Modern Portfolio Management E-Book

Martin L. Leibowitz

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Serie: Wiley Finance Editions
• Sprache: Englisch

Seitenzahl: 633

Veröffentlichungsjahr: 2009

Beschreibung

Active 130/30 Extensions is the newest wave of disciplined investment strategies that involves asymmetric decision-making on long/short portfolio decisions, concentrated investment risk-taking in contrast to diversification, systematic portfolio risk management, and flexibility in portfolio design. This strategy is the building block for a number of 130/30 and 120/20 investment strategies offered to institutional and sophisticated high net worth individual investors who want to manage their portfolios actively and aggressively to outperform the market.

Leseprobe

Table of Contents

Praise

Title Page

Copyright Page

Dedication

Foreword

BETA GRAZERS DRESSED UP LIKE ALPHA HUNTERS

THE CURRENT BUSINESS MODEL FOR ASSET OWNERS AND ASSET MANAGERS

THE FUNDAMENTAL LAW OF ACTIVE MANAGEMENT

CORPORATE GOVERNANCE AND 130/30 INVESTING

CONCLUSIONS AND A CAUTIONARY NOTE

NOTES

Structure of the Book

Acknowledgements

Introduction

THE EVOLUTION OF ACTIVE EXTENSION: THE CALPERS BREAKTHROUGH

BOTH QUANT AND FUNDAMENTAL STYLES

THE FUND-LEVEL CONTEXT

AE-PLUS GENERALIZATIONS

REFERENCES

PART One - Active 130/30 Extensions and Diversified Asset Allocations

CHAPTER 1 - Active 130/30 Extensions and Diversified Asset Allocations

ACTIVE MANAGEMENT WITH ALPHA RANKING MODELS

TRACKING ERROR MODELS

THE ACTIVE EXTENSION

TRACKING ERROR UNDER ACTIVE EXTENSION

INFORMATION RATIOS

USING GENERICS IN ACTIVE EXTENSION

POSITION SIZE CONSTRAINTS

BEYOND THE INFORMATION RATIO

FUND LEVEL RISK EFFECTS

PASSIVE IMPLICIT ALPHAS

BEYOND-MODEL DRAGON RISKS

ACTIVE ALPHAS

RISK AS RISK TO THE POLICY PORTFOLIO

CORRELATION TIGHTENING AND STRESS BETAS

SHORT-TERM RISK AND LONG-TERM RETURNS

THE ALPHA/BETA MATRIX

CONCLUSION

REFERENCES

PART Two - The Role of Quantitative Strategies in Active 130/30 Extensions

CHAPTER 2 - Active Extension—Portfolio Construction

A FRAMEWORK FOR ANALYZING ACTIVE BENCHMARK-RELATIVE PORTFOLIOS

THE STRUCTURE OF ACTIVE LONG-ONLY PORTFOLIOS

MOVING FROM LONG-ONLY TO ACTIVE EXTENSION

BENCHMARK REPRESENTATION

CONSTRUCTING ACTIVE EXTENSION PORTFOLIOS

NOTES

REFERENCES

CHAPTER 3 - Managing Active Extension Portfolios

MEASURING AND MANAGING PORTFOLIO RISK

OUTLIERS AND THE ASSUMPTION OF IDIOSYNCRATIC NORMALITY

ACCOUNTING FOR OUTLIER RISK

MANAGING OUTLIER RISK

MANAGING ACTIVE EXTENSION PORTFOLIOS

NOTES

PART Three - Special Topics Relating to Active 130/30 Extensions

CHAPTER 4 - Active Extension Portfolios: An Exploration of the 120/20 Concept

RATIONALE FOR ACTIVE EXTENSIONS

THE CAPITALIZATION STRUCTURE OF THE EQUITY MARKET

AN ALPHA OPPORTUNITY MODEL

AN ACTIVE MANAGEMENT MODEL

THE ALPHA RANKING MODEL

AN ACTIVE EXTENSION MODEL

OTHER APPLICATIONS

CAVEATS AND GENERALIZATIONS

CHAPTER 5 - Alpha Ranking Models and Active Extension Strategies

LONG-ONLY ALPHA RANKING MODELS

ADDING THE ACTIVE EXTENSION

GREATER SHORTING COSTS

A DOUBLY CONCENTRATED PORTFOLIO

PRECONDITIONS FOR ACTIVE EXTENSIONS

CHAPTER 6 - The Tracking Error Gap

IMPLICIT BETAS AND BETA DOMINATION

THE KEY CHARACTERISTIC OF POSITION SIZE

UNCORRELATED TRACKING ERRORS

CORRELATION EFFECTS

A SIMPLE FACTOR MODEL

SOURCES OF HIGH TRACKING ERROR

HIGH TRACKING ERRORS AS A PROBE

THE LONG/SHORT FUND CONTEXT

APPENDIX

REFERENCES

CHAPTER 7 - Correlation Effects in Active 120/20 Extension Strategies

LONG-ONLY ALPHA RANKING MODELS

LONG-ONLY ALPHA/TEV RATIOS

ADDING THE ACTIVE EXTENSION

OFFSETTING LONG/SHORT CORRELATIONS

INFORMATION RATIOS FOR ACTIVE EXTENSIONS

ALPHA-FOCUSED INVESTMENT

CONCLUSION

APPENDIX

CHAPTER 8 - Alpha Returns and Active Extensions

ALPHA RANKING AND PORTFOLIO WEIGHTING MODELS

CORRELATION EFFECTS

ALPHA/TEV RATIOS

THE ACTIVE EXTENSION

OFFSETTING LONG/SHORT CORRELATIONS

FUND LEVEL EFFECTS

CONCLUSION

CHAPTER 9 - An Integrated Analysis of Active Extension Strategies

EXPONENTIAL WEIGHTING FUNCTIONS

CORRELATION EFFECTS AND EXPONENTIAL WEIGHTINGS

ALPHA RANKING MODELS

ALPHA/TEV RATIOS

OPTIMAL WEIGHTING EFFECTS

THE ACTIVE EXTENSION

OFFSETTING LONG/SHORT CORRELATIONS

ALPHA-FOCUSED INVESTMENT

CONCLUSION

CHAPTER 10 - Portfolio Concentration

WEIGHTING MODELS AND MAXIMUM ACTIVE WEIGHTS

EXPONENTIAL WEIGHTING MODELS

TRACKING ERRORS

ALPHA RANKING FUNCTIONS

THE INFORMATION RATIO FOR CONSTANT ALPHAS

THE INFORMATION RATIO FOR DECLINING ALPHAS

EFFICIENT FRONTIERS FOR DECLINING ALPHAS

SIMPLE FIXED WEIGHTS

APPENDIX

CHAPTER 11 - Generic Shorts in Active 130/30 Extensions

THE SHORT GENERIC MODEL

EXTENSIONS WITH GENERIC OFFSETS

INFORMATION RATIOS

GENERIC COMPLETIONS

APPENDIX

CHAPTER 12 - Beta-Based Asset Allocation

