The EDHEC European ETF Survey 2008€¦ · portfolio performance analysis, and active asset...

The EDHEC European ETF Survey 2008

June 2008

An EDHEC Risk and Asset Management Research Centre Publication

Sponsored by

Foreword ....................................................................................................................3

Executive Summary .................................................................................................5

Résumé ..................................................................................................................... 12

Methodology ......................................................................................................... 21

Part 1: Results

Current Use of ETFs: Survey Results ......................................................................... 251. The Role of ETFs in the Asset Allocation Process ....................................................................... 26

2. ETFs in Practice .........................................................................................................................................31

3. The Pros and Cons of ETFs, Futures, Total Return Swaps, and Index Funds ................... 35

Part 2: Background

New Risk Budgeting Techniques: Applications with ETFs ........................................ 451. The Core-Satellite Approach ............................................................................................................. 46

2. The Dynamic Core-Satellite Portfolio Process ............................................................................. 50

3. New Risk Budgeting Techniques: Conclusion and Outlook .................................................. 65

Conclusion ............................................................................................................... 67

Appendix ................................................................................................................ 73

References .............................................................................................................. 79

About the EDHEC Risk and Asset Management Research Centre .......... 82

About iShares ......................................................................................................... 86

Table of Contents

Printed in France, June 2008. Copyright EDHEC 2008.The opinions expressed in this survey are those of the authors and do not necessarily reflect those of EDHEC Business School or iShares.

3An EDHEC Risk and Asset Management Research Centre Publication

The EDHEC European ETF Survey 2008 - June 2008

The survey we are pleased to present here is part of the EDHEC Risk and Asset Management Research Centre’s Indices and Benchmarking research programme headed by Felix Goltz and Lionel Martellini.

This programme has led to extensive research on indices and benchmarks in both the hedge fund universe and the more traditional investment classes. In 2006, EDHEC published a study of the quality of major stock market indices. Following up on this study, EDHEC is carrying out work that assesses the advantages and disadvantages of various new forms of equity indices.

In view of the growth and development of ETFs in Europe, and in view of their growing popularity as investment media for both index management and the construction of benchmarks, it is only natural that EDHEC should devote significant resources to research into ETFs. In 2006, with the support of iShares, we published the first EDHEC European ETF survey. The present survey, an update and extension of the 2006 survey, sheds light on recent developments and trends in ETF investing.

The survey is divided into two parts. The first part examines the current use of ETFs, as revealed by answers to our questionnaire. Overall, the results suggest that European investors and asset managers are well aware of the advantages of ETFs. Moreover, this awareness has grown in recent years. However, we also find that the full potential of these instruments for asset allocation is not currently being exploited by the majority of investment management professionals.

The second part complements the survey results by discussing advanced techniques involving dynamic allocation strategies carried out with ETFs. In particular, the second part shows how ETFs can be used in a dynamic risk management context and analyses the benefits of this approach.

We would particularly like to thank our partners at iShares for their continuing support. We would also like to express our appreciation to the authors of the survey and to the publishing team led by Laurent Ringelstein.

Foreword

Noël AmencProfessor of FinanceDirector of the EDHEC Risk and Asset Management Research Centre

4 An EDHEC Risk and Asset Management Research Centre Publication


About the authors

Felix Goltz is a senior research engineer and co-head of the indices and benchmark research programme with the EDHEC Risk and Asset Management Research Centre. His research focus is on asset allocation involving alternative investments, such as hedge funds and derivatives, and on equity indexing strategies. Felix holds a Ph.D. in Finance from the University of Nice Sophia-Antipolis, and has studied economics and business administration at the University of Bayreuth, the University of Nice Sophia-Antipolis and EDHEC.

Véronique Le Sourd has a master’s degree in applied mathematics from the Université Pierre et Marie Curie in Paris. From 1992 to 1996, she worked as research assistant in the Finance and Economics department of the French business school HEC and then joined the research department of Misys Asset Management Systems in Sophia Antipolis. She is currently a senior research engineer at the EDHEC Risk and Asset Management Research Centre.

Adina Grigoriu is head of asset allocation at EDHEC Investment Research, where she advises asset managers on constructing their hedge fund of fund portfolios as well as dynamic core-satellite portfolios. She has an actuarial degree and extensive experience in different finance fields, including quantitative modelling. She started her career as a derivatives trader. She then joined a multinational asset management company where she held several positions ranging from product manager to fund manager and head of ALM.

Noël Amenc is professor of finance and director of research and development at EDHEC Business School, where he heads the Risk and Asset Management Research Centre. He has a masters degree in economics and a PhD in finance and has conducted active research in the fields of quantitative equity management, portfolio performance analysis, and active asset allocation, resulting in numerous academic and practitioner articles and books. He is associate editor of the Journal of Alternative Investments and a member of the scientific advisory council of the AMF (French financial regulatory authority).

Lionel Martellini is professor of finance at EDHEC Business School and scientific director of the EDHEC Risk and Asset Management Research Centre. He holds graduate degrees in economics, statistics, and mathematics, as well as a PhD in finance from the University of California at Berkeley. Lionel is a member of the editorial board of the Journal of Portfolio Management and the Journal of Alternative Investments. An expert in quantitative asset management and derivatives valuation, Lionel has published widely in academic and practitioner journals, and has co-authored reference textbooks on alternative investment strategies and fixed-income securities.

ExecutiveSummary

5An EDHEC Risk and Asset Management Research Centre Publ icat ion



The EDHEC European ETF Survey 2008 relies on a questionnaire that elicited responses from 111 European institutions to analyse the current use of ETFs by European investors and asset managers. In addition, we provide an outlook on future use by (i) asking respondents to comment on future developments and (ii) providing a methodology for using ETFs in a state-of-the-art dynamic risk budgeting process. Thus, we hope to provide insight into how ETFs could be used to further benefit investors. The current survey also provides insight into developments over time, as we can compare results with an earlier survey EDHEC published in 2006.

Current Use of ETFsOverall, the analysis of responses we obtained to our questionnaire leads us to the conclusion that, while ETFs are very popular and widely used among European investors and asset managers, the current use of these products stops short of harnessing their full potential. We summarise the main results of this analysis in the following key conclusions.

1. Dominance of broad market ETFsWhen using ETFs in constructing equity core portfolios, 94% of respondents use ETFs on broad market indices, while only 19% use style ETFs that track finer subcategories of the equity market. Consequently, the possibility of using ETFs to construct optimal core portfolios composed of different equity styles or segments is largely neglected. This neglect is surprising, as the advantage of the wide range of ETFs is precisely that it makes it possible to design precise allocations that correspond to the investor’s long term risk/return objectives, as opposed to accepting

an allocation inherent to a broad market index. The dominance of broad market ETFs in respondents’ core portfolios is not limited to equity investments, but is also found with government bond and corporate bond portfolios.

2. ETFs in the satelliteETFs are now widely used in satellite portfolios. 54% of respondents make use of ETFs in their satellite portfolios, which corresponds exactly to the percentage of respondents using ETFs in the core. This result is interesting, as one of the initial precepts of the core-satellite approach was to use very active instruments in the satellite. However, the outperformance of the satellite may of course be generated by exposure to different forms of beta (small- cap stocks, value stocks, credit risk, and so on) rather than to manager alpha. Our survey results suggest that current industry practice acknowledges the role played by such beta management in the satellite portfolios.

3. ETFs for alternative assets on the riseOur survey results likewise suggest a substantial increase in the popularity of ETFs and ETF-like products for investing in alternative assets. The percentage of respondents using ETFs for commodities, real estate, or hedge funds has increased considerably since our earlier survey in 2006. For each alternative asset class mentioned, 30% to 50% of respondents actually use ETFs. In 2006, only 5% to 15% of respondents used ETFs for a given alternative asset class. Thus, it seems that recently launched products such as real estate ETFs, commodity ETFs, and investable hedge fund products are now widely used by European investors and asset managers. Moreover, along with emerging

Executive Summary



market products, ETFs on alternative asset classes rank most highly on the wishlist of European investors and asset managers for future product development. Finally, when asked where they see the greatest increase in their future use of ETFs, 44% state it will be in accessing new types of asset classes.

4. ETFs still focus on equity investingETFs are still most heavily used for equity investing. Indeed, they make up more than 20% of the average respondent’s equity investments. For bond investments, ETFs do not quite make up 10% of assets. This result is confirmed when looking at the percentage of respondents using ETFs for a given asset class. 78% of respondents use ETFs in equity investing, while less than half use ETFs in fixed-income investing, commodities, and real estate. Satisfaction with ETFs is also higher in equity investing. In fact, 92% of respondents are satisfied with their equity ETF investments, while only 66% (85%) are satisfied with their corporate bond ETFs (government bond ETFs).

5. Advanced features of ETFs are underusedPossibilities such as ETF securities lending, trading options on ETFs, and short-selling ETFs are used by only a fraction of respondents. Even if we include the respondents who say they may use the feature in the future, no more than 13% of respondents are current or potential users of any one of these features. Inverse-performance ETFs, by contrast, are or will be used by more than 30% of respondents. The number of respondents who will use these features in the future is high with respect to the number of current users, suggesting that growth can be expected in ETF lending, ETF options trading, and the short selling of ETFs.

6. ETFs and futures are the preferred indexing instrumentsWe ask respondents to compare ETFs to futures, traditional index funds, and total return swaps across a number of criteria. In terms of liquidity, transparency, and cost, ETFs are considered advantageous by respondents, although they are less well regarded than futures with respect to some criteria. ETFs are viewed as the best in terms of available range of indices and asset classes. It is clear then that European investors and asset managers are well aware of the product diversity of exchange-traded funds, which has increased greatly in recent years. Futures are perhaps the most serious rival of ETFs, but ETFs are preferred for their lower minimum subscription, fewer operational constraints, and friendlier tax and regulatory regimes. The implementation concerns with futures (margin calls, applying exact allocations even for small-sized portfolios) give ETFs an advantage.

7. Passive ETFs with full replication are the preferred choice of investorsA large majority of respondents prefer passive ETFs. Active ETFs are preferred by only about 11% of respondents. Likewise, only 16% of respondents say they would like to see the development of more actively managed ETFs. For the construction of passive ETFs, the majority of respondents prefer pure replication of the index by holding all components at the required weight. Synthetic replication and statistical replication are seen as less attractive than full replication. Synthetic replication throughderivatives, however, is significantly more popular (with 20% of respondents) than statistical replication (with 7% of respondents). It should be noted that the low acceptance of statistical replication

Executive Summary



may constitute a potential barrier to the further expansion of ETFs in asset classes with low liquidity, where full replication may not be feasible.

8. “Indexing” is on the rise. ETFs will benefit most, without harming other indexing vehiclesWe ask respondents to identify the instrument they are most likely to use more in the future. Intentions of future use of ETFs, futures, total return swaps, and index funds reveal that future use is trending upward for all four categories. Moreover, ETFs are the instruments that will benefit most from increased use of indexing instruments. 69% of respondents plan to increase their use of ETFs, while only 3% plan to decrease it. For futures, 36% of respondents plan an increase, while 2% plan a decrease. For total return swaps, only 18% plan to increase their use, and 9% of respondents plan to decrease it. Index funds are the only instrument for which an increase in future use is not pronounced: 23% intend an increase and 19% a decrease. Overall, it seems that the anticipated increase in ETF use will not necessarily hinder the further development of other indexing vehicles.

Compared to the earlier survey conducted by EDHEC in 2006, one can see that the perception of the comparative advantage of ETFs has remained similar but that the

use of ETFs has been growing across all asset classes. The table below provides a comparison of the key results of the two surveys.

As these results show, the use of ETFs in the equity universe has increased from 45% to 78%. In addition, for the other asset classes, ETFs are used by 30% to 50% of respondents to our 2008 survey, as opposed to 5% to 15% in the previous survey. Satisfaction with ETFs has remained at high levels or increased slightly for equity and bond ETFs. For ETFs or ETF-like products on alternative asset classes, satisfaction rates have advanced tremendously. Overall, the 2008 survey points to a continuation and even to an acceleration of the trends suggested by the 2006 survey.

New Risk Budgeting Techniques: Applications with ETFsIn addition to providing an analysis of the current use of ETFs in the industry, we provide an overview of novel ways of making ETFs part of portfolio management. We show how various ETFs may be used in the context of dynamic risk budgeting.

Core-satellite portfolios are usually constructed by placing assets that are supposed to outperform the core in the satellite. However, during some periods

Executive Summary

ETF use 2008 vs. 2006: percentage of users and satisfaction

Equity Govt. Bonds Corp. Bonds Commodities Real Estate Hedge Funds

Percentage of respondents using ETFs

2006 Survey 45% 13% 6% 15% 6% 7%

2008 Survey 78% 42% 40% 48% 35% 30%

If you use ETFs or ETF like products, are you satisfied with them?

2006 Survey 92% 80% 58% 65% 50% 27%

2008 Survey 92% 85% 66% 87% 77% 58%



these assets may underperform the core. The dynamic core-satellite approach described in more detail in the full report makes it possible to reduce the impact of a satellite on performance during a period of relative underperformance, while maximising the benefits of the periods of outperformance.

In our illustrations, we implement the dynamic core-satellite approach (Amenc, Malaise, and Martellini 2004). This method allows asymmetric tracking error management. It leads to an increase in the fraction allocated to the satellite when the satellite has outperformed the benchmark. Indeed, the accumulation of past outperformance results in the potential for a more aggressive (and hence higher tracking error) strategy in the future. If the satellite has underperformed with respect to the benchmark, the method leads to a tighter tracking error strategy in an attempt to ensure the guarantee of the relative performance objective.

To provide a feel for the results in the full document, we summarise two examples. In the first, the investor chooses to add an ETF of value stocks to generate outperformance. In the second, we use the dynamic risk budgeting approach to construct an absolute return fund based on ETFs. We now provide a glimpse of the results of these two examples.

1. Optimal packaging of value exposureThe evidence of a value premium in academic finance has led many investors to tilt their portfolios in the direction of high book to market stocks or, more broadly, towards stocks with low valuation ratios. A

straightforward way of accomplishing this value tilt is by adding an ETF based on a value index as a satellite portfolio.

The figure below indicates the cumulative outperformance of the value index over the large-cap index. As the figure shows, a period of underperformance is followed by a period of overperformance of the value index.

Value minus large-cap spread. Cumulative returns

The table below provides risk and return statistics for different investment strategies with ETFs. The first line shows the performance for an investor who holds large-cap stocks over the test period, that is, for the investor who holds a core without a satellite. Line 2 shows the performance of the satellite, represented by an index for value stocks. Line 3 shows the performance for an investor who adds value stocks as a satellite to his large-cap core portfolio in such a way that they make up 25% of the overall portfolio. In this static core-satellite approach, the weight of the satellite is fixed. Line 4 now shows the dynamic core-satellite approach, in which the weight of the satellite is readjusted so as to control tracking error in an asymmetric manner.

As the average returns in the first column of the table show, the value index does

Executive Summary



not provide much outperformance over the entire period. Consequently, the static core-satellite portfolio adds little performance to the core portfolio (6.08% average annualised returns versus 5.95% for the core). However, adding the satellite in a risk controlled manner through non-linear risk management yields an annualised average return of 8.22%. Thus, the dynamic core-satellite approach adds an annual value of roughly 200 basis points over the static core-satellite portfolio.

2. Designing absolute return funds with ETFsAbsolute return funds have become popular in the asset management industry in recent years. These funds claim to provide relatively smooth returns with a limited level of risk. To illustrate how dynamic risk budgeting may be used in designing absolute return funds with ETFs, we combine a core portfolio that invests in medium-term bonds with a satellite portfolio that invests in an equity ETF. The objective of the proposed strategy is to achieve smooth returns because of the low volatility of the core portfolio, as well as to benefit from the returns on the stock market ETF if stocks outperform bonds, all the while ensuring protection from the downside risk of the equity investment.

The figure shows the cumulative returns of the strategy we use, as well as of the core and the satellite portfolios. In addition, to highlight the built-in protection of this investment strategy, the level of the floor is displayed as well.

Absolute return fund: evolution of core, satellite, and DCS portfolios

From this figure, a number of conclusions can be drawn. The dynamics of the core portfolio confirm the conservative character of the core investment. However, we also see that the performance of the bond core was quite flat over the last two years of the period. For the satellite portfolio, returns are higher if we look at the entire period. More importantly, the fluctuations in the value of the satellite portfolio are tremendous, with a sharp rise to the year 2000 and a plunge from then until 2003, followed again by a steady increase until the end of 2007. The dynamic core-satellite approach combines the advantages of each of its ingredients—the smooth performance of the bond core with the upside potential of the equity

Executive Summary

July 1997 - Dec 2007Average Return*

Maximum Drawdown

Volatility * Downside

Risk *

Modified Value-at-

Risk***

Sharpe-Ratio*/**

Info-Ratio*

DJ EURO STOXX 50 (Core) 5.95% 61.60% 19.89% 14.50% 9.42% 0.20 -

DJ EURO STOXX TM Value (Satellite) 6.23% 50.77% 18.59% 15.27% 9.30% 0.23 0.01

Static Core Satellite 6.08% 58.27% 19.26% 14.50% 9.31% 0.21 0.01

Dynamic Core Satellite 8.22% 57.77% 19.53% 14.43% 9.30% 0.32 0.80

* annualised statistics are given** risk-free rate and MAR are fixed at 2%*** non-annualised 5%-quantiles are estimated

Risk and return statistics for investment strategies with ETFs



satellite. As a result, performance is smooth over the entire period, and cumulative returns at the end of the period are actually higher than those of the satellite alone. It is also interesting to note that as the value of the dynamic core-satellite fund increases, the floor is pulled up to increase the level of protection.

As these examples show, the dynamic packaging of beta exposures makes it possible to outperform naïve static allocation to this beta exposure. In the first example, the dynamic core-satellite technique provides access to the outperformance of value stocks, even though that outperformance is not consistent over the entire time period. In the second example, the conservative nature of the core and the dynamic risk management process both attempt to ensure smooth returns over time. Through the risk-controlled exposure to the equity ETF (the satellite), the absolute return fund provides access to the upside potential of the stock market, while conserving the defensive properties of the core.

The wide range of ETFs on potential satellite assets and the tradability of these ETFs make them ideal for these dynamic allocation strategies.

ConclusionThe results of our survey convey a clear message: ETFs are now widely used and practitioners are highly satisfied with their features. However, the use of ETFs is largely limited to passive holdings of broad market indices. The wide range of ETFs for subcategories and styles is not used to its full potential. Likewise, most practitioners do not benefit from the possibilities of

trading options on ETFs, selling ETFs short, or lending them out. ETFs undeniably provide value when it comes to passive exposure to a traditional or alternative asset class. However, we believe that there is considerable value-added in making use of an important feature of ETFs—namely, that they can be bought and sold like stocks. Thus, they are ideally suited for dynamic risk management in portfolio construction. The last part of our study shows that such dynamic risk budgeting has substantial benefits. While the examples provided there are not meant to be complete solutions, we hope to have provided some food for thought on the future use of ETFs.

