309 SAMs, 290 active risk budgeting, 305 fund/manager characteristics, 299 macroeconomic forecasting, 294–296 past performance, 291 portfolio holdings analysis, 302 active managers, 68 active return, 349 arithmetic and geometric form relationship, 349–350 European equity portfolio vs benchmark, 350–351 geometric active return, 349 active risk budgeting, 305–306 active risk puzzle, 307 investors, 308 misinterpretation of, 307 potential benefits of, 306 wisdom, of crowds, 305 active vehicles, 286 administrative fees, 656 advertising guidelines, 668–669 bonus investment company composites, 669–670 cumulative figures, 670 sample presentations, 669 agnostic, 214 investors list, 214 AIMR See Association for Investment Management and Research (AIMR) allocation portfolio, 554 See also global allocation portfolio; selection portfolio return of bond benchmark, 555 return of equity benchmark, 555 Ankrim and Hensel approach, 593 allocation effect, 594–595 attribution analysis, 595–597 currency premium significance, 597 currency return components, 596 data and returns for, 596 currency management effect, 594 forward premium effect, 593–594 interaction effect, 595 selection effect, 595 annualized return, 316 Apple (AAPL), 25 Arbitrage Pricing Theory (APT), 481 AS measure See average style measure (AS measure) asset classes, 286 asset management luck vs skill, 53 fund managers, types of, 8, 53–54 performance evaluation, 55 statistical error types, 53 statistical theory, 55 type II error, 54 zero-alpha manager, 54 asset manager fund manager types, Janus twenty fund, 51 mutual fund managers, 50 return reversals, 51 yearly asset class rankings, 52 asset pricing models, 3–4,11, 263 bond return models bond market factors, 22–23 five-factor model, 24 stock benchmarking procedure, 23 stock market factors, 22–23 stocks and bonds, returns on, 24 BTM, 12 CAPM joint hypothesis problem, 12 model estimation, 6–7 efficient markets, 10 behavioral finance academics, 11 Grossman-Stiglitz efficient, 10–11 QE, 10 four-factor model, 263 fund t-statistic, 263–264 inefficient stock markets, 11 asset pricing, 11 CRSP, 11 E/P ratio, 12 small capitalization stocks, 11 KTWW, 263–264 modern, momentum stocks, 14–15 comparing, 17 economic explanation, 17 monthly and cumulative momentum, 15–16 701 702 Index in U.S, 17 multifactor models, 18 next-period expected return, 19 next-period systematic risk, 19 next-period total risk, 19 small-cap factor, 18 time-series model, 18 problems, 24–25 small capitalization, 12–13 stock return models annual sort, 22 benchmarking approach characteristic, 21 Carhart model for CVX, 20–21 equity information, 21–22 four-factor model, 21 regression-based models, 20 stock characteristic-based model, 21 value-weighted returns, 22 theory-inspired single-factor model, 17–18 value stocks, 13–14 CAPM’s central prediction, 13–14 Fama-French’s ‘beta is dead’ slicing test, 13 stock return predictors, 14 asset-backed securities (ABS), 384–385 Association for Investment Management and Research (AIMR), 632 attribution analysis, 395 See also multi-period attribution analysis; single-period attribution analysis active return decomposition for DAX portfolio, 400 factor-based, 481–486 investment process components, 396 overall portfolio return, 396 net asset value, notation for, 397 segment contribution, 397 position-based approach, 479–480 problems, 487–490 quality, 396 return contributions for balanced portfolio, 398–399 individual stocks to active return, 400–401 overweight sector, 400 of portfolio segments, 396 transaction-based approach, 479–481 and valuation questions, 486–490 attribution analysis on three levels, 561–563 contributions after separating currency effects, 566–567 on MSCI-sector, 566, 568 contributions without separating currency effects, 563–564, 567 allocation contribution, 564 GAC, 565–566 selection contribution, 564 IT shape and share, 565 on single security level, 566 implementation, 567–568 attribution analysis on two levels contributions after separating currency effects, 560, 562 benchmark equity return segment, 560 portfolio terms, 561 contributions without separating currency components, 550–560 allocation portfolio, 554–555 global allocation portfolio, 551–554 IT, 556–560 selection portfolio, 555–556 attribution analysis with derivatives balanced portfolio, 625 cost of carry, 603 basis risk, 603 DAX-future, 604 forward contract, 604 currency premium and surprise, 591 Ankrim and Hensel approach, 593–597 Singer and Karnosky approach, 597–603 DAX in balanced portfolio, 626–628 interest rate swaps, 628 limitations and extensions, 627–628 options and valuation, 620 European call option, 621 European option, 620 geometric Brownian motion, 621 OTC-options, 620 riskless portfolio, 622–623 sensitivity quantifies, 622 options on single stock, 623–625 options on stock indices, 625–627 problems, 629 single-stock futures, 604–605 attribution analysis terms, 608 employment of stocks futures, 609–610 future transactions, 608 initial weights and returns for, 609 portfolio structure, 609 prices of DAX-08 futures contract, 605 single future position, 605–608 transaction impact in, 606 stock market index futures, 610–620 transaction, impact in, 610 swamps, 628 average style measure (AS measure), 119, 146–147 B BAI method, 346–347 balanced mutual fund, 70–71, 82 Bank Administration Institute method See BAI method Barbell strategy, 494–495 Barclays capital indices, 384–385 baseline bootstrap tests economic value-added alpha levels, 195–197 mutual fund alphas, crosssection of, 188–191 procedure, 186–187 baseline models, 33, 35 basis risk, 603 Bayes theorem, 282 Bayesian investor, 283 Bayesian methods, 41–42 Bayesian predictive distribution, 215 behaviorally un-humble manager (BUM), 304 Index benchmark, 28–29, 367–368 peer group benchmarks, 29 portfolio of securities, 368 specifications, 216–217 unpriced benchmarks, 40–41 BHB See Brinson, Hood, and Beebower (BHB) bond indices, 380–384 Barclays capital indices, 384–385 bond index calculation, 386 Citigroup indices, 386 government bond index series, 384–386 Markit iBoxx EUR benchmark indices, 385 bond return models See also multifactor models; stock return models bond market factors, 22–23 five-factor model, 24 stock benchmarking procedure, 23 stock