TOTAL PORTFOLIO BETAS

BETA AND VOLATILITY CLUSTERING

PASSIVE ALPHAS

ALPHA CONSTRAINTS AND DRAGON RISKS

THE ALPHA CORE

LIMITS ON THE ALPHA CORE

THE FIXED CORE FRONTIER

THE VOLATILITY BARRIER AND THE POLICY BOX

CONVERGENCE OF SHORTFALL RISKS

INTEGRATING ACTIVE AND PASSIVE ALPHAS

EQUITY DURATION AND INTEREST RATE EFFECTS

RELATIVE RETURN ANALYSIS

GREATER FLUIDITY IN POLICY PORTFOLIO

REFERENCES

CHAPTER 13 - Beta Targeting: Tapping into the Appeal of Active 130/30 Extensions

BENEFITS OF BETA TARGETING

RELATIVE VOLATILITY AND INTRINSIC TRACKING ERROR

AN EMPIRICAL EXAMPLE

A MONTE CARLO SIMULATION

THE BETA VOLATILITY FORMULA

ALPHA MEASUREMENT AND INFORMATION RATIOS

SETTING COMFORTABLE BETA TARGETS

BETA TARGETING AND TOTAL FUND RISK

THE ALPHA/BETA MATRIX

CONCLUSION

APPENDIX

REFERENCES

CHAPTER 14 - Activity Ratios: Alpha Drivers in Long/Short Funds

STRUCTURE OF LONG/SHORT FUNDS

GROSS, MARKET, AND ACTIVE EXPOSURES

ACTIVITY LEVELS AND ACTIVITY RATIOS

EVOLUTION OF THE IR

MOVING TO THE ACTIVITY RATIO

AR-BASED EFFICIENT FRONTIERS

CORRELATION, POSITION COUNT, AND ALPHA RANKING EFFECTS

IR CURVES FOR DIFFERENT FUND TYPES

MOVING A LONG/SHORT INTO A GENERALIZED ACTIVE EXTENSION

CONCLUSION

APPENDIX

CHAPTER 15 - Generalizations of the Active 130/30 Extension Concept

CHARACTERISTICS OF ACTIVE-EXTENSION-PLUS STRATEGIES

GENERALIZED ACTIVE-EXTENSION-PLUS EXAMPLES

ACTIVE-EXTENSION-PLUS ALPHAS

ACTIVE-EXTENSION-PLUS TRACKING ERRORS AND INFORMATION RATIOS

ALPHA AND BETA RETURNS

THE FUND LEVEL CONTEXT

PART Four - Key Journal Articles

CHAPTER 16 - On the Optimality of Long/Short Strategies

PORTFOLIO CONSTRUCTION AND PROBLEM FORMULATION

OPTIMAL LONG/SHORT PORTFOLIOS

OPTIMAL EQUITIZED LONG/SHORT PORTFOLIO

CONCLUSION

NOTES

REFERENCES

CHAPTER 17 - The Efficiency Gains of Long/Short Investing

FRAMEWORK AND NOTATION

PREVIOUS RESEARCH AND CONTROVERSY

IMPORTANCE OF THE BENCHMARK DISTRIBUTION

THE APPEAL OF LONG/SHORT INVESTING

EMPIRICAL OBSERVATIONS

SUMMARY

ACKNOWLEDGMENTS

APPENDIX A. FURTHER EXPLANATIONS

NOTES

REFERENCES

CHAPTER 18 - Toward More Information-Efficient Portfolios

IMPACT OF CONSTRAINTS

IMPROVING INFORMATION EFFICIENCY BY ALLOWING SHORT SALES

OPTIMAL LEVEL OF SHORTING

TRADE-OFFS AMONG TRACKING ERROR, DEGREE OF SHORTING, AND TRANSFER COEFFICIENT

PRACTICAL IMPLICATIONS

NOTES

REFERENCES

CHAPTER 19 - Allocation Betas

STANDARD MEAN-VARIANCE FRONTIERS

THE ALLOCATION BETA

RETURN DECOMPOSITION: ASSET CLASS

RISK DECOMPOSITION: ASSET CLASS

PORTFOLIO-LEVEL ANALYSIS

ALPHA INDEPENDENCE AND BETA-PLUS

ALPHA CORES AND EFFICIENT FRONTIERS

CONCLUSION

APPENDIX A: ALLOCATION ALPHAS AND BETAS

REFERENCES

CHAPTER 20 - Alpha Hunters and Beta Grazers

TRULY ACTIVE ALPHAS

CHRONIC AND ACUTE INEFFICIENCIES

MARKET IMPACT

REBALANCING AND MARKET EFFICIENCY

RISK AS RISK TO THE POLICY PORTFOLIO

THE ILLUSION OF GROWTH ETERNAL

CONCLUSION

NOTES

REFERENCES

CHAPTER 21 - Gathering Implicit Alphas in a Beta World

IMPLICIT BETAS AND TOTAL FUND BETAS

NEW QUESTIONS

BETA-PRESERVING DIVERSIFICATIONS

IMPLICIT ALPHAS

HUNTING ACTIVE ALPHAS, GATHERING IMPLICIT ALPHAS

THE ALPHA WALL

ALPHA CONSTRAINTS AND BEYOND-MODEL RISKS

THE ALPHA/BETA MATRIX

REVERSING THE STANDARD OPTIMIZATION

RISK AS RISK TO THE POLICY PORTFOLIO

STRESS BETAS

ALPHA EROSION UNDER BETA DOMINATION

GREATER FLUIDITY IN THE POLICY PORTFOLIO

REFERENCES

CHAPTER 22 - Optimal Gearing

SIMPLE MODEL

SIMPLE MODEL EXTENDED: GEARING PENALTIES

CHARACTERISTICS OF UNDERGEARED VERSUS OVERGEARED PORTFOLIOS

EMPIRICAL ANALYSIS

OTHER REAL-WORLD ISSUES

CONCLUSION

APPENDIX A

APPENDIX B

NOTES

REFERENCES

CHAPTER 23 - 20 Myths about Enhanced Active 120/20 Strategies

NOTES

REFERENCES

CHAPTER 24 - Active 130/30 Extensions: Alpha Hunting at the Fund Level

INTRODUCTION

ACTIVE EXTENSIONS

FUND-LEVEL RISK EFFECTS

CONCLUSION

REFERENCES

APPENDIX

CHAPTER 25 - Long/Short Extensions: How Much Is Enough?

THE SHORT EXTENSION MODEL

MODEL PARAMETERS AND IMPLICATIONS

CONCLUSION

APPENDIX A: MODEL DEVELOPMENT

OPTIMAL ACTIVE SECURITY WEIGHTS AND THE FUNDAMENTAL LAW

SIMPLE TWO-PARAMETER COVARIANCE MATRIX

NORMAL PROBABILITY FUNCTIONS IN THE SHORT EXTENSION MODEL

EXPANSION EQUATION FOR SECURITIES OUTSIDE THE BENCHMARK

MARKET-NEUTRAL SPECIAL CASE AND THE SIMPLE APPROXIMATION

CAP-WEIGHTED BENCHMARK MODEL AND EFFECTIVE N

ROBUST APPROXIMATION

ADJUSTMENT FOR SHORTING COSTS

NOTES

REFERENCES

About the Authors

Index

More Praise forModern Portfolio Management

“At a time when boundaries between asset classes are rapidly breaking down and hedge-fund strategies permeate traditional asset management, this book provides a superb framework to understand and analyze the broader implications of this trend in the public equity space.”