Executive Summary



L’EDHEC European ETF Survey 2008 réalise une analyse de l’usage actuel des ETFs par les investisseurs et les gérants d’actifs Européens sur la base d’un questionnaire qui a été rempli par 111 institutions européennes. Cette enquête donne également une perspective sur l’utilisation à venir des ETFs, d’une part en demandant aux répondants d’apporter un commentaire sur les futurs développements des ETFs et d’autre part en fournissant une méthode d’utilisation des ETFs dans un processus sophistiqué de budgétisation dynamique du risque. Nous espérons ainsi donner avec cette étude un bon aperçu de l’utilisation qu’il serait possible de faire des ETFs, au-delà des pratiques actuelles, afin que les investisseurs puissent en tirer le meilleur profit. Par une comparaison avec les résultats d’une précédente étude réalisée en 2006, le présent rapport est également une source d’indications sur l’évolution récente de l’utilisation des ETFs.

Utilisation actuelle des ETFsDans l’ensemble, l’analyse des réponses que nous avons obtenues à notre questionnaire nous a amené à la conclusion que, bien que les ETFs soient très populaires et largement utilisés par les investisseurs et les gérants de fonds européens, leur utilisation courante est loin d’atteindre son véritable potentiel. Les principaux résultats de cette analyse peuvent être résumés en quelques points essentiels.

1. Les ETFs basés sur les grands indices de marché sont dominants 94% des répondants qui utilisent des ETFs pour construire des portefeuilles d’actions core utilisent les ETFs basés sur des indices de marchés larges, alors que seulement 19% emploient des ETFs basés sur des

indices de style, qui permettent pourtant de reproduire des strates plus fines du marché des actions. La possibilité d’utiliser les ETFs pour construire des portefeuilles core optimaux, qui seraient constitués de plusieurs styles ou segments du marché des actions, semble donc être largement ignorée. Cela peut paraître surprenant, sachant qu’il existe un large éventail d’ETFs qui, plutôt que d’accepter l’allocation inhérente à un indice de marché large, offre précisément l’avantage de pouvoir créer des allocations précises, correspondant aux objectifs de rentabilité et de risque de long terme d’un investisseur. La dominance des ETFs basés sur les indices de marché larges dans les portefeuilles core des répondants ne se limite pas à l’investissement en actions, mais se retrouve aussi pour les portefeuilles d’obligations d’Etat et pour les portefeuilles d’obligations risquées.

2. L’utilisation des ETFs dans le satelliteLes ETFs sont maintenant largement utilisés dans les portefeuilles satellites, avec une utilisation par 54% des répondants. Ce pourcentage correspond exactement à celui des répondants qui utilisent les ETFs dans la partie core du portefeuille. Il s’agit d’un résultat intéressant puisque l’approche core-satellite plaidait initialement en faveur de l’utilisation d’instruments très actifs dans le portefeuille satellite. La surperformance du satellite peut cependant être aussi obtenue par différentes expositions bêtas (petites capitalisations boursières, titres value, risque de crédit, etc.), plutôt que par l’alpha du gérant. Les résultats de notre enquête suggèrent que les pratiques courantes de l’industrie semblent reconnaître le rôle joué par cette gestion des bêtas dans les portefeuilles satellites.

Résumé



3. Les ETFs basés sur les actifs alternatifs sont en augmentationLes résultats de notre enquête suggèrent une forte croissance de la popularité des ETFs, et des produits assimilés aux ETFs, investis en actifs alternatifs. Le pourcentage des répondants utilisant les ETFs sur les matières premières, l’immobilier et les hedge funds s’est considérablement accru par rapport à notre enquête précédente réalisée en 2006. Pour chacune des classes d’actifs alternatifs mentionnées ci-dessus, entre 30 et 50% des répondants utilisent actuellement des ETFs. En 2006, seulement 5 à 15% des répondants utilisaient des ETFs pour une classe d’actifs alternatifs donnée. Ainsi, il semble que les produits lancés récemment, tels que les ETFs sur l’immobilier, les ETFs sur les matières premières et les produits sur les hedge funds investissables, soient maintenant largement utilisés par les investisseurs et les gérants de fonds européens. De plus, avec les produits sur les marchés émergents, les ETFs sur les classes d’actifs alternatifs se classent presque aussi haut sur la liste de souhaits des investisseurs et des gérants de fonds européens concernant le futur développement de produits. Pour finir, lorsqu’on leur demande où ils voient la plus forte croissance dans leur future utilisation des ETFs, 44% déclarent que ce sera dans l’accès à de nouvelles classes d’actifs par l’intermédiaire des ETFs.

4. Les ETFs sont toujours concentrés dans le domaine des actionsLe domaine dans lequel l’utilisation des ETFs est la plus développée, est toujours celui de l’investissement en actions. Pour la moyenne des répondants, les ETFs constituent plus de 20% de leurs investissements en actions. Pour les investissements en obligations, les ETFs n’atteignent pas tout à fait 10%

des actifs. Ce résultat est confirmé par l’observation du pourcentage des répondants qui utilisent les ETFs pour une classe d’actifs donnée. 78% des répondants utilisent les ETFs dans l’investissement en actions, alors que moins de la moitié utilisent les ETFs dans l’investissement en obligations, dans les matières premières et l’immobilier. La satisfaction avec les ETFs est aussi plus élevée dans l’investissement en actions. 92% des répondants sont satisfaits par leur investissement dans les ETFs en actions, alors que seulement 66% sont satisfaits par leurs ETFs sur les obligations risquées et 85% par leurs ETFs sur les obligations d’Etat.

5. Les caractéristiques évoluées des ETFs sont sous-utiliséesLes possibilités telles que le prêt de titres des ETFs, la négociation d’options sur les ETFs et la vente à découvert d’ETFs sont utilisées par seulement une fraction des répondants. Même si l’on tient compte des répondants qui ont indiqué qu’ils sont potentiellement de futurs utilisateurs, on ne trouve pour chacune des caractéristiques pas plus de 13% de répondants qui en sont des utilisateurs actuels ou potentiels. Il existe cependant une exception qui concerne les ETFs sur la performance inverse. Ceux-ci sont utilisés, ou seront utilisés dans le futur, par plus de 30% des répondants. Pour toutes ces caractéristiques, le nombre de répondants qui indiquent qu’ils les utiliseront dans le futur est important par rapport au nombre d’utilisateurs actuels, ce qui laisse penser qu’on peut s’attendre à une croissance à la fois du prêt de titres des ETFs, de la négociation d’options sur les ETFs, et de la vente à découvert d’ETFs.

Résumé



6. Les ETFs et les futures sont les instruments d’indexation préférésNous avons demandé aux répondants de comparer les ETFs aux futures, aux fonds indiciels traditionnels et aux total return swaps sur la base d’un certain nombre de critères. En termes de liquidité, de transparence et de coût, les ETFs sont considérés comme aussi avantageux que les futures, bien qu’ils soient moins bien classés que les futures par rapport à certains de ces critères. Les ETFs sont les mieux classés en termes de choix d’indices et de classes d’actifs. Les investisseurs et les gérants de fonds européens semblent donc bien informés de la diversité des ETFs qui a considérablement augmenté au cours des dernières années. Les futures apparaissent comme le challenger le plus sérieux des ETFs, mais les ETFs sont perçus comme supérieurs du point de vue de la souscription minimale, des contraintes opérationnelles, des taxes et du régime réglementaire. Ainsi, les difficultés liées à l’implémentation des futures (telles que les appels de marge, l’application des allocations exactes, même pour les portefeuilles de petite taille) donnent un avantage aux ETFs.

7. Les ETFs passifs avec réplication complète constituent le choix préféré des investisseursLa grande majorité des répondants préfèrent les ETFs passifs. Seuls environ 11% des répondants expriment une préférence pour les ETFs actifs, et seuls 16% des répondants disent qu’ils aimeraient voir se développer plus d’ETFs gérés de façon active. En ce qui concerne la construction des ETFs passifs, la majorité des répondants préfère la réplication pure de l’indice, qui consiste à détenir toutes ses composantes avec le même poids que dans l’indice. La réplication synthétique et

la réplication statistique sont considérées comme moins séduisantes que la réplication complète. La réplication synthétique au moyen de produits dérivés est cependant significativement plus populaire (elle est mentionnée par 20% de répondants) que la réplication statistique (mentionnée par 7% de répondants). Il est à noter que la faible acceptation de la réplication statistique pourrait constituer une barrière potentielle à un développement supplémentaire des ETFs sur les classes d’actifs ayant une faible liquidité, et pour lesquelles la réplication complète ne serait pas réalisable.

8. L’indexation est en augmentation. Les ETFs en seront les premiers bénéficiaires, sans que cela cause du tort aux autres véhicules d’indexation. Nous avons demandé aux répondants l’instrument qu’ils étaient le plus susceptibles d’utiliser dans le futur. Les intentions concernant l’utilisation future des ETFs, des futures, des total return swaps et des fonds indiciels révèlent une tendance à l’augmentation pour les quatre catégories. De plus, les ETFs sont les instruments qui tireront le plus de profit d’une augmentation de l’utilisation des instruments indiciels. 69% des répondants ont le projet d’augmenter leur utilisation des ETFs, tandis que seulement 3% ont le projet de la diminuer. Pour les futures, 36% des répondants ont un projet d’augmentation, contre 2% qui ont un projet de diminution. Pour les total return swaps, seulement 18% ont le projet d’augmenter leur utilisation, et 9% des répondants ont le projet de la diminuer. Les fonds indiciels constituent le seul type d’instrument pour lesquels une augmentation de son utilisation n’est pas prévue, avec 23% des répondants qui comptent augmenter leur utilisation et 19%

Résumé



la diminuer. De façon générale, il ne semble pas que le développement prévisible de l’usage des ETFs entravera le développement des autres véhicules d’indexation.

Comparé à la précédente enquête réalisée par l’EDHEC en 2006, on peut voir que la perception de l’avantage comparatif des ETFs est restée similaire, mais que l’utilisation des ETFs s’est accrue en gagnant toutes les classes d’actifs. Le tableau ci-dessous fournit une comparaison entre les principaux résultats des deux enquêtes.

Ces résultats montrent que l’utilisation des ETFs est passée de 45% à 78% dans l’univers des actions. En ce qui concerne les autres classes d’actifs, les ETFs sont utilisés par 30 à 50% des répondants à notre enquête de 2008, alors que leur utilisation était réduite à 5 à 15% d’utilisateurs dans l’enquête précédente. La satisfaction apportée par les ETFs est restée à des niveaux élevés, ou a légèrement augmenté, pour les ETFs actions et obligations. Pour les ETFs, ou les produits assimilés aux ETFs, basés sur les classes d’actifs alternatifs, les taux de satisfaction ont considérablement progressé. De manière générale, les tendances qui avaient été établies dans la dernière enquête ont persisté et se sont même renforcées.

Les nouvelles techniques de budgétisation du risque : Application aux ETFsCette étude fournit non seulement une analyse de l’utilisation actuelle des ETFs dans l’industrie, mais elle donne une vue d’ensemble des nouveaux moyens d’application des ETFs en gestion de portefeuille. De façon plus précise, nous décrivons dans cette étude comment il est possible d’utiliser lesdifférents ETFs dans un contexte de budgétisation dynamique du risque.

Il est d’usage de construire des portefeuilles core-satellite en plaçant les actifs qui sont sensés surperformer le core dans le satellite. Cependant, il peut arriver durant certaines périodes que ces actifs sous-performent le core, par exemple si les conditions économiques leurs deviennent temporairement défavorables. L’approche core-satellite dynamique décrite avec plus de détail dans le rapport complet, rend possible une réduction de l’impact du satellite sur la performance durant une période de relative sous-performance, tout en maximisant ses profits durant les périodes de sur-performance.

Dans nos exemples, nous avons mis en oeuvre l’approche core-satellite dynamique

Résumé

Utilisation des ETFs en 2008 comparée à 2006 : Pourcentage de satisfaction des utilisateurs

ActionsObligations

d’EtatObligations

risquéesMatières premières

Immobilier Hedge Funds

Pourcentage de répondants utilisant les ETFs

2006 Survey 45% 13% 6% 15% 6% 7%

2008 Survey 78% 42% 40% 48% 35% 30%

Si vous utilisez les ETFs ou les produits assimilés aux ETFs, en êtes-vous satisfaits ?

2006 Survey 92% 80% 58% 65% 50% 27%

2008 Survey 92% 85% 66% 87% 77% 58%



(Amenc, Malaise et Martellini 2004). Cette méthode permet aux investisseurs de réaliser une gestion dynamique de la tracking error. Elle conduit à une augmentation de la fraction allouée au satellite quand le satellite a surperformé le benchmark. Ainsi, la surperformance accumulée dans le passé offre le potentiel d’une stratégie plus agressive dans le futur (et ainsi d’une tracking error plus élevée). Si le satellite a sous-performé par rapport à son benchmark, la méthode conduit à une stratégie de tracking error plus restreinte afin de garantir l’objectif de performance relative.

De façon à donner l’idée générale des résultats du document complet, nous résumons ici deux exemples. Dans le premier exemple, l’investisseur choisit d’ajouter un ETF d’actions value pour générer de la surperformance. Dans le second exemple, nous utilisons l’approche de budgétisation dynamique du risque pour construire un fonds de rendement absolu basé sur les ETFs. Dans ce qui suit, nous donnons successivement un aperçu des résultats obtenus avec ces deux exemples.

1. Structure optimale de l’exposition valueLa mise en évidence d’une prime value dans la finance académique a conduit beaucoup d’investisseurs à orienter leur portefeuille vers les titres ayant un ratio book-to-market élevé ou plus généralement vers les titres ayant des ratios d’appréciation faible. Une façon directe de mettre en œuvre ce tilt est d’ajouter au portefeuille un ETF basé sur un indice value et qui constitue le portefeuille satellite

Le graphique ci-dessous indique la performance cumulée de l’indice value par

rapport à l’indice large cap. On peut voir qu’une période de sous-performance de l’indice value, est suivie par une période de surperformance.

Spread entre les indices value et large cap. Rentabilités cumulées

Le tableau ci-dessous fournit les statistiques de risque et de rentabilité pour différentes stratégies d’investissement utilisant des ETFs. La première ligne présente la performance d’un investisseur qui détient des titres de forte capitalisation boursière sur la période de test, c’est-à-dire qu’il détient un portefeuille core sans satellite. La deuxième ligne donne la performance du satellite, représentée par un indice de titres value. La troisième ligne donne la performance d’un investisseur qui ajoute un satellite, constitué de titres value, à son portefeuille core, constitué de titres de forte capitalisation boursière, de sorte que le satellite représente 25% de l’ensemble du portefeuille. Dans cette approche core-satellite statique, le poids du satellite est fixé. La ligne 4 du tableau montre ensuite le résultat de l’approche core-satellite dynamique, dans laquelle le poids du satellite est réajusté de façon à ce que la tracking error soit contrôlée de manière asymétrique.

La première colonne du tableau, qui contient les rentabilités moyennes, montre que

Résumé



l’indice value ne génère pas beaucoup de surperformance sur la période entière. Par conséquent, le portefeuille core-satellite statique ajoute peu de performance au portefeuille core (6,08% de rentabilité moyenne annualisée contre 5,95% pour le portefeuille core). Cependant, l’ajout du satellite dans un processus de contrôle du risque au moyen d’une gestion non linéaire du risque conduit à une rentabilité moyenne annualisée de 8,22%. Ainsi l’approche core-satellite dynamique ajoute une valeur annuelle d’environ 200 points de base par rapport au portefeuille core-satellite statique.

2. Construction de fonds de rendement absolu avec des ETFsLes fonds de rendement absolu ont connu une forte progression dans l’industrie de la gestion d’actifs au cours des dernières années. Ces fonds déclarent fournir des rentabilités relativement régulières avec un niveau de risque limité. De façon à illustrer comment on peut utiliser la budgétisation dynamique du risque pour construire des fonds de rendement absolu avec des ETFs, nous avons combiné un portefeuille core, investi en obligations d’échéance moyenne, avec un portefeuille satellite, investi en ETFs actions. L’objectif de la stratégie proposée est de parvenir à des rentabilités régulières grâce à la faible volatilité du portefeuille core. De plus, l’objectif est de

bénéficier des rentabilités de l’ETF actions si les actions surperforment les obligations, tout en parvenant à une protection du risque de baisse de l’investissement en actions.

Le graphique montre les rentabilités cumulées de la stratégie qui a été mise en œuvre, ainsi que celles des portefeuilles core et satellite. De plus, afin de mettre en valeur la protection intégrée à cette stratégie d’investissement, nous avons également fait figurer le niveau du plancher (floor).

Rendement absolu: Evolution des portefeuilles core, satellite et DCS (dynamique core-satellite)

A partir de ce graphique, on peut tirer un certain nombre de conclusions. La dynamique du portefeuille core confirme le caractère conservatif de cet investissement. Nous voyons aussi que la performance du core, qui est investi en obligations, a été très plate sur les deux dernières années de la période. Pour le portefeuille satellite, nous observons que ses rentabilités sont plus élevées que celles

Résumé

Juil 1997 - Déc 2007Rentabilité Moyenne*

Risque de perte

maximum Volatilité*

Risque de baisse*

Value-at-Risk

Modifiée***

Ratio deSharpe*/**

Ratio d’Informa-

tion*


DJ EURO STOXX TM Value (Satellite) 6.23% 50.77% 18.59% 15.27% 9.30% 0.23 0.01

Core Satellite Statique 6.08% 58.27% 19.26% 14.50% 9.31% 0.21 0.01

Core Satellite Dynamique 8.22% 57.77% 19.53% 14.43% 9.30% 0.32 0.80

* les statistiques ont été annualisées** le taux sans risque et la rentabilité maximum acceptable (MAR) sont fixés à 2%*** on a estimé les quantiles à 5% sans annualisation

Statistiques de rentabilité et de risque



du core, si l’on prend en compte toute la période. De façon plus importante, il apparaît que les fluctuations de la valeur du portefeuille satellite sont énormes, avec une augmentation prononcée de sa valeur jusqu’à l’année 2000 et un déclin radical à partir de cette date et jusqu’en 2003, suivi à nouveau par une augmentation soutenue jusqu’à la fin de 2007. L’approche core-satellite dynamique combine l’avantage respectif de chacun de ces ingrédients, à savoir une performance régulière du portefeuille core obligataire et le potentiel à la hausse du satellite action. Par conséquent, la performance est régulière sur toute la période, et les rentabilités cumulées à la fin de la période sont réellement plus élevées que celles du satellite. Il est aussi intéressant de regarder les valeurs dynamiques du plancher. Lorsque la valeur du fonds core-satellite dynamique augmente, le plancher est tiré vers le haut de façon à augmenter le niveau de protection.

A partir de ces exemples, on peut voir que le packaging dynamique des expositions bêtas permet de générer de la surperformance par rapport à une allocation statique naïve de cette exposition bêta. Dans le premier exemple, la technique core-satellite dynamique permet de bénéficier de la surperformance des titres value, même si cette surperformance n’est pas consistante sur toute la période. Dans le deuxième exemple, la nature conservative du core et le processus de gestion dynamique du risque servent tous les deux à obtenir des rentabilités régulières au cours du temps. Au moyen d’une exposition, contrôlée du risque, à l’ETF actions (le satellite), le fonds de rendement absolu permet d’accéder au potentiel de hausse du marché d’actions,

tout en conservant les propriétés défensives du portefeuille core.