market factors, 22–23 stocks and bonds, returns on, 24 book-to-market (BTM), 12, 68, 229–232 bootstrap evaluation of fund alphas implementation, on U.S Domestic Equity Mutual Funds, 184–187 baseline bootstrap procedure, 186–187 bootstrap extensions, 187 rationale, 181 individual mutual fund alphas, 181 mutual fund alphas, crosssection of, 181–184 bootstrap extensions, 187 bootstrap technique, sophisticated, 54 BR See Burke ratio (BR) Brinson approach, 421–422, 549 Brinson, Hood, and Beebower (BHB), 106 Brinson-type analysis, 426–427 BTM See book-to-market (BTM) BUM See behaviorally un-humble manager (BUM) Bundled Fee, 660–661 Burke ratio (BR), 585 Burnie method, 471 additivity property, 474–475 overall selection contribution, 472–473 selection contribution comparison, 473 C Calmar ratio, 584 capital asset pricing model (CAPM), 5, 57, 597–598 joint hypothesis problem, 12 model estimation, 6–7 regression graph for CVX, Sharpe’s, 5–6 capitalization weighted indices, 372 calculating market capitalization indices, 372–373 DAX index formula, 376 Deutsche Börse DAX indices, 380 equation derivation, 373–376 major capital markets, 381 major US equity indices, 381–382 MSCI regional indices, 380 select emerging markets, 382–383 CAPM See capital asset pricing model (CAPM) Carhart model See four-factor model Carhart regression approach, 120 Cariño approach, 462, 465–466, 475–476 carve-outs, GIPS, 647 attribution rules, 647–648 cash allocation, 648 cash segment assignment, options to, 648–649 formation, 646, 649–650 mixed portfolio data, 649 tracking-error-limits of portfolios, 647 CDA dataset, 149 CDs See certificates of deposit (CDs) CEFs See closed end funds (CEFs) Center for Research in Security Prices (CRSP), 11, 108–109, 149, 217, 264 certificates of deposit (CDs), CFA Institute See Association for Investment Management and Research (AIMR) characteristic selectivity measure (CS measure), 108, 137–144 benchmarking bonds, 144 benchmarking stocks, 140–141 BTM, 137–138 caveat, 109 differences in alphas, 118 stockholding-based CS alpha, 109–118 style drift sources, 138 Exxon, 138–139 large pension fund, 139–140 large university endowment, 139–140 U.S mutual funds CS measurement comparison, 142–143 CS measurement time series, 144–145 fund manager performance, 141–142 value-weighted returns, 109 characteristic timing measure (CT measure), 118–119, 144–146 Chevron-Texaco (CVX), CAPM regression graph for, four-factor regression for, 20 restricted OLS CAPM regression output for, two-factor regression for, 18 unrestricted OLS CAPM regression output for, 8–9 Cici and Gibson method, 126–127 Citigroup indices, 386 clean price value, 504 closed end funds (CEFs), 293 CM measure See Copeland and Mayers measure (CM measure) commercial mortgage-backed securities (CMBS), 384–385 composite benchmark See customized benchmark composite return accrual accounting, 652 calculation methodology, 653–655 GIPS guidelines for portfolio valuation, 651 GIPS requirements and recommendations, 652–653 GIPS valuation recommendations, 652 standards for portfolio valuation, 650 conditional holdings-based performance measurement, 127–133 703 704 Index See also unconditional holdings-based performance measurement advantage, 130 conditional measurement, 127–128 CWM estimation, 130 FK conditional approach, 129–131 UWM estimation, 130 conditional weight measure (CWM), 131 See also unconditional weight measure (UWM) constant relative risk aversion (CRRA), 34 Copeland and Mayers measure (CM measure), 87–88 Coupon effect, 511 cross-sectional dependence actual pairwise correlations, 262–263 advantages, 262 average correlation, 262 managed funds, 262 cross-sectional t-statistic, 254, 271, 280 CRRA See constant relative risk aversion (CRRA) CRSP See Center for Research in Security Prices (CRSP) CS measure See characteristic selectivity measure (CS measure) CT measure See characteristic timing measure (CT measure) currency forward, 591 currency premium, 592 using equations, 592 safe currency return, 592 currency surprise, 592 currency forward contract, 592 relationship, 593 customized benchmark, 391–392 CVX See Chevron-Texaco (CVX) CWM See conditional weight measure (CWM) D Daniel and Titman approach, 120 data, 187 DAX index formula, 376 DGTW method, 107–108 AS measure, 119 CS measure, 108–118 CT measure, 118–119 factor-based regression approach comparison, 120–121 gross return, 119 normalizing, for turnover, 122–123 performance measure correlation, 124–126 subportfolio performance, 121–122 Dietz method, 346–347 dirty price value, 504 discretionary portfolio identification, 638 asset manager’s work, 641–642 composite creation in connection with tax restrictions, 640 composite formation with restrictions, 639 fund inflow and outflow effect, 642 guidance statement, 640 non-fee paying portfolios, treatment of, 642–643 portfolio elimination, 641 restrictions, 639–640 DMT See duration-matched treasury (DMT) dogmatist, 212– 222 investors list, 214 no-predictability dogmatist (ND), 212 predictability dogmatist (PD), 212 Dow Jones industrial average type, 370 adjustment factor, 371 calculation formula, 371 constituents, 371 by share prices, 372 stocks of General Electric, 372 stocks of General Motors, 372 duration-based approaches attribution model, 537–538 Barra enterprise performance, 538–539 Lord’s approach, 537, 539 duration-matched treasury (DMT), 537 E earnings-to-price ratio (E/P ratio), 12 EFFAS See European Federation of Financial Analysts’ Societies (EFFAS) efficient markets, 289 funds benefits, 289–290 EONIA See Euro Overnight Index Average (EONIA) equal weighting (EW), 154 equity funds, 67 equity indices Dow Jones industrial average type See Dow Jones industrial average type market capitalization weighted indices, 372 calculation, 372–373 DAX index formula, 376 Deutsche Börse DAX indices, 380 emerging markets selection, 382–383 equation derivation, 373–376 major capital markets, 381 major US equity indices, 381–382 MSCI regional indices, 380 equity mutual funds optimal portfolios, 219 fund risk loadings, 222 mutual funds under CAPM, 220–221 NA, 222 ND, 222 predictable manager skills, 223 Sharpe ratio, 218–219 EURIBID See Euro Interbank Bid Rate (EURIBID) EURIBOR See Euro Interbank Offered Rate (EURIBOR) Euro Interbank Bid Rate (EURIBID), 387 Euro Interbank Offered Rate (EURIBOR), 387 interest rates, 387 Euro Overnight Index Average (EONIA), 387 performance index, 388 European Federation of Financial Analysts’ Societies (EFFAS), 661–662 EW See equal weighting (EW) Exxon-Mobil (XOM), 24–25 F factor analysis via Wilshire axiom, 482–484 Brinson-type analysis, 484 Index ‘Global 6-Regions Equity’model, 483 overweight-position, 484 portfolio and benchmark structure, 482 residual return estimation, 485 factor-based attribution analysis advantage, 485 APT model, 481 for bond portfolios, 485–486 macroeconomic models, 481 microeconomic models, 481 via Wilshire axiom See factor analysis via Wilshire axiom factor-based regression approach Carhart regression approach, 120 Daniel and Titman approach, 120 DGTW measures comparison with, 120 Fama-French/Carhart approaches, 121 multi-factor returns-based performance, 120 false discovery rate approach (FDR approach), 45–46, 252, 257 active fund managers, 46–47 advantages, 252–253 alpha p-value distribution, 43, 45 cross-sectional dependence actual pairwise correlations, 262–263 advantages, 262 average correlation, 262 managed funds, 262 estimators, comparison, 260–261 FDR approach, 260 FDR estimators, 261 full luck approach, 259 managed fund performance, luck impact on, 253–263 estimation procedure, 257 in multiple fund setting, 253 measuring luck, 255–256 problems, 284 proportion of zero-alpha funds, 260 U.S domestic equity mutual funds asset pricing models, 263 Bayesian interpretation, 282 data, 264 luck impact, 266, 270 performance measurement, 278–279, 281 performance persistence, 274 Fama-French approach ‘beta is dead’ slicing test, 13 and Carhart approach, 121 with DGTW, comparison of, 121 as dogmatist approach, 210–211 and four-factor model, 120 in net measuring security, 148 three-factor model of, 184–185, 226 12-industry classification, 234 FASB See Financial Accounting Standards Board (FASB) FDR approach See false discovery rate approach (FDR approach) Ferson-Khang conditional approach (FK conditional approach), 129 Financial Accounting Standards Board (FASB), 651 Finanacial Times Stock Exchange index See FTSE index firing managers, 293–294 firm, 636 predecessor firm, performance history of, 638 transparency requirements, 637 fixed income portfolios, attribution analysis for interest rate structure determination, 523–532 investment processes forward rates and implicit forecasts, 497–498 methodology for equity portfolios, 498–503 yield curves, 491–497 OAS valuation approach, 503–523 fixed-income funds, 69–70 FK conditional approach See Ferson-Khang conditional approach (FK conditional approach) flat yield curve, 491 forward premium See currency premium forward premium effect, 593–594 forward rates, 497–498 discount factors, 497–498 market implicit forecasts, 498 four-factor model, 20–21, 38–39, 67 FTSE index, 373 FTSE Global Bond Index Series, 380–384 ‘fund itself’ characteristics, 300–301 fund manager selection managed fund returns agnostic, 214 dogmatist, 212 optimal portfolios, 215 skeptic, 212–213 problems, 242–243 U.S domestic equity funds, 216–219 optimal portfolios of equity mutual funds, 219 out-of-sample performance, 223 superior predictability-based performance, 229–241 survivorship bias, 241 fund t-statistic cross-sectional distribution of, 194–195 heteroskedasticity, 263–264 normal distribution, 253–255 Type II error, 54 fund universes, 388–389 G GAC See global allocation term (GAC) geometric active return, 349 geometric form, attribution analysis in multiplicative, 471, 474 Burnie method, 471 additivity property, 474–475 overall selection contribution, 472–473 selection contribution comparison, 473 Cariño approach, 475–476 Menchero approach, 475–476 geometric attribution analysis, 477–478 portfolio and benchmark data, 477 mutatis mutandis, 476–477 geometric linkage, 315 German Blue Chip Equity, 669 GICS See Global Industry Classification Standard (GICS) 705 706 Index GICS sector classification, 379 GIPS See Global Investment Performance Standards (GIPS) global allocation portfolio, 549–552 See also allocation portfolio; selection portfolio bond segment, 553–554 equity, 553–554 GAC, 551 portfolio return, 553 RGAP, 551 global allocation term (GAC), 551 Global Industry Classification Standard (GICS), 379 Global Investment Performance Standards (GIPS), 631 advertising guidelines, 668–669 bonus investment company composites, 669–670 cumulative figures, 670 sample presentations, 669 AIMR-PPS, 632–633 beneficiaries, 632 carve-outs See carve-outs, GIPS composite maintenance, 671 composite return accrual accounting, 652 calculation methodology, 653–655 GIPS guidelines for portfolio valuation, 651 GIPS requirements and recommendations, 652–653 GIPS valuation recommendations, 652 standards for portfolio valuation, 650 correct application of standards, verification of, 671–672 criteria for portfolio allocation to composites, 643 alternative breakdown of DAX-portfolios, 645 for composite creation, 643 by investment objective, 645 by investment style, 645 presentation of management performance, 644 by tracking error limit, 645 discretionary portfolio identification, 638 asset manager’s work, 641–642 composite creation, in connection with tax restrictions, 640 composite formation, with restrictions, 639 fund inflow and outflow effect, 642 guidance statement, 640 portfolio elimination, 641 restrictions, 639–640 non-fee paying portfolios, treatment of, 642–643 equity composite ‘German blue chips’ of malus-invest, 681 firm, 636 performance history, of predecessor firm, 638 transparency requirements, 637 gross vs net return, 655–661 administrative fees, 656 annualized return comparison, 658 approximation methods, 659 bundled fee, 660–661 fees and/or costs types, 655 gross-of-fees return, 656–657 management fee, 657 net returns to gross returns, 660 relationship, 657 TER, 661 time series adjustment of unit price, 658 homogeneity measurement of investment process, 672–678 asset-weighted standard deviation, 674 global equity portfolios composite, 677 high-low range, 673 measurement combination of dispersion, 676–678 quartile representation, 674–676 standard deviation, 673–674 nations with country sponsors for, 633 performance presentation standards, 634 composite