—Gumersindo Oliveros, Director, World Bank Pension Plan and Endowments

“Martin Leibowitz and his coauthors have produced a must read for financial market practitioners and researchers who want to go beyond the ABCs of 130/30 extensions. The current volume brings together the breadth of research that Leibowitz and Bova produced on active extensions, complementary work by Emrich, and a sampling of seminal contributions on active management from other authors. It is a necessary resource for pension and endowment fund managers.”

—Edgar Sullivan, Managing Director, General Motors Asset Management

Copyright © 2009 by Morgan Stanley. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey. Published simultaneously in Canada.

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, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, [phone], fax [phone], or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, [phone], fax [phone], or online at http://www.wiley.com/go/permissions.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at [phone], outside the United States at [phone] or fax [phone].

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:Leibowitz, Martin L., 1936- Modern portfolio management : active long/short 130/30 equity strategies / Martin L. Leibowitz, Anthony Bova, Simon Emrich. p. cm. - (Wiley finance series) Includes index.

eISBN : 978-0-470-48494-4

This volume is dedicated to all those whose support nurturedour careers over the years and made this work possible.

Foreword

The High and Low of 130/30 Investing

“There are two ways to make money in the stock market. You can buy low and sell high, or you can sell high and buy low.” With this short statement, California Public Employee Retirement System (CalPERS) entered the world of long/short investing.

I made this statement in the middle of a presentation to the CalPERS board of trustees several years ago. At that time, CalPERS was considering an internal active long/short equity product. The CalPERS investment staff prepared a detailed agenda item for the pension fund trustees that explained how:

• Custodians lent out shares to prime brokers on behalf of their customers,

• Securities lending generated fee revenue to the pension fund,

• Hedge fund managers borrowed these shares from the prime brokers to establish their negative alpha bets,

• The short rebate worked, and

• Collateral must be maintained at the prime broker, as well as many other details.

Midway through the presentation, I realized that the amount of detailed information that the staff had prepared was beginning to build into an unwieldy pile for the fund trustees. It was at the point that I decided to distill into two sentences the essence of what the investment staff wanted to accomplish.

Was this statement an oversimplification? Perhaps. Did it convey the exact nature of what the investment staff wanted to do? Definitely. With this anecdote behind us, the real question remains: Why did CalPERS enter the world of long/short investing? To understand this issue, we need to look at a problem common to many investors, not just pension funds.

BETA GRAZERS DRESSED UP LIKE ALPHA HUNTERS

In his great article on alpha hunters and beta grazers, Marty Leibowitz demonstrates that the asset management industry can be broken down into two simple camps: those that generate active returns—demonstrating a level of portfolio manager skill—and those that generate returns that mostly match the market return.1 Even more bluntly, beta grazers are those asset managers that do little more than capture the systematic risk premium associated with an asset class. Passive/index managers are the classic example of a beta grazer, whereas hedge fund managers are often thought of as the best example of alpha hunters.

However, there are many beta grazers out there that try to disguise themselves as alpha hunters. Consider Exhibit F.1. This is a long-only active equity manager whose stated benchmark is the S&P 500. This manager currently has several billion dollars of assets under management. Consider how neatly this manager tracks the broad stock market. The beta of this active portfolio is 1.000 (yes, I really did carry out the beta calculation to three decimal places) and the R-Square measure is 0.994. More visually, notice that compared to the S&P 500, this active manager produces a nice straight line.

This is one of the first lessons of beta management: Beta grazers are linear in their performance. By this, I mean that when you compare a beta grazer to its benchmark, you should see a straight line of the type presented in Exhibit F.1. The straighter the line, the more the active manager is a beta grazer despite any claims to the contrary.

In addition, this alpha hunter maintains well over 200 positions in its portfolio, many held for risk management purposes. This means that many of the stocks in the portfolio are not held for their alpha generating capability, but rather, are held passively to balance the portfolio back to the benchmark. There is no conviction with respect to the bulk of the securities in this portfolio; many of the stocks held are there to capture the systematic risk premium associated with the S&P 500.

Unfortunately, the performance of this product matches its hidden beta grazer status. It has consistently underperformed the S&P 500 for the last five years by about 55 to 60 basis points (bps) per year—approximately equal to its management fee of 55 bps and trading costs of about 10 bps per year.

THE CURRENT BUSINESS MODEL FOR ASSET OWNERS AND ASSET MANAGERS

Unfortunately, the current business model for most asset owners (pension funds, endowment funds, retail investors, and high net worth investors) as well as for asset managers is: beta trumps alpha.

• Most investors first make the strategic allocation to broad asset classes.

• Then asset managers are directed to squeeze alpha out of the asset owner’s strategic benchmarks. But the strategic benchmarks are designed to be efficiently constructed to measure risk premiums associated with different asset classes.

• As a result, alpha and beta are packaged together in traditional long-only products.

• The result is frequently much more beta than alpha (see Exhibit F.1).

• And, alpha risk budgets are typically spent in the most efficient markets, like large cap equity.

To break out of this conundrum, a new business model must be established: Alpha is sought independently of beta:

• Alpha should not be captive to beta.

• Alpha risk budgets should be spent in the least efficient markets:

• High yield, distressed debt, private equity, small cap, emerging markets, absolute return, real estate, corporate governance.

• These subasset classes have the least efficient benchmarks and, therefore, the highest alpha content.

So what does this really mean? Investors must break away from the traditional asset allocation model of trying to extract alpha from beta drivers.

• Beta grazers are not designed to outperform the market—they provide efficient exposure to broad asset classes, and should capture these risk premiums as cheaply as possible.

Conversely, alpha hunters are designed to outperform the market, often without regard for benchmark boxes—style boxes should be used when an investor believes that it has the least amount of talent or insight to add value.

• Investors must reduce their reliance on beta grazers to generate excess returns.

• In seeking active returns, asset owners should commit their investment capital to those sub-asset classes where asset managers have the best opportunity to add value—look for the cracks in between benchmark boxes.

130/30 products are a natural extension away from benchmark boxes. They seek to exploit the cracks that exist between the more efficient long-only market and the less efficient short market. When an active manager conducts research to construct an active portfolio, she inevitably comes across good and bad stock bets. In the traditional benchmark box of the long-only world, the negative information concerning the bad stock bets cannot be fully exploited. 130/30 products allow asset owners and asset managers to break out of this way of thinking to fully exploit an active manager’s information set.

THE FUNDAMENTAL LAW OF ACTIVE MANAGEMENT

The added value produced by portfolio managers can best be summarized by the Law of Active Management. This law was first proposed by Richard Grinold and later expanded by Clarke, de Silva, and Thorley2:

F.1

where IR is the information ratio of an active manager measured ex post by

F.2

where α is the excess return generated by the portfolio manager, and σα is the active risk taking or tracking error (TE) of the manager.