Les ETFS sont un véhicule idéalement adapté à la mise en oeuvre de telles stratégies dynamiques, puisqu’il existe un large choix d’ETFs sur des actifs susceptibles de constituer des satellites, et leur négociation aisée permet d’implémenter des stratégies d’allocation dynamique telles que celle qui a été décrite.

ConclusionLes résultats de notre enquête donnent un message clair : les ETFs sont maintenantlargement utilisés et les praticiens sont hautement satisfaits de leurs caractéristiques. Cependant, l’usage des ETFs est principalement limité à la détention passive d’indices de marché larges. Le large choix d’ETFs disponibles sur les sous-catégories d’actifs et les styles de gestion n’est pas utilisé à la hauteur de son potentiel. De même, la plupart des praticiens ne tirent pas profit de la possibilité de négocier des options sur les ETFs, de vendre des ETFs à découvert ou de prêter les titres des ETFs. Les ETFs fournissent indéniablement une valeur ajoutée en ce qui concerne l’exposition passive à une classe d’actifs traditionnels ou alternatifs. Cependant, nous croyons qu’il existe une valeur ajoutée considérable à faire usage d’une caractéristique importante des ETFs, qui est de pouvoir les acheter ou les vendre comme des actions. Ainsi, ils sont idéalement adaptés pour gérer dynamiquement le risque dans la construction de portefeuille. La dernière partie de notre étude montre qu’une telle budgétisation dynamique du risque a des bénéfices substantiels. Bien que les exemples fournis ne soient pas censés

Résumé



être des solutions complètes, nous espérons qu’ils auront fourni de la matière à réflexion pour le futur usage des ETFs.

Résumé

Methodology




Since its 2001 founding, the EDHEC Risk and Asset Management Research Centre has monitored practices in the European asset management industry. Surveys on the state of the European asset management industry have looked specifically at the use of recent research advances by investment management companies and at best practices in the industry. Among our surveys are inquiries into portfolio risk management, the use of indices and benchmarks, fund of hedge fund management, alternative diversification, exchange-traded funds, and real estate investment.

The present survey focuses on the use of exchange-traded funds by asset management firms, institutional investors, and private wealth managers. Surveying the ETF landscape is a promising venture since new products are being launched frequently, and to us it seemed worth looking into the use asset management practitioners make of these innovations. We also did a survey on this topic in 2006; the aim of the current study is to expand on and to update the earlier one.

Our survey is based on a questionnaire that was addressed to industry participants in Europe from 29 January 2008 to 21 April 2008. The study generated responses from 111 institutions based in Europe. A majority of these institutions—56% of respondents—are investment management or advisory firms. In addition to this group of third-party fund managers, there is a significant proportion of pension funds and similar institutional investors, who make up 17% of the sample. Insurance companies account for another 7%, while respondents from the banking sector make up 10%. Slightly below 2% of our respondents are family offices.

Approximately 8% of the respondents do not identify their sector of activity.

Exhibit 1: Activities of the respondents

55.9%

17.1%

9.9%

1.8%

8.1%

7.2%

Investment management and advisoryInsurancePension fund or Foundation

BankFamily officeNo answer

We then use two proxies—assets under management and the number of staff members devoted to investment analysis and management—to shed light on the size of the respondents to our questionnaire.

As it happens, we elicit responses from institutions with a wide range of assets under management (AUM). Overall, it can be said that we cover all size categories and the distribution over different size categories is relatively smooth. 8.6% of respondents have less than €100 million of assets under management; 24.7% manage between €100 million and €1 billion, 34.6% between €1 billion and €10 billion, 19.8% between €10 billion and €100 billion, and 12.3% more than €100 billion.

Methodology



Exhibit 2: Amount of assets managed or delegated

24.7%

19.8%

12.3% 8.6%

34.6%

< 100 Million100 Million to 1 Billion1 Billion to 10 Billion

10 Billion to 100 Billion> = 100 Billion

The wide range of sizes covered by our sample is also reflected in the number of people assigned to asset management. 44% of participants have a team of fewer than 10 people occupying asset management functions. 39% employ between 10 and 100 people, while 11% employ between 100 and 1,000 in asset management. 6% of respondents are very large institutions with asset management staff of over 1,000 people.

Exhibit 3: Number of people assigned to asset management

44%

39.3%

10.7%

6%

< 1010 to 100

100 to 1000> = 1000

Listings on public exchanges are also indications of company size. Approximately one-third of the institutions that respond to our questionnaire are exchange-listed; the majority are privately held.

Exhibit 4: Is your company listed on the stock exchange?

31.5%

61.3%

7.2%

YesNoNo answer

It is useful to underline that our survey allowed us to obtain responses from major institutions, as shown by the 12.3% of respondents who manage in excess of R100 billion. At the same time, responses from our survey may provide us with a fairly balanced view of asset management practices across size categories.

The breakdown of survey respondents by country suggests that our confidence in the geographic distribution of the sample is well placed. In fact, it is clear that most responses are obtained from institutions based in one of four major European markets—France, Switzerland, the UK, and Italy. Responses from these countries account for anywhere from 12% to 20% of all responses. Germany, the Benelux countries, Northern Europe, and Eastern European countries account for 4% to 9% of respondents. Other European countries (mainly Austria, Spain, Portugal, and Greece) make up 13% of respondents. There may be a small bias towards French companies, which make up 19.6% of respondents. EDHEC, after all, is a French institution and it may obtain higher response rates in its home country. However, in view of the largely representative figures for the

Methodology



UK and Switzerland, the overrepresentation of France is not a major problem.

Exhibit 5: Country distributions of the respondents

19.6%

8.8%

12.7%

3.9%

3.9%

6.9%

12.7%

17.6%

13.7%

FranceSwitzerlandUnited KingdomItalyGermany

BeneluxNorthern EuropeEastern EuropeOthers

Overall, from these breakdowns, we are confident that our survey provides representative insights into the current practices of a range of investment institutions. Not only has the questionnaire been filled out by 111 European institutions, but the breakdown of these institutions by assets under management and by country also shows that we cover a range of institutions. We turn now to the findings of this questionnaire.

Methodology

1. Results




How do investment managers use ETFs? Are they satisfied with the characteristics of ETFs? What are the advantages of ETFs with respect to such competing ways of investing in an index as traditional index funds, futures, or total return swaps? These are the questions that we ask leading European investment managers and institutional investors in our questionnaire. In what follows, we present the resulting replies and comment on the broader trends that they suggest.

1. The role of ETFs in the asset allocation processIn this section, our survey addresses the use of exchange-traded funds in core-satellite methods of asset allocation. The core-satellite strategy is widely regarded as an effective means of organising asset allocation, as will be further detailed in part 2 of this document. Despite its advantages, only 50% of European investment managers have taken a core-satellite approach to portfolio construction. Interestingly, as shown in exhibit 1.1, 15% of respondents are planning to change the organisation of their allocation to the core-satellite approach in the near future. This shows that the core-satellite approach is still gaining popularity. In fact, both the number of users and the number of potential adopters have increased with respect to our 2006 survey on ETFs. Only 5% of respondents in this survey report that they are not familiar with this approach.

Exhibit 1.1: Have you implemented (or are you going to implement) "core-satellite"-type organisation of allocation?

20.7%

1.8%

8.1%

4.5%

15.3%

49.5%

YesNo, but we will soonNo

No, unlikely to in the futureNot familiar with itNo answer

ETFs can be a desirable instrument not only in the core portfolio (for strategic allocation purposes) but also in the satellite (for tactical bets). As exhibit 1.2 shows, respondents actually do use ETFs for both of these purposes. In fact, 54% use ETFs in the core, the same as the percentage that trades them for the satellite. Rather than considering ETFs for either the core or the satellite, 58% of respondents use them to cover a particular geographic zone, asset category, or asset class.

Exhibit 1.2: What role do ETFs play in your allocation?

We use ETFs in the core portfolio for strategic

allocation purposes

We use ETFs to representa particular geographical

zone, asset category or asset classes

We use ETFs in the satellite portfolio

in order to make tactical bets

Other

0

10

20

30

40

50

6054% 54%

58%

4%

ResultsCurrent Use of ETFs: Survey Results



Indexing is sometimes perceived as restricted to the core portfolio, while the satellite should be made up of actively managed vehicles. In part 2 of this document, one of our objectives is precisely to show how ETFs may be used in the satellite of the portfolio. It should be noted that the results on ETF use indicate that ETFs are as popular for the satellite as they are for the core. In the analysis of the next questions, we will further assess the role of ETFs on different asset classes as a component of either the core or the satellite portfolio.

In particular, we ask survey participants to identify the ETFs they use for investment in equity, fixed income, and alternative asset classes. We also ask whether each type of ETF is used in the core portfolio and/or in the satellite portfolio. Note that responses are non-exclusive, as a given type of ETF may be used in both the core and the satellite portfolio and different types of ETFs may be used simultaneously. In addition, the percentages shown refer to the percentages of users of ETFs for the given asset class who use a particular type of ETF. For example, the percentage of users of style ETFs must be interpreted as the percentage of equity ETF users who use style ETFs. This presentation assures that we assess the relative importance of the types of ETFs within the asset class, as opposed to the overall importance of the asset class itself.

The key to diversification benefits is low correlation of portfolio holdings, so the co-movements of returns of portfolios that represent certain styles or categories within a broad asset class must be looked at. For portfolio managers, it is crucial to

determine the approach that offers the greatest diversification benefits. In our survey, we are interested in highlighting the subcategories that are useful to investment practitioners as they structure their portfolio decisions.

In the equity class, size (large cap, small cap, and so on) and style (growth, value) have been shown by Fama and French (1992) to account for a significant portion of the cross-sectional difference in expected returns of equities. Building on these findings are investment strategies based on characteristics such as market capitalisation and investment style. The low correlation of growth stocks and value stocks and of large- and small-capitalisation stocks has led to style diversification. A study by Ibbotson Associates on sector investing (2002) concludes that sectors can also be used to construct an equity portfolio. The study points to the low correlation of different sectors and to the low correlation of sectors and the market. Another study (Hamelink, Harasty, Hillion 2001) shows that the benefits of sector diversification outweigh those of country diversification. Reflecting the concepts of style and sector diversification, ETF providers have moved from providing global market exposure with funds tracking broad indices to style, sector, or country ETFs that track more specific segments of the equity markets.

When the usefulness of ETFs is evaluated, it is clear that European investors prefer broad-based ETFs. Exhibit 1.3 shows that 94% of equity ETF users use these vehicles in the core portfolio, while 51% use them in the satellite. Style and sector ETFs are less widely used. When they are used, it is mostly in the satellite, not in the core.




In the core portfolio, sector ETFs are used by 31% of respondents, while style (growth/value) ETFs are used by 19%.

Style ETFs are clearly less popular than sector or broad-based ETFs, especially for use in the core portfolio. Because the academic literature has insisted on the importance of style factors, this finding is surprising. As investment styles are not highly correlated, and as this correlation is remarkably stable across market states, equity style diversification is in fact one of the most promising ways of building a diversified core portfolio. These results for style ETFs concur with the low number of ETFs available on style indices in Europe. In fact, these results can be explained by the historical importance of style investing in the US and by its relative unpopularity in Europe.

Exhibit 1.3: Concerning your equity investments, please indicate which of the following you use for the core and/or satellite portfolio

Broad market ETFs Style ETFs Sector ETFs

94%

19%

51%

39%31%

71%

0

20

40

60

80

100

CoreSatellite

For government bond investments, our respondents again prefer broad-based ETFs in the core portfolio, with 84% of bond ETF users stating that they use broad market ETFs in the core portfolio (see exhibit 1.4). Maturity-segment ETFs are the most popular form of non-broad-based

ETFs. 72% of bond ETF users use these vehicles in the satellite portfolio and 33% use them in the core. Different maturity segments, after all, are natural media for tactical timing strategies in the satellite. In addition, the time-series behaviour of these instruments, which differs with their differing exposure to interest rate changes, also makes them a useful vehicle for constructing an efficient core portfolio. Inflation-protected bond ETFs are less popular, perhaps owing to their relatively recent introduction or perhaps to the perception that inflation protection comes at a high price.

Exhibit 1.4: Concerning your government bond investments, indicate which of the following you use for the core and/or portfolio

Broad market ETFs Maturity- segment ETFs

Inflation-protected bond ETFs

39%33%

84%

72%

20%

36%

CoreSatellite

0

20

40

60

80

100

The responses for corporate bond investments are broadly in line with those for government bond ETFs. Exhibit 1.5 shows the responses on the use of four kinds of corporate bond ETFs. In the core portfolio, broad market ETFs for corporate bonds were again the most widely used of the corporate bond ETFs. ETFs on indices that subdivide corporate bonds into finer categories, such as countries, sectors, or maturity segments, are used less frequently in the core. However, when it




comes to the satellite portfolio, the most popular ETFs are rating-segment ETFs (used by 49% of the respondents who use bond ETFs), followed by maturity-segment ETFs and broad market ETFs (used by 38% of the respondents who use bond ETFs). These results show that practitioners seem to agree with academic research that points to the significant benefits of active allocation to such finer categories of the bond market as maturity segments (see Amenc, Malaise, and Martellini 2004). However, while maturity-segment ETFs and rating-segment ETFs are relatively widely used in the satellite portfolio, only 26% of the respondents who use bond ETFs use sector-specific corporate bond ETFs in the satellite.

Exhibit 1.5: Concerning your corporate bond investments, please indicate which of the following you use for the core and/or satellite portfolio

Broad market

ETFs

Maturity- segment

ETFs

ETFs by credit

rating segment

ETFs by sector

38% 38%

12%

72%

33%

19%

49%

26%

CoreSatellite

0

10

20

30

40

50

60

70

80

After having looked at traditional investments, we look at how European investors use ETF-like index products for alternative investments. The predominance of commodity ETFs or other ETF-like commodity investment products is clear. Some three-quarters of the users of ETF-like alternative investments use these

products for commodities investing. As one can imagine, investable index products are not as popular in hedge fund investments, possibly as a result of their relatively recent appearance. Only 27% of all users of ETFs for alternative asset classes use them for the core; 28% use them for the satellite. Real estate investment accounts for 54% of overall users of ETFs for alternative asset classes when it comes to the satellite portfolio and for 36% when it comes to the core portfolio.

Exhibit 1.6: Please indicate if you use ETFs or ETF-like products in the core and or satellite for the following alternative asset classes

Real Estate Commodities Hedge funds

54%

75%

36%

76%

28%27%

CoreSatellite

0

10

20

30

40

50

60

70

80

The results obtained from responses to our survey suggest that different types of ETFs are used for each asset class. These results can be explained as the demand-side reflection of the multiplication of ETF offerings on the supply side. While ETFs on finer segments of the respective markets are relatively widely used as satellite vehicles, the dominance of broad market ETFs when it comes to investments in the core portfolio is striking. This dominance of broad market ETFs is not confined to equities alone, as these ETFs also account for the prevailing share, though to a somewhat lesser degree, of the demand for




government bond ETFs and corporate bond ETFs. Perhaps the most important result of our analysis is that, instead of actively managing their long-term beta exposure to obtain the most efficient risk-return trade-off, European institutional investors and asset managers focus on using broad market indices in their core portfolios.

Finally, we look into the relative importance accorded ETFs and other investment instruments in each asset class. For each asset class, exhibit 1.7 shows the percentages of amounts invested that are accounted for by ETFs. As can be seen, ETFs are now a sizable share of overall amounts in the equity universe. Indeed, for the average respondent to this question, they account for 22% of total equity investment and for a broadly comparable (16%) share of commodities investment but for less than 10% of average overall investment in other asset classes. Worth noting is that ETFs account for more than one-fifth of overall equity holdings, but less than one-tenth of bond holdings. Moreover, corporate bond investment is even less likely to be done via ETFs, with average amounts invested in ETFs currently limited to not even 7%.

We now turn to an analysis of satisfaction with ETFs for these different categories, as well as to different ways of using ETFs. Our aim is to shed light on the reasons for the low amounts allocated to ETFs in certain asset classes.

Exhibit 1.7: What percentage of your investments is represented by ETFs or ETF-like products?

Equity

Government bonds

Corporate bonds

Commodities

Real Estate

Hedge funds

9.65%

6.73%

21.99%

16.46%

7.03%6.52%

CoreSatellite

0

5

10

15

20

25





2. ETFs in practice In this section, we examine both the use of ETFs by European asset managers and institutional investors and the levels of satisfaction with ETFs. Furthermore, ETFs stand out for a number of features; the ability to write and buy options with the ETF as the underlying, the lending of ETF units, shorting, and the use of inverse performance ETFs make ETFs different from other indexing instruments. In this section, we also analyse how these features are perceived by European investors and asset managers.

First of all, we are concerned with the satisfaction of ETF users. Satisfaction may differ from one asset class to the next, so for each class we attempt to elucidate respondent satisfaction with the performance and features of the corresponding ETFs. As a preliminary step, we determine the percentage or respondents who use ETFs for each asset class (exhibit 2.1.a).

Exhibit 2.1.a: Percentage of respondents using ETFs

Equity

Government bonds

Corporate bonds

Commodities

Real estate

Hedge funds

42% 40%

78%

48%

30%35%

0

10

20

30

40

50

60

70

80

As exhibit 2.1.a shows, respondents are more likely to use ETFs for equities than for any other asset class. For commodities and bond investments, ETFs are used by

close to 50% of respondents. Finally, for real estate and hedge funds, only about one-third of respondents use ETFs or ETF-like products.

It should be noted that exhibit 2.1.a. is different from the questions asked in section 1, since it shows the percentage of all respondents who use ETFs for different asset classes. The previous questions in section 1 attempt to show the relative importance of different types of ETFs within the same asset class. For example, as section 1 shows, 94% of all equity ETF users put broad-based equity ETFs in the core portfolio. Exhibit 2.1.a shows that 78% of all survey respondents are users of equity ETFs.

Exhibit 2.1.b shows the satisfaction withETFs for different asset classes. Only those respondents who use ETFs in the respective asset class are asked to report their satisfaction. Exhibit 2.1.b shows that, over all asset classes, a large majority of users are satisfied with their ETFs. Satisfaction is most pronounced for equity ETFs, which are also the most widely used, and least pronounced for ETF-like products on hedge funds, which are also the least used. Satisfaction is also very high for government bond ETFs (85%) and commodity ETFs (87%). Corporate bond ETFs, on the other hand, obtain relatively low satisfaction scores, with two-thirds of users stating that they are satisfied. With 77% of users reporting that they are satisfied, real estate ETFs obtain a satisfaction score that is considerably lower than that of commodity ETFs, but considerably higher than that of ETF-like products linked to hedge funds.