definition, change of, 635 minimum period for presentation, 635 of selected portfolios presentation, 634 of selective periods presentation, 635 survivorship bias, 634 presentation of performance results EFFAS recommendations, 662 euro’ basis, 663 by intermediate currency, 664–665 lira/DM time series conversion, 664 official euro exchange rates, 662 problem in, 661 problems, 680–682 requirements and recommendations, 665–666 presentation of, 678–682 risks, 678–682 selective periods, 634 supplemental information, 666–668 turnover coefficient for portfolio/composite, 667–668 ‘Global 6-Regions Equity model, 482 contributions to, 483 Grinblatt and Titman measure (GT measure), 88–89 gross-of-fees return, 656–657 Grossman-Stiglitz efficient, 10–11 GT measure See Grinblatt and Titman measure (GT measure) H hedge fund analysis risk measures based on drawdown Burke ratio, 585 Calmar ratio, 584 drawdown, 583 high watermark, 584 lake ratio, 586–587 local maximum and local minimum loss, 585 Martin ratio, 586 MD, 584 motivation, 583 pain index, 586 SterR, 584 time to recovery, 584 ulcer index, 585, 587 for short position portfolios, 583 Index holdings based performance evaluation, 84 benchmarking, 85 conditional measurement, 127–133 FK conditional approach, 129–131 fund style-orientation, 85 holdings data, 85 portfolio holdings analysis, 85 returns-based methods, 84–85 unconditional measurement, 86–127 Cici and Gibson method, 126–127 DGTW method, 107–126 self-benchmarking method, 87–107 I IAMs See inferior active managers (IAMs) IASB See International Accounting Standard Board (IASB) ICB sector classification, 378 IID See independently and identically distribution (IID) incremental return/risk ratio (IRRR), 306 independently and identically distributed (IID), 182–183 index users, 367–368 indices, 368 benchmarks for portfolios balanced portfolios, 391 combined equity and bond investment, 392 customized benchmark, 391–393 investment risk, 391 multi-asset classes, 391 return and risk of, 392 bond indices, 380–384 Barclays capital indices, 384–385 calculating bond index, 386 Citigroup indices, 386 J.P Morgan government bond index series, 384–386 Markit iBoxx EUR benchmark indices, 385 construction, 369–370 DAX, 370 equity indices Dow Jones industrial average type, 370 market capitalization weighted indices, 372 fund universes, 388–389 group of securities, 369 investment guidelines, 369 money market indices, 386–388 EONIA indices, 387 EURIBOR indices, 387 LIBOR indices, 387–388 MSCI, 370 peer group comparisons, 388–389 bar chart, 390 DPG peer group comparison for, 390 investment objective, 389 peer group median, 389 portfolio’s position, 389 problems, 393 security indices, 369 individual mutual fund alphas, 181 industry allocation analysis Fama-French 12-industry classification, 234 industries, 234 information variables, 237 ND investor, 237 of optimal portfolios, 235–236 predictability-based strategies, 237 time-series average allocations, 234–237 industry attribution analysis, 237–239 alphas, 240 industry passive return, 240 industry timing return, 240 industry-adjusted net return, 241 industry-level regressions, 241 industry-level returns, 240 sources, 241 inferior active managers (IAMs), 290 information ratio, 74 performance measure, 74 advantages, 75 alternative definition, 75 CAPM, 76 foundations of, 76 by using inefficient index, 76–77 information ratio investing, 76– 77 IR practice levels, 79 M-V optimizers, 78 during manager tenure, 71, 79 measurement time period, 75–76 Miller’s IR, 80 solving equations, 78 informed managers, efficient frontiers for, 13 institutional funds, 286 institutional separate account (ISA), 287, 293 interaction term (IT), 406–408, 556, 565 allocation and selection portfolio, 420 allocation factors, 419–420 attribution analysis for portfolio, 422, 425 properties of, 421 balanced portfolio equities, 557–558 for bond segments, 556 for individual segments, 557 initial weights and returns, 418–419 interpretation uses, 424–425 investment process, 425–426 and multiplicative linkage, 469 portfolio manager, choices of, 420–421 sectorial allocation decision, 556 selection and allocation decisions, 424, 556 and interaction combination, 423–426 selection contribution and, 421–422 signs, 417–418 weights and returns for mixed benchmark, 557–558 interest coupon payments, 510–511 interest rate structure, 491 butterfly factor, 531 Nelson and Siegel method, 525 principal component analysis, 529–530 shift and slope method, 527–528 shift factor, 531 Svensson’s method, 526, 529 yield curve, 524 707 708 Index internal rate of return, 319 cash flow, 318 computation in one withdrawal, 317–318 with one contribution, 316–318 time-weighted return computation, 321 International Accounting Standard Board (IASB), 651 inverted yield curve, 491 investment guidelines, 369 Investment Performance Council (IPC), 633, 678 investment process asset-weighted standard deviation, 674 benchmark bond contribution, 502–503 bond classification, 500 bond portfolios, 498–500 Brinston’s attribution analysis, results of, 500–502 components, 396 global equity portfolios composite, 677 high-low range, 673 homogeneity measurement, 672–678 interaction term, 425–426 measurement combination of dispersion, 676–678 portfolio and benchmark data, 499 quartile representation, 674–676 standard deviation, 673–674 yield curves change, 500, 502 investor primarily holds index funds, 222 investor returns, 315 IPC See Investment Performance Council (IPC) IRRR See incremental return/risk ratio (IRRR) ISA See institutional separate account (ISA) IT See interaction term (IT) J J.P Morgan government bond index series, 384–386 Janus twenty fund, 51 Jensen alpha See single-factor alpha Jensen model See CAPM model K Kirievsky and Kirievsky methodology, 448–451 Kosowski, Robert, Timmermann, Allan, Wermers, Russ, White, Halbert (KTWW) See KTWW, asset pricing models KTWW, asset pricing models, 263–264 L Lagrange multiplier method, 463–464 lake ratio, 586–587 left tail funds, 266–268 LIBOR See London Interbank Offered Rate (LIBOR) linking algorithms attribution analysis results, 466 Cariño’s approach, 462, 465–466 disadvantage, 470 Lagrange multiplier method, 463–464 Menchero’s approach, 462, 465 multiplicative linkage, 467–471 nimble