IC is the information coefficient of the portfolio manager. It is a measure of manager skill and typically gauged by

F.3

TC is the transfer coefficient, where TC ≤ 1.0. This is a measure of how efficiently an active manager can translate her active bets into portfolio positions. Any amount of friction in the financial markets with respect to implementing an active portfolio position (portfolio constraints, trading costs, market impact, opportunity, cost) will reduce the TC below the value of 1.

Breadth is the number of independent bets that the active manager places in the portfolio.

Equation F.1 represents the calculus of active management. Every active portfolio manager is beholden to this rule. For example, a portfolio manager can develop a deep insight into a specific sector or industry, such as biotechnology. For her, the number of independent bets in her portfolio will be limited by her knowledge of this one industry; the breadth will be small. However, her IC should be large as she extracts as much competitive information from a smaller investment opportunity set. Conversely, other portfolio managers will follow several industries to increase the number of active bets (breadth) that they may place into the portfolio. Their trade-off is that their IC is likely to be smaller because they are trying to extract an informational advantage over a larger pool of investment candidates.

EXHIBIT F.2 The Generation of Information Ratios (IRs)

Exhibit F.2 summarizes the Fundamental Law of Active Management, and the several moving parts that can have an impact on portfolio performance. The single largest constraint in active portfolio management is the long-only constraint. It is estimated that this constraint alone can reduce the TC by up to 40 percent.3

EXHIBIT F.3 The Capitalization of the Russell 1000 Stock Index

The limitation of the long-only constraint can best be demonstrated by Exhibit F.3. This exhibit shows a breakdown of the capitalization weighted Russell 1000 stock index, a common equity benchmark. One-third of the capitalization of this index is represented by only 30 stocks, where the average contribution to the capitalization of the index is 1.13 percent. The second tranche is made up of 130 stocks, with an average cap weighting in the index of 0.25 percent. The last tranche of the index consists of 840 stocks, with an average weighting of 0.04 percent. In fact, the median weight for a stock in the Russell 1000 stock index is 0.04 percent.

For active portfolio management, overweights in the portfolio must be funded with underweights. With the long-only constraint in place, the most a portfolio manager can underweight a stock in the portfolio is by its weight in the index. For one-half of the stocks in the Russell 1000, this underweight is only 0.04 percent—not much of an active bet. This forces an active manager to sell down more stocks from the first two terciles to fund the active overweights in the portfolio. Or, even more clearly, consider a portfolio manager whose strongest active overweight is with respect to a stock in the first tercile, whereas her most negative bet is with respect to a stock in the third tercile. This means that her ability to fund her strongest overweight is constrained to only 0.04 percent from the most negative underweight.

EXHIBIT F.4 Breaking Free of the Long-Only Constraint

This problem is by no means limited to the Russell 1000 index. The median weight of a stock in the S&P 500 is only 10 basis points. The smallest 250 companies in the S&P 500 have an index weight of less than 10 basis points.

Exhibit F.4 demonstrates the advantage of relaxing the long-only constraint for 130/30 products. Additional funding is created for active overweights in the portfolio through the use of 30 percent short positions. The 30 percent short positions also increase the leverage of the portfolio. The total exposure to the market is 160 percent—the combination of both 30 percent short active positions with 130 percent long active positions.

Furthermore, the relaxation of the long-only constraint in 130/30 portfolios allows a manager to increase her IR along two dimensions. First, according to Equation F.1, the active manager can increase the number of active bets in the portfolio—expanding the breadth. In addition, the manager can increase the size of her bets—in effect, increasing her IC.

The improvement in the IR of an investment manager follows from the concavity of the return versus risk trade-off common to all actively managed investment products. For both traditional actively managed products and 130/30 products, an increase in TE (σ [α]) leads to an increase in expected excess returns (E[α]). With the long-only constraint in place, the relation between active risk taking and expected alpha is not proportional. This means that increases in risk lead to smaller and smaller increases in alpha. Relaxing the long-only constraint leads to a better trade-off between return and active risk taking. Exhibit F.4 demonstrates this trade-off.

To demonstrate the power and appeal of 130/30 investing, consider two active managers who have the same skill level as measured by the IC: one, a traditional long-only manager, and the other, a 130/30 manager. With the IC held constant, there are two ways for the 130/30 manager to add value beyond that of the long-only manager. First, as previously discussed, the long-only constraint is the single greatest constraint on active portfolio management and can reduce the TC (and the IR) by more than 40 percent. Although there are more costs associated with shorting stocks, these costs are small relative to relaxing the long-only constraint.4

Second, the breadth can be expanded by the 130/30 manager. In fact, the breadth can be expanded in two directions. First, more active long-only bets can be placed into the portfolio because the active manager now has the ability to short stocks to fund long positions that might otherwise not be implemented. Second, negative alpha bets that were previously limited through the long-only constraint may now be executed for the portfolio.

The simple mathematics of Equation F.1 demonstrate that if you can increase the TC or the breadth of the portfolio while holding the IC (manager skill) constant, the IR will increase. It really is not a fair fight between a long-only manager and a 130/30 manager.

CORPORATE GOVERNANCE AND 130/30 INVESTING

Large institutional investors such as CalPERS are keen proponents of good corporate governance of public companies. Yet, the asset management industry, similar to any industry, is subject to good and bad corporate governance and the movement toward alpha/beta separation has improved the governance of this industry. Unfortunately, the existing paradigm for most of the asset management industry is still benchmark driven. Although benchmarks are a useful tool for performance measurement, they are also a significant constraint that reduces the IR of active managers. To achieve consistent alpha, investors must think outside the benchmark in the construction of their portfolios.

Consider a manager that is benchmarked to the S&P 500. This manager is allowed an active risk budget of 5 percent (TE of 5%). This means that the remaining 95 percent of the risk of the portfolio is geared to nothing more than matching the volatility of the benchmark. Why pay active management fees for the 95 percent that does nothing more than track the S&P 500? Again, this gets back to the governance in the asset management industry.

Here is another observation. Ten years ago—even five years ago—an active manager who went 130 percent long and 30 percent short would have called himself a hedge fund manager and typically charged a 2 percent management fee and a 20 percent incentive/profit sharing fee. With the advent of 130/30 products, with most of these products coming from the long-only side of the asset management industry, it is no longer enough for a hedge fund manager to short stocks to demand a 2 and 20 fee structure. Simply, 130/30 products have brought a better form of pricing governance to the asset management industry.

Managers of 130/30 products typically charge a management fee of 0.50 percent to 1.5 percent and a modest profit sharing fee—a far better governance arrangement with the client than a hedge fund would ever establish. Indeed, with the growing number of 130/30, 150/50, and 200/100 products coming to the market, one really has to question whether an equity long/short hedge fund can maintain its 2 and 20 pricing structure.