The reasons for satisfaction or dissatisfaction may be manifold. With corporate bonds, liquidity is naturally an issue and may pose a challenge to ETF providers. In alternative investment classes, except for commodities, for which there is a liquid futures market, liquidity issues may also resurface. In addition, constructing truly representative indices in alternative asset classes may be a challenge, especially when doing so involves attempts to attain the liquidity necessary for the construction of an instrument such as an ETF. Despite the existence of solutions to these problems (see Goltz, Martellini, and Vaissié 2007), current product offerings do not fully satisfy investors, as evidenced by the responses to our questionnaire.

Exhibit 2.1.b: If you use ETFs or ETF-like products, are you satisfied with them?

Equity

Government bonds

Corporate bonds

Commodities

Real estate

Hedge funds

85%

66%

92%87%

58%

77%

0

20

40

60

80

100

As mentioned above, ETFs stand out for a number of features. The following exhibits show how these features are perceived by European investors and asset managers. We ask in particular about the use of inverse-performance ETFs, options written on ETFs, short selling of ETFs, and the use of ETF shares in securities lending. Exhibits 2.2 to 2.5 make it increasingly

clear that the majority of respondents do not take advantage of these features.

As exhibits 2.2 to 2.5 show, inverse-performance ETFs attract the most attention among European investors and asset managers. 31% of respondents either use them or are considering doing so; the share of potential and current users for all other advanced features—options, short selling, and securities lending—is close to 10%.

Inverse-performance ETFs are supposed to provide investors with the inverse of the performance of an index, which is achieved through short selling. In addition, these ETFs provide investors with the money market interest on the amount invested and interest earned on the short position. We also ask those respondents who currently use inverse-performance ETFs to identify the reasons for doing so. It turns out that the main reasons for using them are to make tactical bets on downward movements in the underlying index and to meet short-term hedging objectives.

Exhibit 2.2: Do you use inverse ETFs as a hedging tool?

13.5%

17.1%

4.5%

10.8%

4.5%

49.5%

YesNoNo, but we will soon





Options on ETFs began trading on derivatives exchanges shortly after the introduction of ETFs. These instruments are limited to a relatively narrow range of the most successful ETFs. The possible advantages of these options include precise exposure to the underlying fund, minimum investments lower than those required by index options, as well as physical delivery of the underlying asset if the option is exercised (index options, by contrast, are settled in cash). We ask the respondents who currently use options on ETFs to identify their reasons for doing so. Acquiring leveraged exposure on movements in the underlying ETF and the structuring of capital guaranteed products through options-based portfolio insurance emerged as the two main reasons.

Exhibit 2.3: Do you use options on ETFs?

3.6%

4.5%

3.6%

13.5%

2.7%

72.1%



Unlike traditional index funds, ETFs may be sold short. Since ETFs can be borrowed and sold short, long/short strategies are possible. With these strategies, long/short exposure to different style or sector indices can be used to capitalise on return differentials between categories while maintaining low or zero exposure to market risk. Respondents who use short selling of

ETFs reply that they do so mainly to place such tactical bets or to make short-term downward adjustments of the portfolio weights of a given asset class.

Exhibit 2.4: Do you short ETFs?

4.5%

2.7%

10.8%

2.7%

8.1%

71.2%



ETF units held by an investor may be lent out to generate additional income for the portfolio. Interest paid by the borrower of the ETF may compensate for management fees and generate income above the management fees in the ETF. Most of those who do lend ETFs state that they do so as part of a securities lending programme with their custodians.

Exhibit 2.5: Do you lend your ETF units?

7.2%

3.6%

10.8%

1.8%

5.4%

71.2%






An interesting finding is the number of potential users, that is, respondents who state that they will soon use a given ETF feature. For most questions, the number of potential users exceeds the number of current users. The only exception is ETF lending, for which the number of potential users is slightly less than the number of current users. Overall, the large number of potential users relative to current users suggests that such advanced uses of ETFs will increase significantly in the future. In addition, the low levels of familiarity with all of these advanced uses suggest that if ETF users are made aware of the possibilities linked to ETF investing there will be a tremendous potential for growth.

We close out the section on practical ETF investing with two questions concerning actual trading in ETFs. Exhibit 2.6 shows how much of respondents’ ETF trading is done over the counter rather than on exchange. While most (66%) respondents do not trade a significant share of their ETF investments over the counter, 20% of respondents execute more than half of their ETF trading on OTC markets. On average, respondents use about three counterparties.

Exhibit 2.6: How much of your ETF trading is done OTC rather than on exchange?

1.1%

8.8%

9.9%

5.5%

8.8%65.9%

<10%10% to 25%25% to 50%

50% to 75%75% to 90%>90%




3. The pros and cons of ETFs, futures, total return swaps, and index funds In this section, we compare four investment instruments: ETFs, futures, total return swaps (TRS), and traditional index funds. All these instruments allow simple execution of trades in large baskets of stocks. Our criteria for evaluation are loosely based on Rubinstein’s (1989) early examination of such instruments. We look at the advantages and disadvantages of each instrument and then emphasise specific issues concerning total return swaps, futures, and ETFs. In addition, we assess the future use of these instruments by European institutional investors and asset managers to highlight developing trends.

3.1. Rating of ETFs, futures, total return swaps, and index funds according to the following criteriaIn this part, survey respondents are asked to rate exchange-traded funds, futures, total return swaps, and index funds according to various criteria. The responses we obtain allow a few general conclusions. First, in terms of liquidity, transparency, and cost, ETFs are considered advantageous although on some criteria they are less well regarded than futures. Second, ETFs are ranked highest for available range of indices and asset classes. Therefore, European investors and asset managers seem to be well aware of the diversity of exchange-traded funds, which has grown dramatically in recent years. Third, futures are the most serious challenger to ETFs, but ETFs are perceived as superior with regards to minimum subscription, operational constraints, and the tax and regulatory regime. Therefore, it appears that implementation concerns with futures (such as margin calls, and applying

exact allocations even for small-sized portfolios) give ETFs an advantage. Fourth, the respondents believe that ETFs perform much better than total return swaps on each criterion. This belief conflicts with that expressed by Lhabitant, Mirlesse, and Chardon (2006), who conclude that indexation with derivatives provides better performance than exchange-traded funds and that, when considering both costs and tracking error, swaps are the most efficient mechanism for tracking an index. These conflicting beliefs may be explained, to some extent, by a lack of familiarity with total return swaps, as a considerable share of respondents do not answer this particular question. Even among those who do, however, total return swaps are not considered superior.

When interpreting the results, one should bear in mind that a high percentage of survey participants do not answer the questions relative to the ranking for all instruments. Rather than adjusting the rankings for non-responses, we indicate the percentage of non-responses, since this allows us to draw conclusions as to the familiarity of survey participants with each instrument or evaluation criterion.

As exhibit 3.1.1 shows, 65% of respondents believe that futures are very good in terms of liquidity and no respondents state that liquidity is “poor”. The liquidity of exchange-traded funds is viewed as fairly good by 50% of respondents but poor by 9%. 39% of respondents view index funds as fairly liquid and 15% deem their liquidity poor. The liquidity of total return swaps, viewed as the least liquid of these instruments, is considered poor by 17% of respondents. These results show that respondents value the ability—unavailable with standard index




funds—to trade immediately with futures and ETFs. These findings dovetail with those of the literature, which highlights the “value of immediacy” of exchange-traded funds (see Comer et al. 2002).

Exhibit 3.1.1: Rating with regard to level of liquidity

Very good Fairly good Poor No answer0

10

20

30

40

50

60

70

80

65%

50%

16%

31%

39%

9%

17%

15%

10%

19% 22

%35

%

0%

32%

17% 24

%

ETFsFutures

Total return swapsIndex funds

Survey respondents express opinions on the cost of liquidity that are similar to their opinions on liquidity. Exhibit 3.1.2 shows that on this score futures are still held in highest esteem, followed by ETFs. Index funds and, in particular, TRS are viewed as weak.

Exhibit 3.1.2: Rating with regard to cost of liquidity

0

10

20

30

40

50

60

70

80

Very good Fairly good Poor No answer

59%

66%

22%

28%

36%

12% 23

% 22%

12%

20% 24

%38

%

0%

11%

11% 18

%

ETFsFutures


Exhibit 3.1.3 shows that, when it comes to costs, most respondents think futures are the best instrument. 62% think they are very good and only 5% consider them poor. 79% have

positive views of ETFs (very good or fairly good)—30% think they are very good and 11% consider them poor. Total return swaps are viewed as fairly good by 34% of respondents, but a sizeable fraction of investors (17%) think they are poor in terms of cost. Index funds do not get good ratings on cost when compared to ETFs and futures.

It must be kept in mind that the evaluation of the costs of these instruments is specific to the context in which they are used. In particular, the position size and frequency of trading determine the relative merits of each instrument. Kostovetsky (2003), for example, finds that for large investments, ETFs are preferable to index funds, while for small amounts, the high trading costs make ETFs less attractive unless the holding period is long. Our respondents express preferences for ETFs and futures over index funds, preferences that may be accounted for by their large position sizes, as indicated by the average amount of assets under management they report. Gastineau (2001) points out the reasons that make ETFs more cost efficient than index funds. First, ETFs are typically very large funds, allowing for economies of scale and, second, expenses for the transfer agency function of mutual funds are not incurred with ETFs.

Exhibit 3.1.3: Rating with regard to cost

0

10

20

30

40

50

60

70

80


62%

49%

14%

34%

36%

11% 17

% 24%

11%

20% 23

%37

%

5%

30%

12% 17

%

ETFsFutures





With respect to the reliability of tracking error, only a marginal percentage of respondents believe that these four instruments are poor, as shown in exhibit 3.1.4. Interestingly, at 82% good or very good, ETFs obtain the highest percentage of positive responses. They are followed by futures (74% of responses are positive). Futures, however, are considered very good by a higher percentage of respondents. 67% of respondents have positive views (good or very good) of traditional index funds. Total return swaps are considered good as well, although here more than one-third of respondents express no opinion.

Exhibit 3.1.4: Rating with regard to the reliability of tracking error

0

10

20

30

40

50

60


51%

49%

23% 28

%44

%

5% 5%7% 14

%22

% 25%

37%

4%

33%

30%

23%

ETFsFutures


ETFs clearly obtain the best rating for the available range of products, with 47% of respondents stating that the available range is very good. This finding is consistent with recent developments in the ETF industry, and has been stated as an advantage of ETFs in the literature (see Demaine 2002). ETFs now offer exposure to a wide range of indices, unlike futures, which have a less diversified product range. However, as exhibit 3.1.5 shows, futures still come in second, deemed very good as they are by

33% of respondents; they are followed by TRS and index funds.

Exhibit 3.1.5: Rating with regard to range availability

0

10

20

30

40

50


33%

34%

32% 36

%46

%

3% 7%11

%

16%

24% 26

%39

%

11%

47%

18%

17%

ETFsFutures


From exhibit 3.1.6., it can be seen that very few respondents (4%) believe that ETFs or futures are poor in terms of transparency. Total return swaps, on the other hand, are considered poor by 18% of respondents, and index funds by 9%. Looking at the percentage of respondents who view the instrument positively confirms that ETFs (80% positive) and futures (73% positive) are preferred to index funds (65% positive), which in turn are preferred to total return swaps (44% positive).

Exhibit 3.1.6: Rating with regard to the level of transparency

0

10

20

30

40

50

60

70

80


61%

38%

12%

32%

43%

4%4% 9%

16% 23

%39

%26

%39

%

18%

42%

12%

22%

ETFsFutures





ETFs are clearly the preferred instrument when it comes to the minimum subscription requirement (see exhibit 3.1.7). 59% of respondents consider ETFs very good, while only 2% consider them poor. The positive views of ETFs are to be compared with the views of futures, which are considered very good by 32% of respondents but poor by 9%. However, futures are only slightly behind traditional index funds. The highest percentage of respondents (30%) to express dissatisfaction (poor) with the minimum subscription went to total return swaps.

Exhibit 3.1.7: Rating with regard to minimum subscription

0

10

20

30

40

50

60


32%

22%

33%

22%

34%

9%2%

5%

17%

25%

40%

28%

39%

30%

59%

9%33

%

ETFsFutures


As exhibit 3.1.8 shows, exchange-traded funds are viewed less susceptible to operational constraints than the other three instruments. Indeed, half of our respondents believe that exchange-traded funds are very good in terms of such constraints and 32% consider them fairly good. Futures and traditional index funds are ranked behind ETFs, with less than 30% of respondents seeing them as very good, and 36% seeing them as fairly good. TRS are clearly perceived as the instrument most susceptible to operational constraints, with 27% of respondents viewing them as poor.

Overall, the first main result is that ETFs are rated better than index funds and futures. The latter two instruments actually obtain rankings that are very similar. The second main result for this question confirms that there is a pronounced difference between exchange-traded derivatives (futures) and over-the-counter derivatives (swaps).

Exhibit 3.1.8: Rating with regard to operational constraints

0

10

20

30

40

50


27%

32% 36

%23

%36

%

10%

1% 7%

18%

27%

41%

29%

39%

27%

50%

8%28

%

ETFsFutures


Exhibit 3.1.9 indicates that respondents prefer exchange-traded funds (80% view them positively) when it comes to the regulatory regime. Only 3% deem them poor on this score. Interestingly, ETFs also receive the fewest non-responses, suggesting familiarity among respondents with the applicable regulatory regime. Likewise, index funds and futures are mostly regarded positively. However, there is considerable discontent with the regulatory framework for total return swaps.

Survey responses fall into line with the conclusions of academic research papers (Poterba and Shoven 2002) and articles in the professional press (Stock 2006) that stress the tax efficiency of exchange-traded funds. In fact, the majority of respondents (71%) consider ETFs either very good or fairly good on this score.




Exhibit 3.1.9: Rating with regard to regulatory regime

0

10

20

30

40

50

60


45%

28%

25%

33%

28%

4%3% 2%

17%

26%

41%

29%

28%

18%

52%

8%42

%

ETFsFutures


Index funds and futures are considered very good or fairly good by slightly fewer respondents (63% and 64% respectively). Futures, however, are classified as very good for their tax regime by the greatest percentage of respondents. Total return swaps again receive fewer positive responses than the other instruments, which, as the number of non-responses indicates, may be linked to the relative unfamiliarity of respondents with these instruments. In fact, TRS also receive fewer negative responses than the other instruments.

Exhibit 3.1.10: Rating with regard to tax regime

0

10

20

30

40

50


31%

50%

33%

38%

44%

6%9%

6%

20%

30%

42%

31%

5%

21%

14%

19%

ETFsFutures


3.2 Issues with specific instruments

3.2.1. TRSConsidering responses across all criteria, we find, broadly, that total return swaps (TRS) are viewed more poorly than the other three instruments. In addition, TRS are the object of the highest percentage of non-responses, suggesting that respondents are relatively unfamiliar with them. Our survey addresses two outstanding issues with TRS: the requirement for over-the-counter trading and the associated counterparty credit risk. As shown in exhibits 3.2.1 and 3.2.2, trading over the counter is problematic for much the same percentage of respondents for which it is not problematic, but opinions of counterparty risk are not split down the middle (60% of respondents consider it problematic; 15% do not). Interestingly, the percentage of non-responses (22% and 24%) to these specific questions is significantly lower than the percentage of non-responses to the general questions above, where non-responses on TRS come to approximately 40% for most questions. These figures suggest that respondents do not consider the advantages and disadvantages of TRS in detail because counterparty risk dissuades them from using them in the first place.

Exhibit 3.2.1: Is the fact that they are traded over the counter a problem for you?

41.4%21.6%

36.9%

YesNo

No answer




Exhibit 3.2.2: Is the counterparty risk a problem for you?

60.4%24.3%

15.3%

YesNo

No answer

3.2.2. FuturesIn the questions comparing all four instruments above, futures fared remarkably well and can be viewed as the greatest rival of ETFs when implementing indexed investing. When comparing futures and ETFs, a drawback of futures is that they are derivative instruments, require roll-over transactions, and involve margin calls. In the general questions above, futures fall behind ETFs in the evaluation of the operational constraints linked to each instrument.

When asked directly, respondents (to the tune of 36%) report that margin calls and the roll-over of positions are the most serious problems with futures. That futures are derivatives is seen as problematic by only 23% of respondents. Overall, the significant percentage of respondents seeing margin calls and position-rolling as problems jibes with the relatively jaundiced view of futures when it comes to the operational constraints they face. On the other hand, 50% of respondents do not see margin calls and position-rolling as problems, a finding that strengthens the perception of futures as a possible alternative to ETFs. Some 14% of respondents do not express their views on these issues (see exhibits 3.3.1 to 3.3.3).

Exhibit 3.3.1: Is the fact that they are derivative instruments a problem you?

22.5%14.4%

63.1%

YesNo

No answer

Exhibit 3.3.2: Is the requirement to roll over positions a problem for you?

36%13.5%

50.5%

YesNo

No answer

Exhibit 3.3.3: Are the margin calls a problem for you?

36%13.5%

50.5%

YesNo

No answer

3.2.3. ETFsWe ask questionnaire respondents for their opinions on pricing errors with respect to the net asset value (NAV) of the ETF, on the




perceived advantage of securities lending, on active and passive ETFs, and on the ways of using ETFs to replicate indices.

Possible mispricing with respect to the net asset value (NAV) was of concern to 62% of respondents (see exhibit 3.4.1). This finding is somewhat surprising, as an empirical study by Engle and Sarkar (2006) finds that US ETFs have highly efficient prices, though their conclusions for international ETFs are different. In fact, these authors find that the premiums or discounts on fund NAVs are typically small and disappear very quickly, supporting the view that the creation and redemption mechanism of ETFs effectively limits and destroys arbitrage opportunities. However, Engle and Sarkar (2006) also underline that the timing difference between the fund NAV and the ETF price has to be properly accounted for. It may be that the respondents to our survey associate the problem of non-synchronous observations between fund prices and fund NAVs with the problem of mispricing, which is in fact another problem altogether.

Exhibit 3.4.1: Is the mispricing with regard to the net asset value problem for you?

62.2%12.6%

25.2%

YesNo

No answer

It is interesting to note (exhibit 3.4.2) that 27% of respondents see securities lending as an advantage, a finding that suggests

the potential of developing ETF securities lending.

Exhibit 3.4.2: Is securities lending an important advantage when considering ETFs?

27%12.6%

60.4%

YesNo

No answer

As mentioned above, the vast range of ETFs is one of their perceived advantages. Most ETFs are passively managed and replicate indices. However, actively managed ETFs have been launched and are now quoted on European stock exchanges. As exhibit 3.4.3 shows, the large majority of respondents prefer passive ETFs, though active ETFs are preferred by slightly more than 10% of respondents. Active ETFs fly in the face of the investment philosophy that would have the manager eschew stock-picking and concentrate on asset allocation. Active ETFs allow immediate trading in actively managed funds. Therefore, the logical application of such funds would be short-term manager selection, not asset allocation.

Exhibit 3.4.3: What types of ETF do you prefer?