approach, 463–464 portfolio and benchmark data, 467 statistical considerations, 466–467 systematic analysis, 470 Zhang approach, 463 London Interbank Offered Rate (LIBOR), 387–388 interest rates, 387 Lord’s approach, 537 M M-V optimizers See mean-variance optimizers (M-V optimizers) Macauley duration, 540 macroeconomic forecasting, 294–296 average active manager, 296 findings summary, 298, 303, 305 fund’s alpha, 296 macro forecasting area, 296 macro-forecasting strategy, 298 macro-timing strategies, 298 superior insights, 296 variables, 297–298 managed fund performance, luck impact on, 253–263 estimation procedure bootstrap procedure applying, 258–259 cross-sectional t-distribution, 257 FDR approach, 257 p-values estimation, 257–258 unlucky and lucky funds, proportions of, 259 multiple fund setting, luck in, 253 individual fund t-statistic distribution, 254–255 measure performance, 255 skill groups, 253–255 skilled funds, 253 unskilled funds, 253 zero-alpha funds, 253 measuring luck, 255–256 fund populations, 256–257 location of truly skilled, 256 probability of zero-alpha fund, 256 managed fund returns, 210 agnostic, 214 investors list, 214 comparing, 211 dogmatist, 212 investors list, 214 ND, 212 PD, 212 dynamic model for, 212 generating model, 210 investors, 212 optimal portfolios, 215 Bayesian predictive distribution, 215 optimal ICAPM portfolio, 215–216 optimization, 215 predictive moments, 216 unconstrained optimal portfolio, 216 predictable fund risk-loadings, 210–211 skeptic, 212–213 investors list, 214 NS investor, 213 PS investor, 213 superior performance, 210 timing-related fund returns sources, 210–211 Index management fee, 657 manipulation-proof performance measure (MPPM), 32 Markit iBoxx EUR benchmark indices, 385 Martin ratio (MR), 586 maximum drawdown (MD), 584 MBS See mortgage-backed securities (MBS) McLaren approach, 535–537 MD See maximum drawdown (MD) mean squared error (MSE), 258–259 mean-variance optimizers (M-V optimizers), 78 Menchero approach, 462, 465, 475–476 geometric attribution analysis, 477–478 portfolio and benchmark data, 477 Miller’s alpha, 68 mixed equity funds, 70 modern asset pricing models capital asset pricing model, CAPM model estimation, 6–7 applying rules of variances, 7–8 CVX, fitted model, potential problems, 10 regression graph for ChevronTexaco, regression output, 9–10 restricted OLS, unrestricted OLS, 8–9 Sharpe’s CAPM, 5–6 unrealistic assumptions, 5–6 modified Dietz method account statement, 341–343, 345–346 application within one period, 341 exogenous cash flows, 341 as linear approximation to internal rate of return, 343 modification, 345 order of magnitude of error, 342–343 relationship with internal rate of return, 344 Mey market indices, 386–388 EONIA indices, 387 EURIBOR indices, 387 LIBOR indices, 387–388 Mogage-backed securities (MBS), 384–385 MPPM See manipulation-proof performance measure (MPPM) MR See Martin ratio (MR) MSCI indices, 370 regional indices, 380 MSE See mean squared error (MSE) multi-asset class portfolio analysis Brinson approach, 549 decision level, 549–550 global allocation, 549 multi-period attribution analysis See also single-period attribution analysis basic formula I approach Kirievsky and Kirievsky methodology, 448–451 variant 1, 446–448 variant 2, 448–454 basic formula II approach Valtonen linking algorithm, 457–460 Wilshire linking algorithm, 454–457 basic formula III approach Brinson analysis in multiperiod case, 460–461 scaling selection and allocation contributions, 461–462 basis formulas for linking contributions, 439, 442, 445 generalized overall contribution, 443–445 individual methodologies, 445 multi-period benchmark return RB, 441–442 return representation as weighted sum of contributions, 441 for three periods, 443 for two subperiods, 443 linking algorithms attribution analysis results, 466 Cariño’s approach, 462, 465–466 disadvantage, 470 Lagrange multiplier method, 463–464 Menchero’s approach, 462, 465 multiplicative linkage, 467–471 nimble approach, 463–464 portfolio and benchmark data, 467 statistical considerations, 466–467 systematic analysis, 470 Zhang approach, 463 multifactor models, 18 See also bond return models; stock return models next-period expected return, 19 next-period total risk, 19 small-cap factor, 18 time-series model, 18 multiple-factor alpha, 66–67 See also single-factor alpha active managers, 68 balanced mutual fund, 70–71, 82 benchmarks, 67–68 capitalization groups, 68 coefficients, 71 equity funds, 67 fixed-income funds, 69–70 four-factor alphas during manager tenure, 68–69 four-factor model, 67 Miller’s alpha, 68 mixed equity funds, 70 multistrategy funds, 72 seven factor model, 72 SMB, 68 stock market factors, 70 multiplicative linkage interaction terms and, 469 portfolio and benchmark data, 469 portfolio level results, 470 multistrategy funds, 72 mutual fund managers, 50 N NA See no-predictability agnostic (NA) National Association for Pension Funds (NAPF), 632 ND See no-predictability dogmatist (ND) Nelson and Siegel method, 525 Neuer Markt indices (NEMAX), 367–368 709 710 Index Nikkei 225, 370 Nimble approach, 463–464 no-predictability agnostic (NA), 214, 222 no-predictability dogmatist (ND), 212, 222, 226 no-predictability skeptic (NS), 212–213 non-regression approaches Sharpe ratio, 56 agency problem, 58 assumptions, 57 benefits, 56–57 calculations, 59 CAPM, 57 GISW, 59 during manager tenure, 58 manipulating, 59 mean and standard deviation, 57–58 portfolio, 57 portfolio managers, 58 risk and expected return, 58 tracking error, 60 definitions, 60 example, 62 income funds, 62 inefficiency induced by, 15 manipulating, 62 substitute tracking-error gain, 60 TEV, 60 usage, 61 non-stable regression parameters, 39–40 NS See no-predictability skeptic (NS) O OAS See option-adjusted spread (OAS) OLS See ordinary least squares (OLS) option-adjusted spread (OAS), 504 ordinary least squares (OLS), 6–7, 38–39 OTC See over the counter (OTC) OTC-options, 620 out-of-sample performance, 223 by CAPM, 226 experiments, 227 fund risk loadings, 226 hot-hands (H-H) strategy, 223, 227 ND, 226 by NBER characterization, 229 of optimal portfolio strategies, 228–229 of portfolio strategies, 223–226 skew, 223–226 skewness level, 227 strategies performance, 226–227 subperiods, 227–229 