In summary, 130/30 products have brought transparency into the world of long/short investing, and this is one way to improve the governance in the asset management industry. A clear identification of what is beta and what is alpha goes a long way toward establishing fair and proper pricing. Transparency is a key element of every good governance regime and it can mitigate four risks associated with the asymmetric relationship between asset owners and asset managers:

• Asset managers have much better information as to their true level of skill or alpha-producing ability, as measured by their IC, because the IC is not directly observable by asset owners. The implication is that it is incumbent upon asset managers to make their investment process as transparent as possible to the asset owners, which will lead to more efficient pricing of investment products.

• Furthermore, this asymmetry of information between asset managers and asset owners is exacerbated because the investment process or risk taking by the asset manager is not perfectly observable by the asset owners. It is only after the asset manager has produced a return stream and the beta components have been accounted for that alpha can be observed. Therefore, ex ante, asset managers have much better information about their alpha-producing skills, whereas ex post asset owners need to observe this skill.

• Asset owners only get a snapshot of their portfolio at any point in time. The amount or risk that is embedded in the portfolio, as well as the investment process by which the portfolio was constructed, may not be transparent.

• Asset managers know how much beta they deliver with their alpha. In the traditional governance structure, asset owners receive a combination of alpha and beta from asset managers. In fact, because many asset managers are still driven by benchmark-style investing, there is much more beta than alpha in their investment products. This leads to beta grazers dressed up like alpha hunters.

CONCLUSIONS AND A CAUTIONARY NOTE

130/30 investing has grown in popularity, acceptance, and demand. There are several reasons for the surge in this style of investing.

First, smart investors want more active risk taking. As Exhibit F.1 demonstrates, what is sold as active management can often turn out to be a disguised beta grazer. Active risk taking has declined significantly over the last several years as the dispersion across stocks has decreased. Although most investors would say that they want less rather than more dispersion in their portfolio returns, greater dispersion in stock returns provides larger opportunities for active managers to add value. The long-only constraint locks an active manager into an environment of lower dispersion.

Second, both asset owners and asset managers have come to understand the Fundamental Law of Active Management and its implications for long-only portfolios. The long-only constraint is now widely recognized as the most limiting constraint on the ability for an active manager to generate excess returns.

Finally, less constrained investing has become much more accepted. Hedge funds, absolute return managers, private equity, credit derivatives, commodities, and other forms of alternative assets have expanded the investment opportunity set for asset owners. 130/30 products are the natural extension for both traditional active managers seeking less constrained portfolio management, as well as hedge fund managers moving into more mainstream asset management.

We should note that not all is apple pie with 130/30 strategies; there are additional risks. These strategies appeal both to asset managers and asset owners, but there are many moving parts associated with 130/30 strategies. The most important piece is the ability to borrow stock from a prime broker from which to sell short. Stock can sometimes be hard to borrow, particularly those stocks that are in the lower capitalization range. Also, borrowed stock can be recalled by its owner, forcing the portfolio manager to cover her short position before maximizing the value of her position. In fact, 130/30 managers can sometimes be subject to short squeezes in which the covering of their short positions in the open market results in quick increases in the price of the underlying stock and a reduction of the short sale profits. And, short positions, in theory, can have unlimited risk associated with them because the stock price can—again, in theory—increase forever. This last criticism is a favorite one of consultants to throw out in their resistance to long/short strategies, but with the increase of intelligent risk management and trading systems, this risk is significantly minimized.

These cautionary notes are not meant to diminish the appeal of 130/30 strategies. They are simply meant to indicate that although 130/30 strategies have the potential to add significant value, there are also some additional risks associated with their implementation. However, these risks are far outweighed by the opportunity to improve the IR of the asset manager. Their potential is real and valuable. Furthermore, 130/30 products have brought better pricing governance to at least one part of the alternative investment universe—long/short hedge fund investing. So, understand the risks, but enjoy the benefits.

Mark Anson President and Executive Director of Investment Services Nuveen Investments, Inc.

NOTES

1 See M. Leibowitz, 2005, “Alpha Hunters and Beta Grazers,” Financial Analysts Journal, September/October.

2 See R. Grinold, 1989, “The Fundamental Law of Active Management,” The Journal of Portfolio Management, Spring, pp. 30-37; and R. Clarke, H. de Silva, and S. Thorley, 2002, “Portfolio Constraints and the Fundamental Law of Active Management,” Financial Analysts Journal, September/October, pp. 48-66.

3 See Clarke, de Silva, and Thorley, 2002.

4 See M. Leibowitz, 2005, “Alpha Hunters and Beta Grazers,” Financial Analysts Journal, September/October.

Structure of the Book

This book is divided into four parts.

Part One includes Chapter 1, “Active 130/30 Extensions and Diversified Asset Allocations,” by Martin Leibowitz and Anthony Bova. It describes the key features of active extension (AE) strategies and highlights their ability to improve an equity portfolio’s alpha at the cost of increased tracking error (TE).

Part Two, written by Simon Emrich, consists of two chapters: Chapter 2, “Active Extension—Portfolio Construction,” and Chapter 3, titled “Managing Active Extension Portfolios.” A framework is developed that separates a portfolio into a part that is responsible for benchmark returns and one responsible for the excess, alpha-driven returns. In a long-only portfolio, the active component is asymmetric, consisting of a relatively concentrated long position in the overweights and a relatively diversified short position in the underweights. As the long-only constraint is relaxed, both the risk and return in the portfolio increase. First, the weight in the alpha component can be simply scaled up. This will increase the risk and return of the portfolio proportionately, so the risk-adjusted return will not change. Second, the structure of the alpha component can be changed to take the alpha views better into account. Emrich also presents empirical evidence of the non-normality of stock returns over time. These fat tails have important implications for the risk management of AE portfolios.

Part Three is a compilation of various articles written by Martin Leibowitz and Anthony Bova that were published as Morgan Stanley Portfolio Notes over the 2005 to 2008 period. These articles address various aspects of the AE approach to equity management. Each Note is intended to be self-contained, so that the reader can focus his or her attention on specific areas of immediate interest. As a consequence, there is some degree of overlap across these papers.

Chapter 4, “Active Extension Portfolios: An Exploration of the 120/20 Concept,” was the first article written by Leibowitz and Bova on AE. The increased flexibility for active equity management that AE provides allows a wider range of alpha-seeking opportunities for both traditional and quantitative management. Active extensions open the door to a fresh set of actively chosen underweight positions that are limited in long-only portfolios. With proper risk control, an AE should entail TE that is only moderately greater than that of a comparable long-only fund.

Chapter 5, “Alpha Ranking Models and Active Extension Strategies,” shows how alpha ranking models can be useful for analyzing the structure of AEs (as well as providing useful insights for traditional long-only strategies). With a moderately declining alpha ranking, AE provides increasing alpha/TE ratio (information ratio [IR]) benefits that begin to peak with short percentages somewhere in the 40 to 60 percent region. For a more concentrated ranking model, the optimal shorting percentage is in the 10 to 20 percent range.

Chapter 6, “The Tracking Error Gap,” explores the difference between theoretical projections of the TE and actual TEs seen in practice. This TE gap is usually due to some form of correlation or factor effect across the active positions. If the TEs from active equity management are truly uncorrelated, they are likely to be beta dominated and, therefore, play a very minor role in the standard volatility of the overall fund. However, there is a danger that such correlation/factor effects could accumulate across managers and represent a more significant source of fund-level risk.