10.8%12.6%

76.6%

Active ETFsPassive ETFs

No answer




We ask those who prefer passive ETFs what method of index tracking they prefer. In fact, in addition to pure replication (matching the composition of an index), ETFs may rely on synthetic replication (derivatives written on the index) or statistical replication (using an optimisation procedure to match the returns on the index with a selection of securities that may differ from the index). A majority of respondents (51%) express a preference for conventional pure replication ETFs (see exhibit 3.4.4). Synthetic replication and statistical replication are still seen as less attractive than full replication. Synthetic replication through derivatives is, however, significantly more popular (preferred by 20% of respondents) than statistical replication (preferred by 7%). The low acceptance of statistical replication may constitute a potential barrier to the further expansion of ETFs in asset classes with low liquidity, where full replication may not be feasible.

Exhibit 3.4.4: If the answer is 'Passive ETFs', which of the following do you prefer?

51.4%

21.6%

7.2%

19.8%

Pure replicationSynthetic replication (i.e., use of derivatives)Statistical replication (i.e., sampling)No answer

3.3. Methods for evaluating liquidity and tracking errorWhile we have established above how European investors and asset managers rate ETFs and other indexing vehicles with

respect to their liquidity, their reliability of tracking error, and other criteria, we have not addressed how the fulfilment of these criteria is measured by respondents. This section addresses this issue.

Exhibit 3.5 shows that respondents generally use standard measures of tracking quality, such as the tracking error between the instrument’s returns and the index returns (68%) or the correlation between the two assets (44% of respondents). That 14% of respondents express no opinion on this issue may be the result of the assumption that the tracking quality of modern instruments is good (see the good scores for all instruments above); many respondents do not assess this issue in further detail. Interestingly, some respondents use quite advanced measures of tracking error such as the asymmetric or downside tracking error (see Amenc, Malaise, and Martellini 2004) or co-integration analysis (see Engle and Sarkar 2006 for an application to the tracking quality of ETFs). On the other hand, 23% simply compare the mean returns, which is a straightforward way of dealing with non-synchronous data for the two assets.

Exhibit 3.5: How do you assess the tracking quality of the different investment supports?

0

10

20

30

40

50

60

70

80

Correlationanalysis

Tracking erroranalysis

Asymmetrictracking error

analysis

Comparison ofmean returns

Co-integrationanalysis

No answer

Other

44%

68%

12%

23%

9%4%

14%




The second key issue with indexing instruments is liquidity. Practitioners, of course, are highly familiar with liquidity, but the finance literature has yet to come to a consensus on theory and on empirical methodology. For example, practitioners have long used a number of liquidity measures, but academic articles continue to debate the merits of the multiple possible measures of liquidity.

Survey respondents rely largely on market spreads (66%), turnover (50%), and assets under management (30%) as measures of liquidity (see exhibit 3.6). Interestingly, a significant percentage of respondents (18%) rely on the co-movement of liquidity in the instrument and the returns on the index. This measure has been proposed in Acharya and Pedersen (2005) and addresses the fact that low liquidity becomes a more severe problem in times of negative returns, that is, when low liquidity becomes an obstacle to selling off an asset in down markets.

Exhibit 3.6: How do you assess the liquidity of the different investment supports?

0

10

20

30

40

50

60

70

80

Market spreads

Turnover

AUM

Co-movement ofliquidity and returns

(representing concern over low liquidity in "bad times")

No answer

Other

66%

50%

30%

18%

3%

13%

3.4. Looking aheadMost of our survey questions ask respondents to assess the quality of competing indexing

instruments and give information about their use of these instruments. The aim of the last questions in the survey, however, is to elicit answers that will shed light on future developments.

First, we ask respondents to identify the instruments they are most likely to use in the future. As exhibit 3.7 shows, respondents report in general that the use of all four instruments will increase. In fact, the number of respondents who plan to increase their use of a given instrument is far greater than the number of respondents who plan to decrease their use of the same instrument. This may be seen as evidence that European investors and asset managers have a growing interest in “indexing” and are beginning to turn away from stock-picking.

Second, it can be seen that ETFs will benefit most from the increased use of indexing instruments. 69% of respondents plan to increase their use of ETFs, while only 3% plan to decrease it. 36% of respondents plan to increase their use of futures, while 2% plan a decrease. Only 18% plan to increase their use of total return swaps; 9% plan to decrease it. Index funds are the only instrument for which an increase in future use is not pronounced: 23% plan an increase and 19% a decrease. Overall, it seems that the anticipated increase in ETF use will not necessarily hinder the further development of other indexing vehicles.

From exhibit 3.7, an increase in the future use of ETFs seems inevitable. We also ask those surveyed to identify the area in which they predict the greatest increase in the use of ETFs. Exhibit 3.8 shows that the greatest increase (chosen by 44% of respondents)




will be in the area of accessing new asset classes.

Exhibit 3.7: How would you predict you future usage of the tollowing instruments

0

10

20

30

40

50

60

70

80

Increase Stay the same Decrease No answer

36%

15.3

%42

.3%

38.7

%36

.9%

1.8%2.7%

9%

20.7

%

12.6

% 19.8

%34

.2%

18.9

%

69.4

%

18% 23.4

%

ETFsFutures


The other main areas of increased use are optimal portfolio construction and risk management/hedging. An increase in using ETFs for cash equitisation is predicted by approximately 4% of respondents. Respondents’ predictions of future uses seem to justify the strategy of ETF providers, who aim to cover new asset classes such as listed real estate, listed private equity, and commodities. Many respondents, however, see the biggest increase in more technical uses of ETFs, such as portfolio construction.

Exhibit 3.8: In which area do you predict the greatest future increase in your use of ETF's?

44.1%

14.4%

3.6%

12.6%

25.2%

Exposure to new asset classes through ETFsConstructing optimal portfolios of ETFsHedging and risk management with ETFsCash equitisation with ETFsNo answer

To conclude this section, we now turn to the new products that European investors and asset managers would like to see developed. As shown in exhibit 3.9, emerging markets equity and alternative asset classes are the top concerns of respondents. About half of them would like to see new products developed in this area. With 32% of respondents, emerging market debt and commodities also rank high on the wishlist. Roughly a quarter of respondents would like to see equity style and high yield bond ETFs, while ethical investment (10%) and actively managed equity (16%) are not in high demand.

Exhibit 3.9: What type of ETF product would you like to see developed in the future?

0

10

20

30

40

50

Emergingmarkets

equity ETFs

Emergingmarkets

bond ETFs

AlternativeAsset

Classes

Commodities

Equitystyle funds

High yieldbond funds

Actively managed

equity ETFs

Ethical ETFs

No answer

50%

32%

47%

32%

27%

22%

16%

10%

19%


2. Background




After having provided some new insights into the current use of ETFs by European asset managers and investors, we now turn to an overview of novel ways of using ETFs in portfolio management. In the present section, we will introduce the optimal organisation of portfolio management, i.e., the core-satellite approach, first in a static context and then extended to a dynamic process. We will provide a range of examples illustrating how various ETFs may be used in the context of dynamic risk budgeting.

1. The Core-Satellite ApproachThe core-satellite approach has been advocated by consultants as a new paradigm in investment management and has received increasing attention from investors, who are trying to structure their portfolio management process in a coherent manner. The objective of this section is to clarify what is meant by the core-satellite approach. We introduce the core-satellite portfolio management paradigm and provide a theoretical justification for its adoption by investors.

1.1. Introduction to core-satellite portfolio managementMost of the time, active management is subject to very restrictive constraints, with managers allowed to deviate only slightly from their benchmark. As a result, only a limited part of the portfolio is actively managed, while the essential part of the portfolio passively replicates its benchmark, resulting in an overall portfolio with a low tracking error. For example, an active manager with a 5% tracking error constraint is in fact 95% passive. Meanwhile, high active management fees

are paid on the whole portfolio. On the one hand, this approach is not efficient and leads to high management fees. On the other hand, it is not likely to favour outperformance, since managers are prevented from actually using their abilities to make active bets. Instead of proceeding in this way, it is possible to use the core-satellite approach to get the best of both active and passive management.

The core-satellite approach involves dividing the portfolio into a passively managed core that fully replicates the investor’s specifically designed benchmark and an active component made up of one or more satellites that is allowed higher tracking error. The requirements for the core portfolio are therefore quite different from those for the satellite portfolio.

The raison d’être of the core portfolio is to provide the best possible risk-return trade-off while respecting the constraints faced by the investor. For a pension fund, for example, the core portfolio will be heavily influenced by the liabilities and the factors affecting the evolution of liabilities. More generally, the investor may have a certain number of constraints with respect to long-term risk exposures, which must be taken into account when constructing the core portfolio. The core portfolio can be made up of a pure index fund. However, it is important to note that any long-term allocation to different indices can constitute the benchmark, which is not necessarily limited to market indices, but can also be made up of enhanced index products. The satellite portfolio does not face the same constraints as the core portfolio.

BackgroundNew Risk Budgeting Techniques: Applications with ETFs



In the satellite, the investor seeks to outperform the core. The counterpart of the potential for outperformance is the risk of deviating from the investor's long-term risk exposure as defined in the core portfolio. While satellite portfolios are often actively managed, and typically invested in less efficient markets that require more specialised managers, any source of outperformance may be used in the satellite portfolio. This includes indexing instruments for specific asset classes or styles that are expected to yield superior returns.

The choice of allocation to the core and to the satellite allows the investor to target a level of tracking error for the entire portfolio, while authorising potentially large deviations from the long-term target exposure in a limited part of the portfolio, the satellite.

This approach allows for the improvement of investment management efficiency. The role of the core is to control manager risk, while limiting management fees, whereas the role of active satellites is to provide investment diversification and to outperform the benchmark. In other words, beta is managed in the core and alpha in the satellite(s). In its purest form, this approach leads to an investment in market-neutral managers who provide only portable alpha benefits without passive exposure to the index. High management fees will therefore be paid only for the part of the portfolio that is truly actively managed and supposed to generate abnormal returns, while management fees will be reduced for the passively managed core, so active managers will not be compensated for their active exposure to rewarded

sources of risk. The core-satellite approach to portfolio management has become the gold standard for asset managers in search of performance. This technique is also well suited to pension funds, which are typically active funds with a low tracking error with respect to their benchmark.

In what follows, we first present in greater detail the analytical formulation of the core-satellite portfolio design. We then illustrate the benefits of the core-satellite approach. Finally, we describe how ETFs appear to be useful instruments for deriving core-satellite portfolios in an optimal way.

1.2. The arithmetic of core-satellite investingIn this section, we will set the problem in a simple mean-variance analysis to show how to derive the optimal proportions to invest in the satellite(s) and in the core.

We first consider a core-satellite approach with a single satellite portfolio, the simplest case. The arithmetic of a core-satellite approach is then straightforward. The overall portfolio, a combination of the core and the satellite, is expressed as follows:

P =wS + 1−w( )Cwhere w is the fraction invested in the satellite (S), and 1-w is the fraction invested in the core (C). Computing the difference between the portfolio and its benchmark as follows:

P −B =wS + 1−w( )C −B =w S −B( ) + 1−w( ) C −B( )

P −B =wS + 1−w( )C −B =w S −B( ) + 1−w( ) C −B( )

and assuming for the sake of simplicity




that the core replicates the benchmark perfectly, we get C = B, then we have:

P −B =w S −B( )Using this formulation, we can now calculate the portfolio tracking error with regard to its benchmark B as follows:

This formulation makes it possible to assess the profitability of a core-satellite portfolio with regards to tracking error management. Suppose, for example, that an investor is allowed a 2.5% tracking error. There are then two options. Either the investor hires one manager with a tracking error equal to 2.5% for the entire portfolio, or the investor forms a passive core portfolio, consisting of 80% of the overall portfolio, and leaves 20% in an aggressively managed satellite with a tracking error chosen so that the overall portfolio tracking error meets the risk budget constraint, here, a 12.5% tracking error, as given by the following computation:

TE( S) =

TE P( )w

=2.5%20%

=12.5%

The next step consists of deriving the optimal proportions w* to invest in the satellite and in the core. This problem can be solved using simple mean-variance analysis by optimising the following utility function: 1

where IR(P) is the information ratio of the portfolio P with respect to the benchmark:

IR P( ) = E(P −B)

σ(P −B)=

E(P −B)TE P( )

(see Grinold and Kahn 2000) and λ is the investor’s risk aversion.

It should be noted that when the core portfolio perfectly replicates the benchmark, the information ratio of the overall portfolio IR(P) is actually independent of the proportions in the core and in the satellite and equal to the information ratio of the satellite portfolio IR(S) (as long as the proportion w invested in the satellite is strictly positive). This can be easily seen from the following:

IR P( ) = E(wS + 1−w( )C −B)

σ(wS + 1−w( )C −B)=

wE( S −B)wTE S( )

= IR S( )

Replacing TE(P) by its expression as a function of TE(S) in equation (1), we may rewrite the optimisation program as follows:

U w( ) = IR ×w×TE( S) − λw 2TE 2( S )

and the first order condition, enabling us to obtain the optimal value of the proportion w* to be invested in the satellite portfolio, reads:

∂U∂w

w*( ) = 0 ⇒w* =IR

2λTE S( )

For example, suppose that the tracking error of the active fund is 5%, that the information ratio IR is 0.5, and that the coefficient of risk-aversion with respect to relative risk λ is 20. Then, the optimal proportion invested in the active portfolio is:

w* =

IR2λTE S( )

=0.5

2 ×20 ×5%= 25%

The tracking error resulting from this allocation is: TE P( ) = 25% ×5% =1.25%


TE P( ) = var P −B( ) =w var S −B( ) =wTE S( )

U = E(P −B) − λσ 2(P −B) = IR P( ) ×TE P( ) − λTE 2 P( )



So far we have assumed that the satellite portfolio is a single portfolio. As explain above, the satellite is often divided among several managers. The point is then to find the optimal allocation to the active managers. Extending the analysis to the case of a satellite S invested in a number n of active portfolio managers Si

according to the proportions wi is straightforward.

If S = wi Si

i=1

n

∑ ,

the excess return on the satellite portfolio is then:

S −B = wi Si −B( )

i=1

n

∑

and the tracking error of the satellite portfolio reads:

TE S( ) = wiwjσij

i , j=1

N

∑ −2 wiσiB +σB2

i=1

N

∑⎛

⎝⎜⎞

⎠⎟

12

where σij is the covariance between portfolio managers Si and Sj , and σB is the volatility of the benchmark.

One can then find the optimal fraction invested in each active manager within the satellite portfolio so as to achieve the highest possible information ratio. One can show (see Scherer 2002) that the optimal condition is that the ratio of return to risk contribution is the same for all managers:

wkαk

wk2σαk

2 + wkwjσkjj

∑⎛

⎝⎜⎞

⎠⎟

TE( S)

=wl αl

wl2σα l

2 + wlwjσlj

∑⎛

⎝⎜⎞

⎠⎟

TE( S)

wkαk

wk2σαk

2 + wkwjσkjj

∑⎛

⎝⎜⎞

⎠⎟

TE( S)

=wl αl

wl2σα l

2 + wlwjσlj

∑⎛

⎝⎜⎞

⎠⎟

TE( S)

1.3. Benefits of the core-satellite approachThe core-satellite approach allows separate control of the tracking error of the satellite and that of the core, while respecting the authorised level of deviation from the benchmark. This strategy may be used by institutions wanting to provide diversified asset management, while holding on to potential for higher returns with a selection of active asset management strategies. This approach provides more freedom than usual benchmarked active asset management, in which tracking error must remain within a defined range and significant deviation from the composition of the benchmark is not possible. If passive and active management are distinct, active management may be allowed higher tracking error. For the level of tracking error required for the overall portfolio, one need only strike the proper balance between the tracking error of the active part of the portfolio and its weight in the overall portfolio. For a given level of tracking error, the more the satellite part of the portfolio is managed offensively (high tracking error), the more its contribution to the overall portfolio will be limited. It is then possible to profit from very dynamic management of a part of the portfolio, which is a source of value added for the portfolio performance, while keeping levels of tracking error required for benchmarked asset management. Moreover, separating the portfolio management into active and passive parts can reduce management fees, as they are paid only by the active part of the portfolio.

Assume, for example, that an investor has a relative risk tracking error budget equal to 5%. The first solution is to allocate 100%




of the portfolio to an active manager who will commit to respecting the 5% tracking error constraint. We assume that the management fees charged by this active manager are equal to 40 basis points. The second solution consists of allocating 75% of the portfolio to a purely passive product, e.g., an exchange traded fund (ETF) or preferably to a strategy that is based on an efficient benchmark, and allocating the remaining 25% of the portfolio to a 20% tracking error manager. We assume that the management fees of a core portfolio passively invested in an ETF are equal to 16 basis points.

Exhibit 1 (below) illustrates the benefits provided by the core-satellite approach in terms of tracking error control and management fees.

The second solution, which is consistent with a core-satellite approach to active asset management, offers two benefits. First, allowing the active manager to deviate significantly from the benchmark leads to a better use of the manager’s skills. If the manager has reliable views on market patterns and makes good bets on future trends, a 5% tracking error constraint leaves him with too little room for active decisions consistent with these views. Beating the market is notoriously tough, and attempting to do so with one

hand tied behind your back is hardly the ideal starting point. The second benefit of the core-satellite approach is that it allows for a clear distinction between the value added by the design of the strategic asset allocation represented by the benchmark (core portfolio) and the outperformance generated by active portfolio management.

Finally, the core-satellite approach makes it possible to control management fees. With our realistic assumptions of 40 basis points for the active manager and 16 basis points for the core, we get total fees for the entire portfolio of 22 basis points, lower than the fees typically charged by active managers with the same tracking error constraints (5%). With the great competition among asset managers, such a difference could prove decisive.

2. The dynamic core-satellite portfolio processIn this section, we describe the methodological backbone of dynamic risk budgeting—the dynamic core-satellite approach—a novel approach to risk management first described in Amenc, Malaise, and Martellini (2004). It is of interest that—through this non-linear risk management technology—payoffs that involve a type of relative return guarantee can be achieved through dynamic trading in ETFs. We provide two categories of


Exhibit 1. Benefits of the core-satellite approach

Core Satellite Global

Weight 75% 25% 100%

Tracking Error 0% 20% 5%(0%×0.75+20%×0.25)

Management fees 16bps 40bps 22bps(16×0.75+40×0.25)

This table illustrates how the core-satellite approach provides benefits to asset managers in terms or tracking error control and management fees optimisation.



applications using a range of equity and fixed income ETFs. The first set of examples deals with relative risk management, while the second extends the approach to the design and management of absolute return funds.

2.1. Dynamic core-satellite investing 2.1.1. Dynamic management of the tracking error budgetThe core-satellite concept essentially makes it possible to manage the tracking error of the overall allocation. If the investor has a given tracking error budget, the target tracking error can be achieved by defining static proportions to be invested in the core and in the satellite. However, management of the tracking error budget can also be made dynamic; the proportion invested in the active portfolio can vary as a function of the current cumulative outperformance of the global portfolio with respect to the benchmark. This objective is achieved by transporting the traditional constant proportion portfolio insurance method (CPPI) to the context of core-satellite portfolio management, so as to allow for more efficient control of relative risk.