outperform retail funds, 286 over the counter (OTC), 620 P PA See predictability agnostic (PA) pain index (PI), 586 par yield curve, 496 parallel effect, 508, 512 passive management approach attribution analysis with split-off, 433 for 2-country-portfolio, 430–435 currency contribution separation, 430–435 local currency and base currency, 432 portfolio data, 430 return contributions, 431 return in local currency and in base currency relationship, 432 2-country-portfolio data, 434 PD See predictability dogmatist (PD) peer group comparisons, 388–389 bar chart, 390 DPG peer group comparison for, 390 investment objective, 389 peer group median, 389 portfolio’s position, 389 performance measures correlation, 124–126 performance persistence, 203–208 Performance Presentation Standards (PPS), 632 performance-decomposition methodology, 136–137 AS measure, 146–147 CS measure, 137–144 CT measure, 144–146 measuring net return selectivity, 148 trade execution costs, 147–148 PHM See portfolio holdings measures (PHM) PI See pain index (PI) portfolio allocation criteria, to composites, 643 according to investment style, 645 tracking error limit, 645 alternative breakdown of DAXportfolios, 645 for composite creation, 643 by investment objective, 645 management performance, presentation of, 644 portfolio holdings measures (PHM), 86–87 portfolio strategy attributes, 229 agnostics strategies, 233 characteristic-based benchmarks, 234 lagged net inflows, 233 mutual fund types, 232 optimal portfolios, 230–231 portfolio holding attributes, 229–232 predictability-based strategies, 233 predictable skills, 232–233 skeptics strategies, 233 PPS See Performance Presentation Standards (PPS) predictability agnostic (PA), 214 predictability dogmatist (PD), 212 predictability skeptic (PS), 212–213 price-weighted indices See Dow Jones industrial average Prime Construction Performance Index, 414–415 PS See predictability skeptic (PS) pull-to-par effect, 508 Q quantitative easing (QE), 10 quarterly currency returns, yen to euro, 593 questions and answers (Q&As), for asset manager, 632 quintile representation, 674–676 sorts, 137–138 R regression-based models, 20 regression-based performance measures, 62 See also returns-based performance measures Index conditional regression models, 73 Jensen model equation, 74 macroeconomic variables, 73–74 information ratio performance measure, 74 advantages, 75 alternative definition, 75 CAPM, 76 foundations of, 76 inefficient index, 76–77 information ratio investing, 77–78 IR practice levels, 79 M-V optimizers, 78 during manager tenure, 71, 79 measurement time period, 75–76 Miller’s IR, 80 solving equations, 78 multiple-factor alpha, 66–67 active managers, 68 balanced mutual fund, 70–71, 82 benchmarks reveals, 67–68 capitalization groups, 68 coefficients, 71 equity funds, 67 fixed-income funds, 69 fixed-income funds, 70 four-factor alphas during manager tenure, 68–69 four-factor model, 67 Miller’s alpha, 68 mixed equity funds, 70 multistrategy Funds, 72 seven factor model, 72 SMB, 68 stock market factors, 70 potential bias in, 23 selectivity performance measures, 73 single-factor alpha, 62 benchmark choice, 64 bias with incorrect benchmark, 65 comparing, 66 equal-weighted benchmark, 64–66 estimation error, 64 finding, 63 during manager tenure, 63 performance measures geometry, 16 security market line, 16 Treynor’s ratio as performance measure, 66 value-weighted benchmark, 65–66 timing performance measures, 72–73 return calculation, 513, 518 account statement without cash flows, 313–314 basis formula for, 313 different segments effects, 522–523 investor returns, 315 return of investment, 313–314 weights and local returns, 522–523 return gap, 135 return of investment, 313–314 return smoothing, 37–38 return-on-equity (ROE), 13–14 returns-based analysis alpha p-value distribution, 43, 45 alpha variation, 43 baseline models, 33, 35 Bayesian methods, 41–42 Carhart model, 38–39 FDR, 45–47 goal, 34 models with timing factors, 36–37 non-normal alphas, 38 non-stable regression parameters, 39–40 performance measurement, 42–44 problems, 34–35 return smoothing, 37–38 SDF, 44–45 timing-related biases, 36 unpriced benchmarks, 40–41 returns-based performance evaluation model benchmark, 28–29 manager types, 33 MPPM, 31–32 peer group benchmarks, 29 performance measure properties, 30 continuous, 30 exhibit monotonicity, 30 fit, 30 fund scoring, 31 scalable, 30 risk-aversion role, 34 type I error, 32–33 type II error, 32–33 returns-based performance measures goal, 55 luck vs skill, 53–55 non-regression approaches Sharpe ratio, 56 tracking error, 60 problems, 55, 80 risk adjusted attribution analysis Ankrim approach, 578 betas for stocks, 580 return of risky asset, 579 without separating currency effects, 576, 581 based on information ratio, 581 active return, 582 information ratio, 582 tracking error, 582 based on systematic risk, 569 Brinson approach, 576, 579 Brinson process, 569–570 intuitive derivation via investment process, 570 allocation beta, 572 allocation on segment level, 575 balanced portfolio, 570 interaction beta, 572 portfolio beta and Jensen’s alpha relationship, 573 resulting betas and overall beta relationship, 570 security level data, 577 selection beta, 572 without separating currency effects, 575–576 single-index model, 570 risk budgeting, 305 risk factors, 17–18 risk-averse investors, 57 ROE See return-on-equity (ROE) roll-down effect, 506–507, 511–512 S SAMs See superior active managers (SAMs) SDF See stochastic discount factor (SDF) second-order Newton approximation method, 347–349 711 712 Index security indices, 369 security parameters, 76 selection portfolio, 555 See also allocation portfolio; global allocation portfolio bond segments, 555 individual equity, 555 self-benchmarking method benchmark error, 90 BHB approach, 106–107 CM approach, 87–88 GT measurement, 88–90 GT performance evaluation, 91, 96–99 mutual fund performance estimate, 104–105 stockholding-based returns, 100–103 sensitivity analysis, 197 cross-sectional bootstrap, 198–199 data records length, 199 estimated alphas vs bootstrapped alpha, 192–193 residual and factor resampling, 198 stockholdings-based alpha bootstrap test, 200–202 time-series dependence, 197–198 seven factor model, 72 Sharpe measure See