Chapter 7, “Correlation Effects in Active Extension 120/20 Strategies,” explores how factor correlations can affect the potential rewards from AEs. These correlations, even at a minimal level, can have a significant effect on the TE and can, therefore, have a meaningful impact on portfolio performance. In AE portfolios, these correlations may lie within the long positions, within the short positions, and/or between the long and short positions. One of the benefits of AE is the opportunity to use the short positions to offset factor effects within the long portfolio. Such offsets can sharpen the intended exposures by removing extraneous risk factors, thereby leading to materially improved IRs.

Chapter 8, “Alpha Returns and Active Extensions,” presents empirical evidence that a wide range of active portfolios can be approximated by exponentially declining alpha rankings and position weightings. The actual sequential weights seen in practice provide confirmation that portfolios are at least roughly structured along these lines. These alpha/weighting models can be used to explore how AEs (and active portfolios in general) can generate alpha returns subject to prescribed risk limits.

Chapter 9, “An Integrated Analysis of Active Extension Strategies,” looks at the impact of various weighting patterns for long and/or short active positions. With the assumption of a constant residual volatility for each active position, the theoretically optimal weighting for each position should be proportional to its alpha ranking. However, one key finding is that for a moderately declining alpha ranking, the alpha/TE ratio is little changed by different, but still reasonable, weighting patterns.

Chapter 10, “Portfolio Concentration,” further explores how various active weighting patterns relate to different alpha rankings. It turns out that higher alphas and still near-optimal IRs can be derived from weights that are significantly more concentrated than the theoretically optimal. Because most funds have significant unused capacity for active risk, more concentrated active structures can enhance the alpha prospects while sustaining near-optimal IRs. Optimal solutions are usually defined in terms of a maximum IR of alpha to TE, but there may be situations in which a sponsor may seek a greater alpha at the expense of a higher TE and a lower IR.

Chapter 11, “Generic Shorts in Active 130/30 Extensions,” discusses the use of customized generic shorts in AE portfolios. Active portfolios often embed factor exposures that are less than fully productive in alpha terms. An appropriate basket of generics can limit unwanted factor effects, lower TEs, and improve IRs. These generics can be thought of as style/sector-specific instruments, such as exchange-traded funds (ETFs) or tailored baskets that are tied to an existing factor in the long-only portfolio. Even though these generic shorts may have zero alphas, they can still provide benefits in terms of providing reinvestable funds and correlation offsets.

Chapter 12, “Beta-Based Asset Allocation,” demonstrates that U.S. equity is the primary risk factor in most U.S. institutional portfolios. The explicit equity percentage is exposure as an inadequate risk gauge of beta risk. The correlations of each asset class with U.S. equities can provide an implicit beta measure that can be used to determine a fund’s total beta. This total beta approach suggests that most U.S. institutional funds share three surprising characteristics:

1. Total volatilities in the 10 to 11.50 percent range,

2. 90 percent or greater correlation between fund performance and U.S. equities, and

3. Total implicit beta values between 0.55 and 0.65.

Chapter 13, “Beta Targeting: Tapping into the Appeal of 130/30 Active Extensions,” shows how having a well-defined beta, even if different from the beta-1 standard, can provide 130/30 extension-like characteristics. The essential feature is having a targeted beta that can act as an expected value, together with a sufficiently low beta volatility. This expectational form of beta targeting allows a broader range of active equity strategies to fall in the AE category. Beta-targeted strategies also help to more clearly identify the true level of alpha performance.

Chapter 14, “Activity Ratios: Alpha Drivers in Long/Short Funds,” focuses on a fund’s activity level—the aggregate weight of all meaningfully sized active long and short positions as the determinant of the basic alpha characteristics. It turns out that the IR depends largely on the activity ratio (AR)—the short activity divided by the long activity. With a given AR, the expected alpha and TE both increase (or contract) proportionally with the long activity level acting as a scaling factor. Thus, funds with the same AR can be viewed as simply rescaled versions of one another with respect to their intrinsic alpha-producing potential. By moving from active to generic positions, or vice versa, a fund can adjust its activity levels to achieve a given AR and activity scale. With beta and AR flexibility, some long/short funds can be reshaped to serve as more generalized versions of a 130/30 or 150/50 AE.

Chapter 15, “Generalizations of the Active 130/30 Extension Concept,” discusses a number of generalizations of the basic AE format that offer the promise of higher alphas while still retaining the AE’s essential structural features and risk characteristics. These generalized AE-Plus strategies may not necessarily have a 100 percent net exposure or beta target of 1. The key to this broader extension umbrella is to establish a well-defined beta-target, which may be smaller than beta-1, and a stable net investment basis (even though that need not be 100 percent). It is the clear-cut distinction between beta and alpha risk that represents the hallmark of such generalized AE-Plus funds.

Part Four reproduces, in chronological order, papers that were published from 1998 to 2008 in external journals, such as the Financial Analysts Journal, Journal of Portfolio Management, and Journal of Investment Management.

Numerous authors have contributed to the theory and methodology of long/short strategies. Part Four contains two papers from three groups of the most influential authors on this topic. Their unique perspective should aid the reader in better understanding the complexities around this important topic. These authors discuss both long/short investing in general as well as hone in on the particular issues that arise from active 130/30 extension portfolios. Each group of authors has chosen both a classic paper that discusses key principles of long/short investing, and a more contemporary paper that reflects their more recent thoughts. The three papers from 1998 to 2004 reflect the earlier work of these influential authors on long/short active equity and the origins of the 130/30 concept.

Bruce Jacobs and Ken Levy derived formulas for optimal active long/short portfolios. The relative sizes of the active and benchmark exposures depend on the investor’s desired residual risk within the framework of a minimum variance active portfolio. They also show that all long and short positions must be simultaneously integrated in the construction of optimal long/short portfolios. The correlations between long and short positions play a large role in determining the risk of a portfolio.

Richard Grinold and Ronald Khan focused on their Fundamental Law of Active Management and its role in determining the IR of a portfolio. They show empirically how the movement into long/short investing offers significant benefits over long-only investing. Long/short implementations are particularly advantageous when the universe of assets is large, asset volatility is low, and the strategy has high active risk. The long-only constraint induces capitalization biases, limits the manager’s ability to act on good insights, and reduces the efficiency of active strategies relative to enhanced index (low-risk) long-only strategies.

Roger Clarke, Harindra de Silva, Steven Sapra, and Steven Thorley extended Grinold and Kahn’s Fundamental Law of Active Management with the development of the portfolio transfer coefficient. They show that the ability to take even modest short positions provides an important structural advantage that can improve the information efficiency of traditional long-only portfolios. Investors do not need to relax the long-only constraint completely to reap substantial benefits as relaxing the constraint by just 10 to 20 percent can be advantageous. This modest relaxation of the long-only constraint results in a disproportionate improvement in the information transfer from security valuation to active portfolio weights.