The standard CPPI procedure, which was introduced by Black and Jones (1987) and Black and Perold (1992), allows for the production of option-like positions through systematic trading rules. This procedure dynamically allocates total assets to a risky asset in proportion to a multiple of the cushion, i.e., the difference between current wealth and a desired protective floor. This produces an effect similar to that of owning a put option. With this strategy, the portfolio’s exposure tends to zero as the cushion approaches zero; when the cushion is zero, the portfolio is

completely invested in cash. Thus, in theory, the guarantee is perfect: the strategy of exposure ensures that the portfolio never descends below the floor, in the event that it touches the floor, the fund is “dead”, i.e., it can deliver no performance beyond the guarantee.

This procedure of constant proportion portfolio insurance techniques, originally designed to ensure absolute performance, can be used with relative returns. Amenc, Malaise, and Martellini (2004) describe a new method, which dynamically adjusts the fractions invested in the core and in the satellite(s), that allows investors to gain full access to good tracking error, while keeping bad tracking error at bay. This method, a structured form of active management, appears to be a natural extension of constant proportion portfolio insurance techniques.

An approach similar to standard CPPI can be taken to offer the investor a relative performance guarantee and a limit on underperformance with respect to the benchmark. The techniques of traditional CPPI still apply, provided that the risky asset is re-interpreted as the satellite portfolio, which contains relative risk with respect to the benchmark, and the risk-free asset is re-interpreted as the core portfolio, which contains no relative risk with respect to the benchmark.


Exhibit I.4: Traditional CPPI versus Relative Approach CPPI

This table compares the traditional CPPI to the relative approach CPPI

Traditional CPPI Relative Approach CPPI

Risky Asset Satellite Portfolio

Risk-free Asset Core Portfolio



Suppose, for example, that the benchmark is a passive investment, e.g., a bond index. The guarantee is set at 90% of the benchmark value and we assume that the multiplier is equal to four.

At the initial date T0, portfolio value and benchmark value are normalised at 100, with a floor set at 90% of the benchmark value. The floor is thus 0.9×100 = 90. The cushion is therefore equal to 100 – 90 = 10. The investment in satellite is then 10 × 4 = 40, which results in 100 – 40 = 60 in the core. On date T1, let us assume that the difference between the satellite and the benchmark is +10%, resulting, for example, from the following scenario: S = 0%, C = -10%. In this case, the position invested in the core has decreased by 10% from 60 to 54. Besides, the active portfolio value has remained stable at 40, while the benchmark has also decreased by 10%, from 100 to 90. The difference between the fund value (94 = 54 + 40) and the benchmark value (90) is now equal to 4. The floor has dropped from 0.9 × 100 to 0.9 × 90 = 81. Thus the cushion is now 94-81=13. The new optimal fraction to invest in the satellite is 13 × 4 = 52, which leaves 94 – 52 = 42 in the core. On date T1 the resulting allocation is therefore 52/94 = 55% in the satellite and 42/94 = 45% in the core.

Suppose, on the other hand, that the difference between the satellite and the benchmark is -10%, a result of the following scenario: S = 0%, C = +10%. In this case, the position invested in the core has increased by 10%, from 60 to 66. The active portfolio value has remained stable at 40, while the benchmark has also increased by 10%, from 100 to 110. The floor is now at 0.9×110 = 99. The

difference between the fund value (106 = 66 + 40) and the floor (99) is now equal to 7, meaning that the cushion has decreased from its initial value of 10. The new optimal fraction to invest in the satellite portfolio is 7 × 4 = 28, which leaves 106 – 28 = 78 in the core portfolio. On date T1 the resulting allocation is therefore 28/106 = 26% in the satellite and 78/106 = 74% in the core.

As this example shows, the method leads to an increase in the fraction allocated to the satellite (from 40% to 55% in the example) when the satellite has outperformed the benchmark. Indeed, the accumulation of past outperformance has resulted in an increase in the cushion, and thus greater potential for a more aggressive (and hence higher tracking error) strategy in the future. If, on the other hand, the satellite has underperformed with respect to the benchmark, the method leads to a tighter tracking error strategy (through a decrease of the fraction invested in the satellite portfolio) in an attempt to ensure that the relative performance objective is met.

This approach allows dissymmetric mana-gement of tracking error, ensuring that the underperformance of the portfolio with respect to the benchmark will be limited to a given level, while letting the investor gain fuller access to the excess returns potentially generated by the active portfolio.

The advantage of this approach is that a dynamic version of core-satellite allocation allows an investor to truncate the relative return distribution so as to allocate the probability weights away from severe relative underperformance to the profit of more potential for outperformance.




2.1.2. Dynamic core-satellite allocationCore-satellite portfolios are usually constructed by placing assets that are supposed to outperform the core in the satellites. However, if economic conditions become temporarily unfavourable for these assets, they may underperform the core. The dynamic core-satellite approach described above makes it possible to reduce a satellite’s impact on performance during a period of relative underperformance, while maximising the benefits of the periods of outperformance.

Indeed, observation of investor behaviour shows that investor expectations are not necessarily symmetric. When stock market indices perform well, investors are happy to be engaged in relative return strategies, but when these indices perform poorly, they prefer absolute return strategies. Techniques such as Value-at-Risk minimisation or volatility minimisation allow only symmetric risk management. For example, the minimum variance process leads to a renunciation of part of the upside potential in the performance of commercial indices in exchange for lower exposure to downside risk, through tracking error constraints. While this strategy allows for long-term outperformance, it can lead to significant short-term underperformance. It is also very hard to recover from severe market drawdowns. In what follows, we introduce a set of techniques to focus on asymmetric risk management.

From an absolute return perspective, it is possible to propose a trade-off between the performance of the core and satellite. This trade-off is not symmetric, as it consists of maximising the investment

in the satellite when it is outperforming the core and, conversely, of minimising the weight of the satellite when it underperforms the core. The aim of this kind of dynamic allocation is to outperform static core-satellite allocation. This dynamic allocation first requires a lower limit on underperformance with respect to the benchmark on the terminal date, i.e.,

V(T) >kB(T) , where k is lower than one, e.g., k = 90%, and B(t) is the benchmark value at date t. Then, it is necessary to provide access to potential outperformance of the benchmark by investing in a satellite whose value on date t is denoted by S(t).

As mentioned above, Amenc, Malaise, and Martellini (2004) introduced a method, known as the dynamic core-satellite management process, which allows asymmetric tracking error management. This objective is achieved by a suitable extension of the CPPI to relative risk management. The concept of this process was introduced above. Let V(t) be the portfolio on date t. It can be broken down into a floor and a cushion, according to the relation V(t) = F(t) + C(t). The floor is given by F(t) = kB(t). Take the investment in the satellite, i.e., the risky asset in a relative context, to be E(t) = mC(t), with m being a constant multiplier, while the remaining part of the portfolio V(t) - E(t) is invested in the benchmark.

The process for cushion growth tells us about the upside potential and allows us to calibrate an optimal value for m.

dCt =dVt −dFt = Et ×

dSt

St

+ (Vt −Et ) ×dBt

Bt

−dFt





It is useful to write explicitly the final value of the portfolio as well as the cushion. We consider a floor given as F(t) = e-r(T-t)K,where K is the guaranteed capital. Let the overall portfolio value be A(t). Moreover, we consider a risky portfolio that is invested in equity. The fraction of the portfolio invested in equity E(t) is given by the fixed rule E(t)=mC(t)=m(A(t)-F(t)), while the remainder of the portfolio A(t)-E(t) is invested in the risk-free asset.

We have the following dynamics for asset and portfolio value

dSt

St

= μdt + σdWt

dBt

Bt

= rdt

dAt = Et ×dSt

St

+ At − Et( ) × dBt

Bt

The value of the portfolio at terminal date is then A(T) = F(T) + C(T) = K + C(T).

To estimate C(T), note that

or

One can solve the previous stochastic differential equation to obtain

To conclude, let us reiterate the rationale behind the dynamic core-satellite approach, before applying the approach to a practical portfolio management context starting with the next section. The core makes it possible to respect investors' long-term risk return objectives, while the satellite provides access to upside potential. The dynamic allocation process will allow for systematic increases in exposure to the satellite portfolio when it does well, while controlling risks by shifting to the core when the satellite does poorly. Thus, this approach allows an investor to truncate the relative return distribution so as to allocate the probability weights away from severe relative underperformance in favour of more potential for outperformance.

As a result, the approach allows asymmetric tracking error management, ensuring that the underperformance of the portfolio with respect to the benchmark will be limited to a given level, while giving the investor access to the excess returns potentially generated by the active portfolio.

2.2. A relative return perspective: applications with ETFs on European equity indicesWe take an investor who is concerned with portfolio risk relative to his benchmark or core portfolio. In the two examples that follow, we take an investor who chooses a large-cap euro-zone equity index, the

dCt= dA

t− dF

t= E

t

= mCt

×dS

t

St

+ At

= Ct+ F

t

− Et

= mCt

⎛

⎝

⎜⎜

⎞

⎠

⎟⎟×

dBt

Bt

− dFt

= Ft×

dBt

Bt

dCt

Ct

= mdSt

St

+ (1− m)dBt

Bt

⎛

⎝⎜⎞

⎠⎟=

dBt

Bt

+ mdSt

St

−dBt

Bt

⎛

⎝⎜⎞

⎠⎟

CT = C0

ST

S0

⎛

⎝⎜⎞

⎠⎟

m

exp r − m r −12σ 2⎛

⎝⎜⎞

⎠⎟− m2 σ

2

2

⎛

⎝⎜⎞

⎠⎟T

⎡

⎣⎢⎢

⎤

⎦⎥⎥

AT = FT + C0

ST

S0

⎛

⎝⎜⎞

⎠⎟

m

exp r − m r −12σ 2⎛

⎝⎜⎞

⎠⎟− m2 σ

2

2

⎛

⎝⎜⎞

⎠⎟T

⎡

⎣⎢⎢

⎤

⎦⎥⎥

dCt =mCt ×

dSt

St

+ (Ct +Ft −mCt ) ×dBt

Bt

−Ft ×dBt

Bt

dCt =Ct m

dSt

St

+ (1−m)dBt

Bt

⎛

⎝⎜⎞

⎠⎟




Euro Stoxx 50, as his core portfolio. He would like to enhance his performance by adding a satellite portfolio that outperforms the benchmark. While he may choose a satellite made up of actively managed funds or alternative asset classes, a more straightforward way of enhancing performance is to seek exposure to well known risk premia in equity markets. These risk premia have also been labelled “anomalies” since some stock portfolios earn returns in excess of what is justified by their market exposure or beta.

Perhaps one of the most striking empirical findings in asset pricing over the last decade is that there are a number of empirical anomalies, a finding that suggests that equity performance is not well described by standard capital market equilibrium models such as Sharpe's (1964) CAPM. In particular, it has been reported (e.g., Fama and French 1992) that both the market capitalisation (size factor) and the ratio of book-to market equity (B/M factor) account for a significant fraction of the cross-sectional difference in expected returns.

The existence and significance of value and size premia are currently at the heart of debate in asset pricing research; there is still no consensus on the economic interpretation of these empirical results.

One interpretation is that the value premium is compensation for risks that the CAPM fails to take into account (e.g., Fama and French 1992, Cochrane 2001). In this view, the book-to-market effect is a proxy for a “distress” factor. Fama and French (1992) suggest that multi-factor models, justified through arbitrage arguments such as Ross's (1976) arbitrage pricing

theory or equilibrium arguments such as Merton's (1973) continuous-time CAPM, may account for these differentials in style returns.

On the other hand, recent research has argued that the documented ability of size and B/M to explain the cross-section of stock returns is not necessarily inconsistent with a single-factor conditional CAPM. In particular, Jagannathan and Wang (1996) use a conditional CAPM framework to show that the beta-premium sensitivity, defined as the slope coefficient from a regression of conditional beta on expected risk premium, affects average returns along with the unconditional beta. In other words, this research argues that, while there is not much difference between the unconditional beta of value and growth, there is evidence that the conditional market beta of value (growth) stocks co-vary positively (negatively) with the expected risk premium (see Petkova and Zhang 2002). Hence, the value premium would be explained by the fact that value stocks are more (less) risky than growth stocks in bad (good) times when marginal utility of consumption is high (low).

A completely different interpretation is that a behavioural bias would explain the value premium as naive investors irrationally undervalue distressed stocks and overvalue growth stocks. In particular, Lakonishok, Shleifer, and Vishny (1994) argue that value strategies yield higher returns because these strategies exploit the less than optimal behaviour of the typical investor, not because they are fundamentally riskier.



Whatever the explanation for the outperformance of value stocks and small-capitalisation stocks, investors may benefit from these empirical regularities by tilting their portfolio towards such high return stocks. Practitioners have been building on the academic findings, and investment strategies based on characteristics such as market capitalisation and investment style (growth, value) are now widely established. Likewise, providers of ETFs cover all types of market segments, and thus allow investors to exploit empirical findings on high return stocks. In our example, we will look at the inclusion of small-cap stocks and value stocks as satellites.

Although the risk premia for small-cap and value stocks have historically been sizeable, these stocks have also underperformed the broader market for prolonged periods. To reap the benefits in times of higher returns, while limiting the risk of underperformance, we include these satellites in the context of a dynamic core-satellite approach. First, we will look at the case of the small-cap satellite and then we will move to the example with the value stocks as a satellite.

2.2.1. Optimal packaging of small-cap exposureIn this first example, we consider a core-satellite portfolio in which the core is made up of the Euro Stoxx 50 index and the satellite of the Euro Stoxx small-cap index. This example allows us to show how to manage the risk of underperformance of small-cap stocks with respect to the large cap Euro Stoxx 50. Since small-cap stocks have historically outperformed over long time periods, periods of underperformance

of small-cap stocks may be characterised as anomalies in the small-cap/large-cap spread.

We look at monthly returns data from January 1994 to December 2007 for the two equity indices. The starting date is limited by the availability of the data for the small- cap index. From figure 1a, it can be seen that small-cap stocks did not outperform their large-cap counterparts over the period we are analysing. A portfolio that is long small caps and short large caps, would be valued below €90 at the end of 2007 for €100 invested at the start of 1994. Looking at the evolution of the value of this long/ short portfolio (i.e., the cumulative small-cap minus large-cap spread), we can see that small caps underperformed sharply up to the year 2000, then kept up with large caps, and finally started to rally in 2002. This rally ended in 2007, when small-cap stocks again underperformed through to December 2007.

Figure 1a: Small-cap minus large-cap spread. Cumulative returns

Our objective is to see how the dynamic risk budgeting process described above fares in such market conditions. The parameters we use are a multiplier of m=5 and a level of the relative guarantee of k=0.95.





Consequently, the initial allocation to the satellite is 25% of the overall portfolio. We also impose a reset rule, in which we lower the parameter k by 2.5% if the weight of the satellite falls below 5%. This enables us to generate new dynamics when margin for action has been decreased too much and allows us to keep a minimum invested in the small-cap index.

Table 1 shows the risk and return statistics for the two indices, and for the dynamic risk budgeting approach. In addition, the table shows the same statistics for a static mix of the core and the satellite that corresponds to the initial weightings of 75% and 25% respectively.

As table 1 shows, the fixed-mix portfolio, which invests 75% in the core and 25% in the satellite each period yields an annualised return of 8.4%. With the dynamic core-satellite technique, the investor obtains an annusalised average return of 10.32%, thus reaping an enviable benefit of nearly 200 basis points a year. It should be highlighted that this outperformance is achieved through a sophisticated packaging of small-cap exposure based on a systematic risk management technique as opposed to a naïve static exposure.

Figure 1b shows the evolution of the allocation to the core and to the satellite in the dynamic core-satellite strategy. Over the first half of the period, the allocation to small caps gradually decreases, as small-cap stocks underperform the large-cap benchmark. Starting in 2000, the dynamic core-satellite strategy increases the allocation to small-cap stocks, as the satellite outperforms and the cushion increases. This increased allocation in times of outperformance, combined with downside protection in times of underperformance, accounts for the spectacular performance of this strategy. Figure 1c indicates the floor (95% of the value of the benchmark) and the cushion (the difference between the portfolio value and the floor) of the strategy.

Figure 1b: Evolution of allocation (small cap)

Table 1: Risk and return statistics for small-cap satellite portfolios

Jan 1994 - Dec 2007Average Return*

Maximum Drawdown

Volatility* Downside

Risk*

Modified Value-

at-Risk***

Sharpe-Ratio*/**

Info-Ratio*


DJ EURO STOXX TM SMALL (Satellite)

8.15% 48.54% 16.51% 11.58% 7.57% 0.37 -0.05

Static CS 8.40% 57.91% 17.32% 12.73% 8.04% 0.37 -0.05

Dynamic CS 10.32% 58.08% 17.93% 12.59% 8.22% 0.46 0.38

* annualised statistics are given ** risk-free rate and MAR are fixed at 2% *** non-annualised 5%-quantiles are estimated




Figure 1c: Core-satellite portfolio evolution (small-cap portfolio)

2.2.2. Optimal packaging of a value tiltWe also analyse the choices of an investor who, for outperformance, adds an ETF of value stocks rather than a small-cap ETF. As mentioned above, the evidence of a value premium in academic finance has led many investors to tilt their portfolios in the direction of high book-to-market stocks or, more generally, towards stocks with low valuation ratios. A straightforward way of accomplishing this value tilt is to add an ETF based on a value index as a satellite portfolio. In the present example, we implement the dynamic core-satellite approach as above (see the first example for a description of the parameters), replacing the small-cap satellite with a value satellite. Note that the period under consideration now starts later (January 1997), since data for the value index are not available before that date.

Figure 2a indicates the cumulative outperformance of the value index over the large-cap index. It can be seen that a period of underperformance is followed by a period of overperformance of the value index.

Figure 2a: Value minus large-cap spread. Cumulative returns

Table 2 again provides risk and return statistics for the core, the satellite, and two competing approaches to core-satellite portfolio management.

From the average returns in the first column of table 2, it can be seen that the value index does not provide much outperformance over the entire period. Consequently, the static core-satellite portfolio adds little performance to the core portfolio (6.08% average annualised returns versus 5.95% for the core). However, adding the satellite

Table 2: Risk and return statistics for value satellite portfolios

Jul 1997 - Dec 2007Average Return*

Maximum Drawdown

Volatility * Downside

Risk*

Modified Value-

at-Risk***

Sharpe-Ratio*/**

Info-Ratio*


DJ EURO STOXX TM Value (Satellite)

6.23% 50.77% 18.59% 15.27% 9.30% 0.23 0.01

Static CS 6.08% 58.27% 19.26% 14.50% 9.31% 0.21 0.01

Dynamic CS 8.22% 57.77% 19.53% 14.43% 9.30% 0.32 0.80

* annualised statistics are given** risk-free rate and MAR are fixed at 2%*** non-annualised 5%-quantiles are estimated



in a risk controlled manner through non-linear risk management yields an annualised average return of 8.22%. Thus, as with the optimal packaging of small-cap exposure, the dynamic core-satellite approach adds an annual value of roughly 200 basis points over the static core-satellite portfolio.

Figures 2b and 2c show the allocation to the satellite over time, as well as the evolution of the value of the dynamic core-satellite portfolio and of the the floor. It can be seen from these figures that the increase in the value allocation is explained by the accumulation of the cushion as the value index outperforms over the later part of the time period.