Sharpe ratio Sharpe ratio (SR), 56, 223–226 agency problem, 58 assumptions, 57 benefits, 56–57 calculations, 59 CAPM, 57 GISW, 59 during manager tenure, 58 manipulating, 59 mean and standard deviation, 57–58 portfolio, 57 portfolio managers, 58 risk and expected return, 58 simple bootstrap procedure, 258–259 Singer and Karnosky approach, 597–598 active currency selection, 601 allocation, 601 attribution analysis, 599–600, 602 hedge selection, 601 interaction, 601–602 selection, 602 strategic options, 598 U.S investor, 598 single-factor alpha, 62 See also multiple-factor alpha benchmark choice, 64 bias with incorrect benchmark, 65 comparison, 66 equal-weighted benchmark, 64–66 estimation error, 64 finding, 63 during manager tenure, 63 performance measures geometry, 16 security market line, 16 Treynor’s ratio as performance measure, 66 value-weighted benchmark, 65–66 single-period attribution analysis See also multi-period attribution analysis active management approach currency contribution separation, 435–437 portfolio and benchmark data, 436 allocation, 402–403 analysis for DAX-portfolio, 411–412 analysis of portfolio, 410 attribution analysis components, 409 contributions over single period, 408 decisions, 403–404 manager, 404–405 portfolio, 404 weightings within benchmark, 403 weightings within portfolio, 404 components, 409 contribution breakdown on single-stock-level, 430 allocation contribution, 428–430 attribution analysis of DAXportfolio, 429 Brinson-type analysis, 426–427 interaction terms, 428–430 DAX breakdown into sectors, 402 dealing with titles with zero weight, 414 allocation contribution calculation, 416 attribution analysis with stocks outside benchmark, 415–417 formula for allocation contribution, 415–417 Prime Construction Performance Index, 414–415 relative attribution analysis, 416 global performance attribution in CFA course, 437–439 interaction term See interaction term (IT) investment decisions, 404 overall return allocation portfolio RAP, 405–406 benchmark RB, 405–406 selection portfolio RSP, 406–408 passive management approach See passive management approach return contributions of DAX portfolio, 401–402 selection, 402–403 analysis for DAX-portfolio, 411–412 analysis of portfolio, 410 contributions over single period, 408 decisions, 403–404 manager, 404–405 portfolio, 404 weightings within benchmark, 403 weightings within portfolio, 404 transaction cost effect, 413–414 weighting of DAX sectors, 403 skeptic, 212–213 investors list, 214 NS investor, 213 PS investor, 213 skilled funds, 253 skilled manager, 33, 53–54 See also unskilled manager spread change effect, 510, 512 Index SR See Sharpe ratio (SR) Sterling ratio (SterR), 584 stochastic discount factor (SDF), 44 alpha estimation, 44 using GMM, 44 limitation, 45 stock characteristic-based model, 21 stock market index futures transaction impact in, 610 underlying asset of futures contract different from equity benchmark, 614–620 euro bund future, 616–619 identical to equity benchmark, 610–614 interest rate futures, 616 return effects with, 619 stock return models See also bond return models; multifactor models annual sort, 22 benchmarking approach characteristic, 21 Carhart model for CVX, 20–21 equity information, 21–22 four-factor model, 21 regression-based models, 20 stock characteristic-based model, 21 value-weighted returns, 22 STOXX indices, 377–380 style drift sources, 138 Exxon, 138–139 large pension fund, 139, 140 large university endowment, 139–140 superior active managers (SAMs), 290 active risk budgeting, 305–306 active risk puzzle, 307 investors, 308 misinterpretation of, 307 potential benefits of, 306 wisdom of crowds, 305 approaches, 290–291 fund characteristics, 299 expense ratios, 301 hedge funds, 301–302 macro-based studies, 301 optimal risk allocations, 16 fund manager characteristics, 299–300 ‘fund itself’ characteristics, 300–301 fund management company, 300 manager alpha, 297 macroeconomic forecasting, 294–296 average active manager, 296 findings summary, 298, 303, 305 fund’s alpha, 296 macro forecasting area, 296 macro-forecasting strategy, 298 macro-timing strategies, 298 superior insights, 296 variables, 297–298 past performance assessing, 294 asset classes and vehicles, 293 CEFs, 293 competitive pressures, 291 firing managers, 293–294 ISAs, 293 non-normal returns, 292–293 persistence in, 13, 291–292 persistence studies, 294–295 using seven-factor model, 293 using style-adjusted returns, 291 portfolio holdings analysis, 302 contrarian managers, 304 publicly-disclosed holdings, 302 qualifications, 304–305 risk shifting funds, 302–304 superior predictability-based performance, 229 industry allocation analysis Fama-French 12-industry classification, 234 industries, 234 information variables, 237 ND investor, 237 of optimal portfolios, 235–236 predictability-based strategies, 237 time-series average allocations, 234–237 industry attribution analysis, 237–239 alphas, 240 industry passive return, 240 industry timing return, 240 industry-adjusted net return, 241 industry-level regressions, 241 industry-level returns, 240 sources, 241 portfolio strategies attributes, 229 agnostics strategies, 233 attributes of optimal portfolios, 230–231 characteristic-based benchmarks, 234 lagged net inflows, 233 mutual funds types, 232 notable findings, 232 portfolio holdings attributes, 229–232 predictability-based strategies, 233 predictable skills, 232–233 skeptics strategies, 233 survivorship bias, 241 conditional on realized fund returns, 241–242 swap curves, 497 synthetic cash, 626 T TER See total expense ratio (TER) TEV See tracking-error variance (TEV) time-weighted return, 322 account statement with one contribution and one withdrawal, 320 additive adjustment, 324 calculation, 322–323 dividend treatment, 323–324 investment funds returns, 322 multiplicative adjustment, 324–325 backward adjustment, 325–326 forward adjustment, 325 multiplicatively adjusted time series, 325, 328 unit price method, 324–326 time-weighted return and internal rate of return comparison with one contribution and higher return in first subperiod, 330–332 in second subperiod, 331–332 with one contribution and variable net asset value X, 333 713 714 Index with one withdrawal and higher return in first subperiod, 332 