It was the convergence of these insights that leads to the realization that the 20 percent level of shorting in a 120/20 fund can become broadly acceptable while still capturing a large percentage of the benefits available from more flexible long/short portfolios.

The later 2007 to 2008 papers present the more recent thinking of these authors and their associates who have played such a key role in the development of this area. In addition, we have included several journal articles by Leibowitz and Bova that relate to AE and related issues.

Acknowledgments

The authors would like to acknowledge Morgan Stanley, generally, and Research Management, specifically, for their encouragement and support of these studies on 130/30 active extension. The development of this concept was greatly facilitated by early writers who explored how loosening the long-only constraint could lead to a much wider range of risk-controlled equity portfolios. This extraordinary analytic background was framed by such thought leaders as Rob Arnott, Peter L. Bernstein, Roger Clarke, Harinda de Silva, Richard Grinold, Thomas Hewett, Bruce Jacobs, Ronald Kahn, Kenneth Levy, Robert Litterman, Harry Markowitz, Richard Michaud, Kenneth Winston, and many others.

We are especially grateful to the authors and their co-authors who have made their landmark journal articles available for publication in this volume. These papers have continuing relevance within the investment field, and their inclusion surely represents a substantial contribution to the value of this volume.

This book would not have gotten to the starting gate without the initiative, vision, and “push” of Bill Falloon. Helen of Troy may have launched many ships, but Bill Falloon has surely launched even more books—and hopefully for better causes! In addition, we must express our gratitude to the many wonderful people at Wiley who put forth a very special effort—Emilie Herman, Laura Walsh, Meg Freeborn, and all those others who worked behind the scenes to bring this work to fruition in a timely fashion.

Finally, the authors would like to thank the many clients and colleagues who helped to forge the ideas and to shape the practice of this exciting new initiative in active investment management.

Introduction

Evolution of the Active Extension Concept

The early motivation behind the development of active extensions (AEs) came from plan sponsors who wanted to generate more active returns from their basic equity allocation. In a number of cases, these plans were intrigued by the attractive performance results of various long/short and market neutral strategies. However, most investment offices and their boards were not yet prepared to embrace high levels of shorting and forgo the benchmark-centric discipline of traditional long-only portfolios.

The benefits of moving to a long/short framework had been well described by several theoretical and empirical studies published by Jacobs and Levy (1993, 1995, 1999, 2006), Kahn and Grinold (2000a, b, c), and others. In 2002, building on these earlier studies, Clarke, de Silva, and Thorley (2002) introduced the concept of the “transfer coefficient” that could measure how efficiently active insights are projected into a given portfolio structure. Their study showed that a significant enhancement of the transfer coefficient could be achieved by having the flexibility to short a modest 20 percent of the original asset value. With productive active insights and appropriate risk discipline, the higher transfer coefficient should translate into enhanced alpha returns.

Moreover, this benefit could be achieved while maintaining the key risk characteristics of long-only funds, that is, 100 percent net long, a beta-1 target, and a relatively modest tracking error (TE). In addition to this theoretical argument, funds that had been previously long-only found this 20 percent figure to represent a more palatable “baby step” into the new realm of shorting.

In a 120/20 fund, the basic 100 percent long-only format is extended by allowing shorts amounting to 20 percent of the original asset value, with the 20 percent proceeds being reinvested back into new or existing long positions. The shorting/reinvestment process is carried out to maintain the same beta-1 value as the original long-only portfolio. The resulting portfolio is thus 120 percent gross long and 20 percent gross short. Hence, the designation: 120/20 fund.

Virtually every equity market is highly concentrated, with a small number of stocks with large capitalizations and a much larger number of lower-capitalization companies. In a long-only portfolio, the ability to take significant underweight positions is usually limited to the small number of stocks with large capitalizations. By allowing a limited facility to short stocks within a risk-controlled framework, 120/20 strategies allow more significant views to be expressed in the lower-cap stocks.

The return enhancement benefits for these strategies are derived from a number of interacting sources:

1. More appropriate sizing of previously existing active underweights across the broader range of securities having low percentage weights within the reference index;

2. The additional opportunities in the low capitalization (and less intensively researched) companies that become “freshly” available because they are now candidates for significant active underweighting;

3. New or enhanced active long positions funded by reinvestment of the short proceeds;

4. Portfolio benefits from a wider “breadth” of potential active positions on both an overweight and underweight basis;

5. Use of shorts to offset unproductive sector and style effects within the long portfolio (and vice versa);

6. More intensive active positions made possible by removing extraneous factors and sharpening the focus on key decision parameters.

THE EVOLUTION OF ACTIVE EXTENSION: THE CALPERS BREAKTHROUGH

One story told by a manager at an industry conference was about how his firm got involved in AE. The firm had been using the same methodology to run both a large base of long-only assets and a much smaller market neutral portfolio. Over a recent period, the market-neutral fund had generated better returns. An existing client who was only involved with their long-only product asked how they could achieve the superior performance of the market-neutral fund. When the manager replied that all the client had to do was simply to switch his fund to the market-neutral product, the client responded that his board would never accept that high level of shorting. The manager then asked whether they might consider a little shorting—10 percent, 20 percent? After some analysis, it became evident that a significant (and actually disproportionate) benefit could be obtained by allowing only 20 percent shorts within a properly deigned management framework.

As concerns receded about the ability of previously long-only managers to efficiently manage a subportfolio of shorts, 120/20s morphed into 130/30s and 140/40s, and even the occasional 175/75s. At this point, the majority of portfolios fall into the 130/30 to 150/50 range. As more varied levels of shorting became common, these funds soon became known by more generic names such as “active extensions.” The term active extension (AE) was coined to convey that, rather than being a quantum leap into alternative assets, these strategies were designed as an incremental “extension” of the risk structure of standard active long-only funds.

A major breakthrough occurred in February 2006, when the $212B California Public Employees’ Retirement System (CalPERS), under the leadership of Mark Anson, approved the issuance of a Request-for-Proposal for US AE managers. To date, $3B has been deployed by CalPERS into AE strategies. This “Sacramento Blessing” provided comfort to other US pension funds that relaxing the long-only constraint within this risk-controlled framework was an acceptable way to pursue higher levels of alpha.

BOTH QUANT AND FUNDAMENTAL STYLES

Quantitative managers were the first to become significantly involved in AE, and they continue to represent the majority of assets under management. For the “quants,” moving from long-only to AE was relatively easy because their models already provided ranking scores for a large universe of stocks. Active positions in the more highly ranked securities could always be appropriately sized by overweights. However, within long-only portfolios, the lower rankings could be expressed through minimal or zero holdings. For smaller capitalization stocks, such nonholdings represented a frustrating limit on the more severe underweightings called for by the quant models.

With the flexibility to short, these portfolios could now put in place underweight positions that were more appropriate for these low-scored securities. The cash proceeds from the short sales could then be reinvested into their more highly scored stocks.

For fundamental managers, rankings are typically expressed more implicitly in terms of conviction “tiers.” With the underweights in a fundamental portfolio, the question arises of whether the manager is actively avoiding certain stocks or using the (generally fragmented) underweights simply as a source of funds.