Figure 2b: Evolution of allocation (value)

Figure 2c: Core-satellite portfolio evolution (value)

From the two examples above, it can be seen that the dynamic packaging of beta exposure makes it possible to generate

outperformance on the order of 2% more per year than naïve static allocation. In the two examples, the dynamic core-satellite technique allows the investor to benefit from the outperformance of small-cap or value stocks, even though that outperformance is not consistent over the entire time period.

2.3. An absolute return perspective: constructing absolute return funds with ETFs

2.3.1. Management techniques for absolute return fundsAbsolute return funds have seen widespread growth in recent years. These funds claim to provide relatively smooth returns with a limited level of risk. Not unlike hedge funds, and rather than stating a benchmark in terms of a market index or a peer group, absolute return funds attempt to obtain a given absolute return level with a target level of volatility, independent of prevailing market conditions.

While such products do not fail to attract the attention of investors, a crucial point is to see how the manager is able to realise the objectives. Traditionally, two broad techniques for securing returns above the risk-free interest rate and with limited volatility have been used.

First, to obtain a broadly diversified portfolio across asset classes that allows significant risk reduction, managers may use techniques inspired by modern portfolio theory. In this strategic asset allocation approach, the correlations between the asset classes under consideration are needed as an input to the optimisation




problem. While low correlation between asset classes or sub-classes effectively allows enhanced risk reduction, it should be noted that these correlations are not known ex ante. Rather, future correlations must be estimated from historical data. This poses two problems.

First, inputting correlations estimated from historical data involves estimation risk: the correlations used as inputs may be different from the true correlations and lead to erroneous portfolio weights. While the problem of estimation risk can be alleviated by imposing some structure on the correlation matrix or by using statistical shrinkage techniques, such methods are not widely used in the industry, as shown by a recent EDHEC survey.1

Second, correlation coefficients may change over time or across states of the economy. In practice, one observes that estimated correlation coefficients change drastically over time. Assets that have been uncorrelated in the past may become highly correlated. Such time dependence can be modelled using multivariate GARCH models (see Engle 2002 for a popular model) Correlations are not only time-dependent but also state-dependent. For example, Longin and Solnik (1995) have shown that the correlation of stock market returns in different countries is not constant and that it increases in a volatile market environment.

A second technique used in absolute return funds is tactical asset allocation, a technique based on the prediction of the returns of different asset classes or sub-classes over short time horizons. The return predictions must thus be

transformed into portfolio holdings that benefit from the scenario in which the return prediction is accurate. In order to be profitable, such strategies thus rely on the capacity of the manager to predict future price movements, either through an econometric model or through qualitative assessment. More importantly, to generate not just high returns, but also the smooth paths associated with absolute return funds, the manager must make correct predictions very consistently.

In practice, these two forms are frequently mixed, with managers performing strategic asset allocation based on historical correlations and making some tactical bets. However, an alternative to such heavy reliance on historical estimates of correlation coefficients and unsure predictions about future returns is to use the dynamic risk management techniques outlined above. The advantage is that these techniques rely only on the observable price paths of asset classes and on a range of predefined parameters, thus circumventing the problems of estimation and prediction uncertainty associated with the two traditional methods.

Below, we show an application of dynamic risk budgeting to the management of an absolute return fund. We also compare this approach to traditional active management, in which the manager has views on the outperformance of an asset class and adjusts the weights accordingly. We will see that to attain the degree of risk control attained with dynamic risk budgeting the traditional form of active management requires great good fortune in predicting returns.


1 - See EDHEC, 2008, European Investment Practices Survey



2.3.2. Designing absolute return funds with ETFsTo illustrate how dynamic risk budgeting may be used in designing absolute return funds with ETFs, we combine a core portfolio that invests in medium-term bonds with a satellite portfolio that invests in an equity ETF. The objective of the proposed strategy is to achieve smooth returns with the low volatility of the core portfolio. An additional objective is to benefit from the returns on the stock market ETF if stocks outperform bonds, while achieving protection from the downside risk of the equity investment.

The details of the strategy under consideration are as follows: the core is made up of the EuroMTS for bonds with three to five years to maturity and the satellite invests in the EuroStoxx 50. The data used consist of monthly returns including coupon or dividend payments for the period from January 1999 to December 2007. The starting period is different from the applications above because the bond data are available starting only with the introduction of the euro, as is typical for data on euro-denominated bonds. The value of the multiplier is set at six, while the floor is defined at 90% of the value of the bond core portfolio. The weight of the satellite is to be capped at 60% of the overall portfolio.

Furthermore, we use an extension to the basic dynamic core-satellite approach to achieve the absolute returns objective. In particular, we introduce a maximum drawdown limit equal to 10% to take into account the investor’s aversion to drawdowns in the absolute value of the overall portfolio. In fact, there are suitable

extensions of the standard dynamic asset allocation strategies that can accommodate the presence of maximum drawdown constraints. These strategies have been introduced by Estep and Kritzman (1988), who call them time invariant portfolio protection strategies (TIPP), and later formalised by Grossman and Zhou (1993) and Cvitanic and Karatzas (1995). By imposing a maximum drawdown constraint of 10% and using our multiplier of six, we note that the maximum allocation to the satellite portfolio is 60%. Hence, the defensive bond portfolio used as the core will always constitute at least 40% of the overall portfolio.

Figure 3a shows the cumulative returns of the strategy we implemented, as well as of the core and the satellite portfolio. In addition, to highlight the built-in protection of this investment strategy, the level of the floor is displayed as well.

Figure 3a: Absolute return fund: evolution of the core, the satellite, and the DCS portfolios

From this figure, a number of conclusions can be drawn. The dynamics of the core portfolio confirm the conservative character of the core investment. However, we also see that performance of the bond core was quite flat over the two most recent years of the period. For the satellite portfolio, we observe that returns are higher if we look at the entire period. More importantly, fluctuations




in the value of the satellite portfolio are tremendous, with a sharp increase in value up to the year 2000 and a radical decline from then until 2003, followed again by a steady increase until the end of 2007. The dynamic core-satellite (DCS) portfolio combines the advantages of each of its ingredients: the smooth performance of the bond core and the upside potential of the equity satellite. As a result, performance is smooth over the entire period, and cumulative returns at the end of the period are actually higher than those of the satellite. It is also interesting to look at the dynamics of the floor. As the value of the dynamic core-satellite fund increases, the floor is pulled up in order to increase the level of protection. It is also instructive to look at the performance in the stock market downturn beginning in the year 2000. In fact, the dynamic core-satellite portfolio is little affected. As the portfolio value approaches the floor, allocation shifts to the core. This behaviour is illustrated in figure 3b, which shows the weights held in the core and the satellite over time.

Figure 3b: Absolute return fund: evolution of the allocation between the core and the satellite portfolios

It is important to recall the objective of the strategy analysed here. The conservative nature of the core and the dynamic risk management process both aim to achieve smooth returns over time. Figure 3c shows the return obtained over rolling periods of one year. From this figure, we see that the dynamic core-satellite portfolio achieves positive returns over most rolling windows of one year. Even in the period after the year 2000, returns are close to zero, though slightly negative. In fact, the maximum loss over a one-year rolling period is -6.68%. This is in stark contrast to the satellite, which displays returns below -20% over a range of annual periods. In fact, the behaviour of the dynamic core-satellite portfolio is similar to that of the defensive bond portfolio that makes up the core.

Figure 3c: Absolute return fund: performance of the core, the satellite and the DCS over a one-year rolling period

Risk and return statistics for the dynamic core-satellite strategy confirm the conclusions from the figures above. In particular, table 3 shows that the average return exceeds those of the core by roughly 337 basis points, while maintaining low levels of risk.


Table 3: Absolute return fund: risk and return statistics for the core, the satellite, and the DCS

* annualised statistics are given ** non-annualised 5%-quantiles are estimated *** risk-free rate and MAR are fixed at 2%

Average Return*Maximum Drawdown

Volatility* Modified

Value-at-Risk**CVaR** Sharpe-Ratio***

Core Satellite

4.00%5.16%

-3.08%-59.90%

2.40%18.41%

0.83%8.56%

1.15%12.90%

0.830.17

DCS - base case 7.37% -9.01% 6.42% 1.97% 2.57% 0.84



2.3.3 A comparison with active management based on return forecasts. We have seen that the dynamic risk budgeting approach is able to provide sound absolute return management. The remarkable trait of the approach is the absence of any prediction. The systematic allocation based on past values of the core and satellite portfolios means that the investor need not bear any forecasting risk.

An alternative to this approach is to forecast the relative returns of the satellite. If the satellite is expected to have higher returns than the core, the weight of the satellite should be increased; if not, it should be decreased. Predictions may be based either on an econometric process or on qualitative assessment by the manager or an outside expert. Of course, the performance of this investment process will depend on the accuracy of predictions. If the forecast is right most of the time, we expect the portfolio to display attractive performance. In this section, we first assess the performance displayed by a manager with positive prediction skills and then analyse how the results change if we increase this skill.

The setup of the analysis is detailed below. We simulate an active manager’s approach with the following assumptions:• If the manager thinks that the satellite will outperform the core on the following month, he will allocate 60% of his portfolio to the satellite. This weight corresponds to the maximum weight allowed in the dynamic risk budgeting process above. The remaining 40% is allocated to the satellite.• If the manager thinks that the core will outperform the satellite on the following month, he will allocate 100% of the

portfolio to the core• The manager rebalances his holdings on a monthly basis and his predictions are correct 7 times out of 12 on average

Using the same time period as above (January 1999 to December 2007) and the same core and satellite, we simulate 1,000 scenarios. Each scenario corresponds to a time series of returns for the active manager, given his bets. As the basis for the simulation, we use a hit ratio of 58.3%, i.e., the average active manager put his money on the right horse, as it were, 7 months out of 12. Thus the 1,000 scenarios represent the returns obtained by 1,000 hypothetical active managers who have a hit ratio of 58.3%.

Table 4 displays risk and return statistics for the average over these 1,000 scenarios and compares them to the base case strategy from above, i.e., the dynamic core-satellite strategy. The average for all scenarios corresponds to an equal-weighted portfolio of the 1,000 active managers. The first two lines of table 4 reproduce the statistics for the core and the satellite portfolio.

It is instructive to compare the time series of the actively managed portfolio to the series of the dynamic core-satellite strategy. The average annualised return of the dynamic core satellite (7.37%) is slightly higher than the average return for this portfolio (7.25%). Moreover, the maximum drawdown, volatility, VaR, and CVaR are significantly lower for the dynamic core satellite. The higher risk statistics for the portfolio of active managers show the impact of bad predictions. In fact, even though these managers are right most of




the time, they err five months per year, thus exposing the investor to significant downside risk.

It should be noted that this result holds for the equal-weighted portfolio that is diversified across 1,000 managers. Using a single manager with the same ability leads to higher uncertainty, as results may be much better or much worse for two reasons. First, the results obtained by a single manager depend on the actual hit ratio displayed over the sample period as opposed to his true long-term hit ratio. Second, given a realised hit ratio, portfolio performance depends on the consequences of his predictions. Predicting outperformance over a month in which the satellite underperforms by 1% is not the same as predicting outperformance over a month in which the satellite underperforms by 10%, even though both are instances of a forecast error. Likewise, predicting outperformance over a month where the satellite outperforms by 10% is more valuable than predicting outperformance over a month where the setellite outperforms by a mere 1%, though both are instances of a correct forecast.

The gap between managers with the same forecasting ability is shown in the lower line of table 4. The worst performing manager (or scenario) achieves average annualised returns of 0.64% while the

best achieves 13.35%. Likewise, maximum drawdown and other risk measures vary widely from one manager (or scenario) to another.

Overall, the results show that an actively managed portfolio based on largely accurate forecasts leads to results less attractive than those generated by the dynamic core-satellite process. Moreover, choosing a particular manager leads to additional risk. In fact, a particular manager may produce poor results despite his forecasting ability.

A closer look at the parameter for forecasting ability would not lack interest, but it is not our objective to make a statement about realistic degrees of forecasting accuracy or how correct forecasts can be obtained. Rather, it is interesting to assess the results assuming even higher hit ratios. Above, we have assumed a hit ratio of 58.3%. Below, we repeat the simulation using average hit ratios of 7/12, 8/12, 9/12, 10/12, and 11/12.

We simulate the same active management approach, with increasing hit ratios and observe the hit ratios necessary to achieve the same maximum level of risk (maximum drawdown and worse performance over a rolling one year period) and the same probability of losing more than 10% of the capital over a one-year period. Table 5 shows the results.


Table 4: Risk and return statistics for the core, the satellite and the active management scenarios.

* annualised statistics are given ** non-annualised 5%-quantiles are estimated *** risk free rate and MAR are fixed at 2%

Average Return*Maximum Drawdown

Volatility* Modified

Value-at-Risk**CVaR** Sharpe-Ratio***

Core Satellite

4.00%5.16%

-3.08%-59.90%

2.40%18.41%

0.83%8.56%

1.15%12.90%

0.830.17

DCS - base case 7.37% -9.01% 6.42% 1.97% 2.57% 0.84

Average of 1,000 scenarios 7.25% -14.72% 7.60% 2.84% 5.36% 0.70

Worst performing scenarioBest performing scenario

0.64%13.35%

-29.61%-2.81%

8.12%6.25%

4.16%0.76%

6.44%1.53%

-0.171.81




The results show that a hit ratio of seven to twelve is necessary to generate results that are equivalent to the dynamic core-satellite approach in terms of average returns. To obtain the same level of maximum drawdown as the dynamic core-satellite approach, a hit ratio of nine to twelve is necessary. If we look at the probability of losing more than 10% of capital over a one-year period, a hit ratio of eleven to twelve is necessary to match the dynamic core-satellite approach. It should be noted that the results in table 5 are for the equal-weighted portfolio of 1,000 hypothetical managers. The additional risk of manager selection faced by an investor who uses such an approach is thus ignored.

Overall, the results suggest that dynamic risk budgeting techniques make for a suitable approach to managing absolute return funds. In particular, this approach makes it possible to avoid the dependence on forecast accuracy of conventional tactical active management. Therefore, to generate absolute returns dynamic risk budgeting techniques are a more reliable alternative even if one has managers with excellent forecasting ability.

3. New Risk Budgeting Techniques: Conclusion and Outlook

We have argued that the core-satellite approach can be extended to dynamic investment, allowing investors to protect their portfolio from excessive loss. Core-satellite management also turns out to be the optimal investment choice in a benchmarked portfolio context, justified by optimal portfolio selection theory, as demonstrated in the mathematical appendix. Introducing a general portfolio model in the presence of benchmark-related objectives, we show that the standard core-satellite portfolio approach to asset management is rationalised as a special case of a fund separation theorem. It is also argued that the presence of implicit or explicit constraints on underperformance of the portfolio with respect to the benchmark induces a dynamic extension of the standard core-satellite approach.

Applications of dynamic core-satellite portfolio management show that it allows efficient management of portfolios of ETFs whose aim is to benefit from the small-cap or value premium. In addition, it is shown that ETFs on stock and bond indices can be used to construct absolute return funds based on this dynamic allocation approach.

Table 5: Evolution of the risk and return statistics as a function of the hit ratio of the active manager

DCS 7/12 8/12 9/12 10/12 11/12

Average return 7.37% 7.25% 10.02% 12.85% 15.75% 18.68%

Max DD -9.01% -14.72% -11.70% -9.18% -7.08% -4.71%

Worst performance over a rolling one-year period -6.68% -30.5% -30.5% -24.7% -19.0% -11.4%

Probability of losing more than 10% over a rolling one-year period

0.00% 3.02% 1.29% 0.38% 0.11% 0.00%

Conclusion




Our main objective is to shed light on current perceptions of the qualities of ETFs as compared to those of other indexing vehicles. In addition, we show how ETFs for different asset classes are actually used. First, we will summarise below some of the key findings of the survey of European investors and asset managers. Second, since the present survey is an update of the EDHEC European ETF Survey 2006, we will compare some of the key findings of the two surveys. In addition to reporting the survey results, we outline several possible applications of ETFs in a state-of-the-art portfolio process based on dynamic risk management techniques. A third objective of this conclusion is to provide a summary of these applications and explain how they relate to the responses elicited by our questionnaire.

Overall, analysis of responses from 111 European investors and asset managers shows that, more often than not, ETFs are the preferred vehicle for the implementation of indexing strategies. The most serious challengers to ETFs are futures instruments. Respondents attribute good qualities to futures, especially in the areas of cost, transparency, and liquidity. ETFs are also rated highly on these criteria and are preferred to futures when it comes to minimum subscription requirements, regulatory requirements, and operational constraints. Another point that respondents appreciate with ETFs is the considerable range of products and asset classes available with these instruments.

Despite these advantages, however, most ETF use is concentrated around equity indices. In fact, more than three-quarters of respondents use ETFs in the equity arena; only around 40% do so for bond investing.

In real estate and hedge fund investing, approximately one-third of respondents use ETFs or ETF-like products; for commodity investing, the figure is closer to one-half. As will be detailed below, the use of ETFs has grown tremendously across all these asset classes. ETF use is centred on the equity universe. It is also centred on a specific form of equity indices—broad market indices—especially when it comes to respondents’ core portfolios. In fact, 94% of respondents use ETFs based on broad market indices, while only 31% use sector ETFs and only 19% use style ETFs for the construction of core portfolios. We conclude that European investors and asset managers do not use customised allocation to finer sub-segments of the equity markets to exploit to the full the potential to improve the efficiency of their core portfolios. Instead, they prefer to hold a broad index ETF. This conclusion carries over to other asset classes, in particular government bonds. For example, 84% of respondents use broad market bond ETFs in their core portfolios, while only 33% use ETFs that reflect the returns of a given maturity segment. As detailed coverage by maturity segment makes it possible to control the interest rate risk exposure of the core portfolio and to customise it to the investor’s precise objectives—the broad bond market ETF simply reflects the average duration of the bond market—investor preferences as revealed by our survey may provoke no little surprise.

ETFs stand out from other indexing instruments not only for the wide range of products on offer but also for the number of possibilities for managing the risk and return of a portfolio of ETFs. These possibilities include trading in options directly written on an ETF, using inverse-performance ETFs

Conclusion



or shorting the ETF itself, and lending out ETF units to generate income from the short interest obtained. We find that, with the exception of inverse-performance ETFS, these possibilities are used by only a very small fraction of European investors and asset managers. Broadly, possibilities such as options on ETFs, short selling, and securities lending are currently used by approximately 5% of investors.

When comparing these results to our survey from 2006, one can see that the perception of the comparative advantage of ETFs has remained similar but that the use of ETFs has been growing across all asset classes. Table 1 provides a comparison of the key results of the two surveys.

As table 1 shows, ETF use in the equity universe has increased from 45% to 78%. For the other asset classes, ETFs are used by 30% to 50% of respondents to our 2008 survey; the figures for 2006 were 5% to 15%.

Satisfaction with ETFs has remained at high levels or increased slightly for equity and bond ETFs. For ETFs or ETF-like products on alternative asset classes, satisfaction rates have advanced tremendously.