and variable net asset value X, 335 with positive returns in both subperiods, 336 returns and net asset values for account statement, 331–332 time-weighted return and internal rate of return, 335–336 and valuation X relationship, 333–334 time-weighted return approximation method computation, 338 active return, 349–352 based on internal rate of return, 339 Dietz method and BAI method, 346–347 exogenous cash flows, 337–338 internal rate of return, 339–340 modified Dietz method, 341–346 linear approximation to internal rate of return, 343–345 modification, 345–346 second-order Newton approximation, 347–349 subperiod between cash flows, 339 time-weighted return calculation, 337–338 without unique internal rate of return, 340–341 TNA See total net assets (TNA) total expense ratio (TER), 661 total net assets (TNA), 152 tracking error, 60 definitions, 60 example, 62 income funds, 62 inefficiency induced by, 15 manipulating, 62 substitute tracking-error gain, 60 TEV, 60 usage, 61 tracking-error variance (TEV), 60 Treynor’s ratio as performance measure, 66 U U.S domestic corporate mutual funds, 172–173 U.S domestic equity fund data, 216–217 CRSP, 217 dividend yield, 217 equity mutual funds optimal portfolios, 219–223 investment objectives, 217 no-load equity mutual funds, 218 out-of-sample analyses, 219 out-of-sample performance, 223–229 predictability, 218–219 return-history category, 217 superior predictability-based performance, 229–241 survivorship bias, 241 U.S domestic equity mutual funds, 264, 268–269 asset pricing models, 263 conditional four-factor model, 263 fund t-statistic, 263–264 KTWW, 263–264 average mutual fund to Vanguard Index 500 fund, comparison, 168–170 baseline mutual fund return decomposition, 163–168 Bayesian interpretation, 282 on mutual fund performance, 282–283 mutual fund selection, 283 optimal Bayesian decision, 283 benchmark-adjusted mutual fund returns, 157–161 correlation between performance measures, 161–163 data, 264 aggressive growth funds, 264–265 equally weighted portfolio of, 264–263 WRDS, 264 luck impact, on long-term performance, 266–267 average fund characteristics, 268 growth and aggressive funds, 268 left tail funds, 266–268 proportions of unskilled and skilled funds, 269 superior short-term alphas, 270 zero-alpha funds, 266 luck impact, on short-term performance, 270–271 BG model, 272–274 fund characteristics and performance dynamics, 273 short-term fund manager skills, 272 unskilled funds, 272 mutual fund returns, 154–157 performance measured with asset pricing models, 281 using Fama-French model, 282 loadings on omitted factors, 281 performance measured with preexpense returns, 278–279 luck impact, on long-term, 279–280 pre-expense skilled funds proportion, 279–281 stockpicking skills, 279 performance persistence, 274 average out-of-sample performance, 277 based on false discovery rate, 276 comparison, 277–278 false discovery rate, 274–277 portfolio alpha estimation, 277–278 portfolio turnover, 277 portfolios construction, 275 skilled funds, proportion of, 274 U.S domestic equity mutual funds, application CDA dataset, 149 CRSP database, 149 growth-oriented funds, 154 TNA, 152 yearly mutual fund universe statistics, 149–152 U.S equity funds bootstrap analysis, 188 individual fund alphas normality, 188 U.S mutual funds Index CS measurement comparison, 142–143 time series, 144–145 fund manager performance, 141–142 ulcer index (UI), 585–586 unconditional holdings-based performance measurement See also conditional holdingsbased performance measurement benchmark portfolio, 87 Cici and Gibson method, 126–127 DGTW method, 107–126 factor model, 87 hypothetical performance, 86 PHM, 86–87 self-benchmarking method, 87–107 unconditional weight measure (UWM), 131 See also conditional weight measure (CWM) uninformed managers, frontiers for, 13 unskilled funds, 253 unskilled manager, 33, 53–54 See also zero-alpha manager UWM See unconditional weight measure (UWM) V Valtonen linking algorithm See also Wilshire linking algorithm attribution analysis results, 459 family of linking formulas, 459 mutatis mutandis, 457–458 x-cumulative linkage method, 459 value stocks, 13–14 CAPM’s central prediction, 13–14 Fama-French’s ‘beta is dead’ slicing test, 13 stock return predictors, 14 value-weighted returns, 22 Van Breukelen approach, 532–535 Vanguard 500 Index Fund and average mutual fund, comparison, 168–170 four-factor model, 69 W Wharton Research Data Services (WRDS), 264 Wilshire linking algorithm See also Valtonen linking algorithm allocation contribution, 454–457 attribution analysis results, 457 contribution effects, 454–456 winner-minus-loser portfolio (WML portfolio), 216–217 WML portfolio See Winnerminus-loser portfolio (WML portfolio) WRDS See Wharton Research Data Services (WRDS) X XOM See Exxon-Mobil (XOM) Y yield curves, 492 curvature, 494–495 interest rate structure, 491 movement of EUR-German, 495 U.S treasury, 496 not appearing in standard forms, 492–493 parallel movement, 493 rotation, 494 swap curve vs U.S treasury curve, 497 yield curves based attribution analysis attribution relative to benchmark, 515 bond-specific effects, 505–506 characteristic movements, 513 coupon effect, 511 currency segments, 521 interest coupon payments, 510–511 interest rates, 505 OAS, 504 parallel effect, 508, 512 pull-to-par effect, 508, 512 return effects, 514–515 returns calculation, 513 roll-down effect, 506–507, 511–512 separation of IT, 520–521 spread change effect, 510, 512 structural effect, 509–510, 512 structural effect breakdown, 512–513 total yield of bond, 506 Z zero-alpha funds, 253 zero-alpha manager, 33 See also skilled manager Zhang approach, 463 715 ... economics KENNETH J ARROW and MICHAEL D INTRILIGATOR Performance Evaluation and Attribution of Security Portfolios by Bernd Fischer and Russell Wermers Academic Press is an imprint of Elsevier The Boulevard,... most important and widely used performance attribution techniques, which allow an accurate measure of all of the sources of investment returns, and which are necessary for precise performance reporting... Copeland and Richard Roll of UCLA and Josef Lakonishok of University of Illinois (and LSV Asset Management) for early inspiration, as well as Wayne Ferson, Robert Stambaugh, Lubos Pastor, and