Fundamental managers typically have more concentrated portfolios without the “breadth” found in quant portfolios. However, fundamental managers can certainly take advantage of the other benefits available in an AE framework—“fresh” underweights, enhanced and sharpened long positions, offset correlations,and so forth. Indeed, in some ways, the offset potential from the shorts may be even more valuable to fundamental managers who want to more precisely shape their concentrated exposures.

To this date, a relatively small number of fundamental managers have implemented AE strategies and have generally experienced good results. The more significant trend involves the larger number of both quantitative and fundamental managers that have “seeded” products internally and are intensively considering launching various AE products.

Many of the early AEs represented sponsor-initiated conversions of existing long-only mandates. Sponsors who already had a current relationship with active equity managers generally felt comfortable with their risk-control procedures and their ability to produce positive alphas over the long term. In essence, these sponsors were eager to enhance the alpha returns from the existing allocations to these managers. From the manager viewpoint, AEs call for a wider range of active positions and more intensive monitoring for the short positions. The reward was a deeper relationship with the sponsor and higher fees. The move to AE also opened the door to the potential for performance fees. As AE becomes more widely accepted, the earlier sponsor “push” gave way to managers playing more of a role in proposing conversions or using their experience to attract new relationships.

The already implemented AEs represent a significant trend, but it is even more impressive to see the large pipeline of both quantitative and fundamental products expected to come to the market over the next few years.

THE FUND-LEVEL CONTEXT

It is worth trying to understand the basis for the evident attractiveness of the AE concept. At the outset, there is the increasingly pressing need to extract higher alphas from the basic equity allocation—an allocation that remains sizable in even the most diversified portfolios. There is also the related desire to tap into the typical fund’s unused capacity for taking productive active risk. The AE framework is specifically designed to have TE that is largely uncorrelated with the underlying equity beta risk. With most institutional portfolios having beta as the overwhelmingly dominant risk factor, such uncorrelated TE risks are suppressed and translate only weakly into fund-level risk.

By applying a standard return/covariance matrix to a range of policy portfolios seen in practice, this beta effect can be seen to be the source of three common-risk characteristics shared by most funds: (1) total volatilities ranging from 10 to 11 percent, (2) a 90 percent or greater correlation with US equity movements, and (3) total “correlation-based” betas between 0.5 and 0.65. These high total beta values account for an overwhelming percentage of fund volatility. With funds all having similar total beta values, and with these betas accounting for most of the fund-level volatility, it is not surprising that virtually all funds have similar levels of volatility risk.

Given beta’s central role, it is obviously important for an investment strategy to have the greatest possible clarity on its beta benchmark and for any incremental active risk to be reliably uncorrelated with the dominant beta risk. Active 130/30 extension strategies fit neatly into both these criteria. The beta is well defined, usually at the same beta-1 value as the standard long-only equity portfolio. This beta-1 specification allows AEs to be viewed as residing within the traditional equity space. Moreover, in the AE design, the alpha is clearly delineated so that the associated TE should be basically uncorrelated with the beta risk.

With this minimal beta correlation, any increased TE from AE should have only a small impact on fund-level volatility. Moreover, most institutional funds have far less active management than could be readily accommodated within their overall volatility limits. The challenge is to find active strategies that are:

• Productive (i.e., have the expectation of positive alpha over the long term), and

• Risk “contained” (i.e., where there is a definitive beta target and where the TE, even if sizable, has a reliably low correlation with the underlying beta risk).

It may be helpful to see where AE lies along a spectrum of strategies that stretches from indexing up to the most aggressive forms of macro management.

Beta “grazing” through passive indexing is the most alpha-free form of equity investment. A more active form is alpha hunting to capture skill-based incremental returns relative to a tightly specified benchmark. The TE may vary, but both the alpha return and TE are clearly intended to be defined within a benchmark-centric framework. A rather different approach entails moving beyond the traditional asset class boundaries to “gather” the higher returns available in new asset classes. This alpha gathering may or may not be “benchmark-centric,” depending on the level of active management and the extent to which the fund’s policy portfolio incorporates the new assets. Finally, there is the “foraging” for excess returns however and whenever they can be found. With this ultimately boundary-free flexibility, relative return and alphas obviously become harder to identify or measure.

Within this spectrum, AEs fall solidly within the benchmark-centric “alpha hunting” category.

AE-PLUS GENERALIZATIONS

It can be argued that it is the constrained nature of alpha hunting that accounts for much of AE’s appeal. With its beta-1 equity risk, its 100 percent net long base, and its clearly delineated alphas, AE represents only an “incremental” expansion of the standard forms of long-only active equity. Indeed, it is these more familiar and more comfortable features that enable AE strategies to be kept within the basic equity allocation rather than having to be thrust into the generally smaller allocation dedicated to “alternatives.”

As noted earlier, TE that is largely uncorrelated with the fund’s dominant beta risk will have very little impact on fund-level volatility risk. (This fund level effect is one reason why the information ratio should not constitute the sole yardstick for judging the benefits from a given AE).

The basic motivation behind the AE initiative was the desire for more alpha return without taking on directional leverage or moving too far afield from standard equity management. To pursue higher alphas usually entails accepting higher TEs. However, as long as the mandates are reasonably diversified across management styles, the suppressive effect of the dominant beta will continue to hold for TEs considerably greater than the 3 to 4 percent associated with the typical 130/30 AE. This raises the question of whether more intrinsically active benchmark-centric strategies can still be accommodated within the general AE guidelines. It turns out that there are a number of generalizations of the basic AE format that offer the promise of higher alphas while retaining AE’s essential structural features and risk characteristics so acceptable.

These generalized AE strategies may not necessarily have a 100 percent net exposure or beta target of 1. The key to this broader “extension” umbrella is to establish a well-defined beta-target, which may be smaller or larger than beta-1, and a stable net investment basis that need not be 100 percent. The targeted beta level could also be set at different levels, and it may not have to coincide with the net long position. One example of such generalized “AE-Plus” funds may be a 175/75 structure that remains 100 percent net long but has a somewhat lower beta of 0.7. This strategy may have the appeal for funds seeking higher alphas while reducing their overall beta exposure. Another AE-plus format may have the same 0.7 beta but with a 130 percent gross long and 60 percent gross short for a 70 percent net long position.

Rather than being a rigidly fixed value, the beta target can be more of a design objective that may vary from period to period. As long as the average realized beta matches the target value, it can be shown that the fund volatility will closely approximate the estimated level over time. With a stable beta target, an AE-plus strategy can be viewed as benchmark-centric with well-defined risk characteristics and clearly delineated alphas. Clarity of the beta risk level and the clear-cut distinction between beta and alpha risk represent the hallmarks of a generalized AE fund.

Many long/short managers as well as long-only managers have portfolio styles that are not pinned to specific beta values. However, their strategies often rotate around some average beta value. Such funds could be brought within the generalized AE framework by simply formalizing this preexisting average as a “beta target.” This beta target need not be rigidly realized in every period, as long as it can be construed as a reasonable average value. With this flexibility, the managers can retain their basic investment style and have a basis to access the evident appeal of the AE framework.