Overall, as the examples we discuss show, the trends noted by the previous survey are maintained or have become more pronounced. Core-satellite investing has become increasingly popular, as the percentage of current and potential users is now 65%, as opposed to 57% in the 2006 survey. More than 50% of respondents use ETFs for both strategic and tactical purposes; less than 30% of respondents reported doing so in the earlier survey.

When it comes to comparing competing indexing instruments and ETFs, we observe that the percentage of non-responses has decreased considerably, suggesting greater familiarity with instruments such as ETFs, futures, and total return swaps. However, the views on the relative merits of each

instrument have not changed much since the previous survey. Table 2 indicates how respondents rate ETFs, futures, total return swaps, and traditional index funds on liquidity, cost, and reliability of tracking error.

Conclusion

Equity Govt. Bonds Corp. Bonds Commodities Real Estate Hedge Funds

Percentage of respondents using ETFs

2006 Survey 45% 13% 6% 15% 6% 7%

2008 Survey 78% 42% 40% 48% 35% 30%

If you use ETFs or ETF-like products, are you satisfied with them?

2006 Survey 92% 80% 58% 65% 50% 27%

2008 Survey 92% 85% 66% 87% 77% 58%

Table 1: ETF use 2008 vs. 2006: percentage of users and satisfaction



The increased awareness of indexing vehicles is seemingly accompanied by a desire to limit ETFs to this field, as opposed to seeing them venture into active management. 77% of respondents to the latest survey say they prefer passive ETFs to active ETFs, compared to 64% in the earlier study. While advanced uses of ETFs have not increased

overall, securities lending seems to have increased, as 12% of respondents now state that they are current or potential users, as opposed to 6% in the earlier survey. 2008 findings on the use of advanced features of ETFs are shown in table 3 and compared to the findings of the 2006 survey.

Conclusion

Rating with regard to the level of liquidity

2006 2008

very good

fairly good

poorno

answer

very good

fairly good

poorno

answer

ETFs 28 46 6 21 32 50 9 10

Futures 59 10 0 31 65 16 0 19

Total return swaps 8 32 18 42 17 31 17 35

Index funds 18 33 15 34 24 39 15 22

Rating with regard to cost

2006 2008

very good

fairly good

poorno

answer

very good

fairly good

poorno

answer

ETFs 21 40 13 26 30 49 11 11

Futures 46 14 4 37 62 14 5 20


Index funds 15 29 18 38 17 36 24 23

Rating with regard to reliability of tracking error

2006 2008

very good

fairly good

poorno

answer

very good

fairly good

poorno

answer

ETFs 34 32 5 29 33 49 5 14

Futures 32 22 5 40 51 23 4 22


Index funds 26 28 5 41 23 44 7 25

Table 2: Ratings 2008 vs. 2006: percentage of respondents who rate the instrument in the corresponding category

Table 3: Use of advanced ETF features 2008 vs. 2006: percentage of respondents answering positively to the question “Do you use the following features of ETFs?”

2006 2008 2006 2008

Answer given⇒“Yes” “No, but soon” “Yes” “No, but soon”

“Yes” or “soon”

“Yes” or “soon”Feature⇓

Trading options on ETFs 4 6 4 5 10 9

Shorting ETFs 4 13 5 8 17 13

Lending ETFs 5 1 7 5 6 12



On the other hand, shorting ETFs has become less popular. 13% of respondents to the 2008 survey report that they are current or potential users of short-selling, while 17% did so in the previous survey. However, this decrease may be explained by the emergence of inverse-performance ETFs, of which more than 30% of respondents report that they are current or potential users. As these results show, inverse-performance ETFs are potentially successful precisely because European investors and asset managers are reluctant to short ETFs themselves. Instead, they prefer to use inverse-performance ETFs as a substitute for short selling.

The percentage of respondents who report that they would like to see new products in areas such as emerging markets and alternative investment classes has not decreased from the previous survey despite the large number of new products that have since appeared. The market may thus expect continued product development growth despite the wide range of products currently on offer.

38% of respondents say that optimal portfolio construction or risk management will account for the greatest increase in their future use of ETFs. The second part of this study illustrates several optimal portfolio construction and risk management applications. We describe, for example, a dynamic risk management process that makes it possible to gain access to the upside potential of a satellite ETF while limiting the risk of underperforming a core portfolio of ETFs. Several applications of this dynamic core-satellite portfolio management show the benefits of dynamically adjusting the weights of the core and the satellite. ETFs are a natural vehicle for implementing these

dynamic asset allocation strategies since they are easily traded and make it possible to choose very precisely defined segments of asset markets, such as equity styles. We show that the inclusion of ETFs on small-cap or value stocks in the dynamic portfolio process captures outperformance while limiting the downside in periods of underperformance. Moreover, dynamic risk management makes it possible to form absolute return funds using ETFs on such traditional asset classes as stocks and bonds.

The results of our survey convey a clear message: ETFs are now widely used and practitioners are highly satisfied with their features. However, the use of ETFs is largely limited to passive holdings of broad market indices. The wide range of ETFs for subcategories and styles is not used to its full potential. Likewise, most practitioners do not benefit from the possibilities of trading options on ETFs, selling ETFs short, or lending them out. ETFs undeniably provide value when it comes to passive exposure to a traditional or alternative asset class. However, we believe that there is considerable value-added in making use of an important feature of ETFs, their ability to be bought and sold like stocks. They are thus ideally suited to dynamic risk management in portfolio construction. The last part of our study shows that such dynamic risk budgeting has substantial benefits. While the examples provided there are not meant to be complete solutions, it is our hope that they spur further reflection on the future use of ETFs.

Conclusion

Appendix




In this appendix, we introduce a general portfolio selection model in the presence of benchmark-related objectives. In addition to the less formal description of the core-satellite approach and its benefits in part I, section 2, of this document, we show that the standard core-portfolio approach to asset management is rationalised as a special case of a fund separation theorem. We also argue that the presence of implicit or explicit constraints on under-performance of the portfolio with respect to the benchmark induces a dynamic extension of the standard core-satellite approach, which was also presented in part I, section 2.

A.1. A Stochastic Model for the Asset and Benchmark ValuesLet [0,T] denote the (finite) time span of the economy, where uncertainty is described through a standard probability space (Ω,A,P) and endowed with a filtration Ft ; t ≥0{ } , where F∞ ⊂ A and F0 is trivial, representing the P-augmentation of the filtration generated by the n-dimensional Brownian motion

W1,...,W n( ) .

We consider n risky assets (or asset classes), the prices of which are given by:

dPt

i =Pti μi dt + σij dWt

j

j=1

n

∑⎛

⎝⎜⎞

⎠⎟, i =1,...,n

We shall sometimes use the shorthand vector notation for the expected return (column) vector

μ = μi( )i=1 ,....,n

' and matrix

notation σ = σij( )

i , j=1 ,....,n for the asset return

variance-covariance matrix, assumed to be non-singular. We also denote by 1=(1,…,1)’ an n-dimensional vector of ones and by

W = W j( ) '

j=1 ,....,nthe vector of Brownian

motions. A risk-free asset, the 0th asset, is

also traded in the economy. The return on that asset, typically a default free bond, is given by dPt

0 =Pt0rdt , where r is the

risk-free rate in the economy.

We also introduce a separate process that represents the dynamics of some exogenously given benchmark:

dBt =Bt μB dt + σB , j

j=1

n

∑ dWtj +σB ,εdWt

ε⎛

⎝⎜⎞

⎠⎟

where

Wtε( ) is a standard Brownian motion,

uncorrelated with W, that can be regarded as the projection residual of benchmark risk onto asset price risk and represents the source of uncertainty that is specific to benchmark risk, which cannot be replicated by existing securities. μB represents the expected return on the benchmark portfolio;

σB ,ε represents pure benchmark volatility, while the vector

σB = σB , j( )

j=1 ,....,n

'

can be regarded as measuring the factor loadings of benchmark risk onto asset price risks. In the complete market case, when the benchmark can be fully replicated by existing assets,

σB ,ε = 0. In general, however, the benchmark may not be replicable by existing assets, e.g., in the case of a broad benchmark such as the MSCI world index. In this case, the correlation between the benchmark and the benchmark-hedging portfolio (i.e., the portfolio with the highest correlation with the benchmark) is strictly lower than one.

A.2. Objective and Investment PolicyThe investment policy is a (column) predi-ctable process vector

wt

' = w1t ,...,wnt( )( )t≥0

that represents allocations to risky assets, with the reminder invested in the risk-free asset. We define by At

w the asset process,

Appendix:The Theory of Core-Satellite Investing



i.e., the wealth at time t of an investor following the strategy w starting with an initial wealth A0 .

We have that:

In a benchmarked portfolio management context, what matters is not the value of the assets per se, but how the asset value compares to the benchmark value. We capture this element by assuming that the investor’s objective is written in terms of relative wealth (relative to the benchmark), as opposed to absolute wealth:

max

wE0 U At

w Bt( )⎡⎣ ⎤⎦ . Hence, the “funding” ratio of assets with respect to the benchmark, denoted by Ft = At

w Bt , appears as a key state variable in this model.

Using Itô’s lemma, we can also derive the stochastic process followed by the funding ratio under the assumption of a strategy w:

which yields:

A.3. Solution Using the Dynamic Programming ApproachDefine the indirect or derived utility process at time t:

Jt =max

wEt U FT( )⎡⎣ ⎤⎦

where Et •[ ] denotes the expectation conditional on information available at time t, such as described by the filtration generated by the n Brownian motion driven asset prices and the (n+1)th Brownian motion driving pure liability uncertainty.

A.3.1. General solutionFor a Markovian control process

wt( )

t≥0 and

a function ϕ t ,Ft( ) ∈C1 ,2 the infinitesimal generator of the funding ratio process is:

Awϕ t ,Ft( ) = ϕt +FϕF μF

w + 12

F 2ϕFF σFw( ) 2

,where the derivative of a function f with respect to variable x is denoted as fx.

Given the objective function, the appropriate Hamilton-Jacobi-Bellman equation associated with this problem is:

sup

w

AwJ t ,Ft( ){ } = 0 ,

subject to J T,FT( ) =U Ft( ) .

Optimising with respect to w yields:

FϕF

∂μFw

∂ww*( ) + 1

2F 2ϕFF

∂ σFw( ) 2

∂ww*( ) = 0

or

FϕF μ − r1( ) −σσB( ) +F 2ϕFF w*'σσ '−σσB( )( ) = 0

with solution:

w* =w* t ,Ft( ) = − σσ '( ) −1

μ − r1( )ϕF t ,Ft( )

FϕFF t ,Ft( )+ 1+

ϕF t ,Ft( )FϕFF t ,Ft( )

⎛

⎝⎜

⎞

⎠⎟ σ '( ) −1

σB

w* =w* t ,Ft( ) = − σσ '( ) −1μ − r1( )

ϕF t ,Ft( )FϕFF t ,Ft( )

+ 1+ϕF t ,Ft( )

FϕFF t ,Ft( )⎛

⎝⎜

⎞

⎠⎟ σ '( ) −1

σB


dAt

w = Atw 1−w' .1( ) dBt

Bt

+w'dPt

Pt

⎡

⎣⎢

⎤

⎦⎥ = At

w r +w' μ − r1( )( )dt +w' σdWt⎡⎣ ⎤⎦

dFt =d

Atw

Lt

⎛

⎝⎜⎞

⎠⎟=

1Lt

dAtw −

Atw

Lt2

dLt −1

Lt2

dAtwdLt +

Atw

Lt3

dLt( ) 2

dFtw

Ftw

=

dAt

w

Bt

⎛

⎝⎜⎞

⎠⎟

Atw

Bt

= r − μB +σB' σB +σB ,ε

2( )dt +w' μ − r1( ) −σσB( )dt + w' σ −σB'( )dWt −σB ,εdWt

ε

dFtw

Ftw

=

dAt

w

Bt

⎛

⎝⎜⎞

⎠⎟

Atw

Bt

= r − μB +σB' σB +σB ,ε

2( )dt +w' μ − r1( ) −σσB( )dt + w' σ −σB'( )dWt −σB ,εdWt

ε



We thus obtain a three-fund separation theorem, in which the optimal portfolio strategy consists of holding two funds, one with weights

wM =

σσ '( ) −1μ − r1( )

1' σσ '( ) −1μ − r1( )

and another one with weights

wB =

σ '( ) −1σB

1' σ '( ) −1σB ,

the rest being invested in the risk-free asset.

The first portfolio is the standard performance-seeking portfolio. Note that the amount invested in that portfolio is directly proportional to the investor’s Arrow-Pratt coefficient of risk-tolerance

−

ϕF

FϕFF

(the inverse of the relative risk aversion). This makes sense: the higher the investor’s (relative) risk tolerance, the higher the allocation to that portfolio will be.

In order to better understand the nature of the second portfolio, it is useful to remark that it is a portfolio that minimises the local volatility σF

w of the funding ratio. To see this, recall that the expression for the local variance is given by

σF

w = w' σ −σB'( ) '

w' σ −σB'( ) +σB ,ε

2( )1

2

,

which reaches a minimum for

w* = σ '( ) −1

σB , with the minimum being

σB ,ε2 . As such, it is the equivalent of the

minimum variance portfolio in a relative return-relative risk space, also the equivalent of the risk-free asset in a complete market

situation in which benchmark risk is entirely spanned by existing securities ( σB ,ε = 0 ). Alternatively, this portfolio can be shown to have the highest correlation with the benchmark. As such, it can be called a benchmark-hedging portfolio, in the spirit of Merton (1971) intertemporal hedging demands.1

A.3.2. Specific solution in case of CRRA utility and constant parameter valuesLet us now consider a specific utility function of the CRRA type:

U FT( ) = FT( )1−γ

1−γ

We try a solution to the non-linear Cauchy problem:

ϕt +FϕF μF

w *

+ 12

F 2ϕFF σFw *( ) 2

= 0 ,

which is separable in F and can be written as:

ϕ t ,Ft( ) = g t ,T( ) Ft( )1−γ

1−γ , with:

g t ,T( ) = exp T − t( ) −1

21 −

1

γ

⎛

⎝⎜⎞

⎠⎟θ − σB( )' θ − σB( ) − γ σB ,ε

2 −θBσB ,ε( ) + 1 − γ( ) 2 − γ( )2

σB ,ε2

⎛

⎝⎜⎞

⎠⎟⎡

⎣⎢⎢

⎤

⎦⎥⎥

Here σ −σB is defined as the matrix whose general term is equal to that of σ outside the diagonal and is equal to σii −σB , also written as σi

2 −σB , on the diagonal.

Given that −

JF

FJFF

=1γ

, we finally obtain:

w* =w* t ,Ft( ) = 1

γσσ '( ) −1

μ − r1( ) + 1−1γ

⎛

⎝⎜⎞

⎠⎟σ '( ) −1

σB

As is well known, it should be noted that when γ =1 , i.e., in the case of the log


1 - When the benchmark is liability driven, the benchmark-hedging portfolio is then called a liability hedging portfolio, and can be implemented with cash instruments (nominal and real bonds) and/or with derivatives instruments (interest and inflation swaps).

g t ,T( ) = exp T − t( ) −1

21 −

1

γ

⎛

⎝⎜⎞

⎠⎟θ − σB( )' θ − σB( ) − γ σB ,ε

2 −θBσB ,ε( ) + 1 − γ( ) 2 − γ( )2

σB ,ε2

⎛

⎝⎜⎞

⎠⎟⎡

⎣⎢⎢

⎤

⎦⎥⎥

situation in which benchmark risk is entirely spanned by existing securities ( ). Alternatively, this portfolio can be shown to have the highest correlation with the benchmark. As such, it can be called a benchmark-hedging portfolio, in the spirit of Merton (1971) intertemporal hedging demands.1

A.3.2. Specific solution in case of CRRA utility and constant parameter valuesLet us now consider a specific utility function of the CRRA type:

We try a solution to the non-linear Cauchy problem:

,

which is separable in F and can be written as:

, with:

Here is defined as the matrix whose general term is equal to that of outside the diagonal and is equal to , also written as , on the diagonal.

Given that , we finally obtain:

As is well known, it should be noted that when , i.e., in the case of the log



investor, the intertemporal hedging demand is zero (myopic investor). In general, again, the optimal strategy consists of holding two funds, in addition to the risk-free asset, the standard mean-variance portfolio and the benchmark hedging portfolio, and the proportions invested in these two funds are constant in time.

Interestingly, this is strongly reminiscent of what is known as the core-satellite approach to benchmarked asset management. In practice, the benchmark hedging portfolio is also known as the core portfolio, while the performance-seeking portfolio is typically called the satellite portfolio.

In other words, we have shown that a fraction of the assets should optimally be allocated to relative risk management and invested in the portfolio (known as the benchmark-hedging portfolio, or core portfolio) that exhibits the highest correlation with respect to the benchmark, while another fraction of the asset is allocated to performance generation and invested in the portfolio (known as the satellite portfolio) that exhibits the optimal risk-return trade-off, irrespective of the benchmark.

This approach stands in sharp contrast to more traditional benchmarked portfolio management methods, where both objectives (benchmark risk management and performance generation) are pursued simultaneously by an active manager subject to tracking error constraints in an attempt to achieve the portfolio with the highest possible information ratio. Also, what the formal analysis of benchmarked asset management suggests is that the optimal strategy should involve two components, the core portfolio, dedicated to risk management with respect

to the benchmark, and the satellite portfolio, dedicated to outperformance. On the other hand, the analysis does not imply that the satellite should necessarily involve active portfolio strategies.

A.4. From Static to Dynamic Portfolio ManagementIt can be desirable from an investor’s standpoint to set a strict constraint on the potential underperformance of the portfolio with respect to the benchmark. There can be two types of constraints, explicit or implicit. In a program with explicit constraints, marginal indirect utility from wealth discontinuously jumps to infinity:

Maxws ,t≤s≤T

Et

AT

BT

⎛⎝

⎞⎠

1−γ

1−γ

⎡

⎣

⎢⎢⎢⎢

⎤

⎦

⎥⎥⎥⎥

such that AT ≥kBT almost surely. On the other hand, in a program with implicit constraints, marginal utility goes smoothly to infinity:

Maxws ,t≤s≤T

Et

AT

BT

−k⎛⎝

⎞⎠

1−γ

1−γ

⎡

⎣

⎢⎢⎢⎢

⎤

⎦

⎥⎥⎥⎥

It can be shown that the solution to the programme with implicit constraints yields the following time-dependent solution:2

w * =w * Ft( ) = 1

γ1 −

k

Ft

⎛

⎝⎜⎞

⎠⎟σσ '( ) −1

μ − r1( ) + 1 −1

γ1 −

k

Ft

⎛

⎝⎜⎞

⎠⎟⎛

⎝⎜

⎞

⎠⎟ σ '( ) −1

σB

w * =w * Ft( ) = 1

γ1 −

k

Ft

⎛

⎝⎜⎞

⎠⎟σσ '( ) −1

μ − r1( ) + 1 −1

γ1 −

k

Ft

⎛

⎝⎜⎞

⎠⎟⎛

⎝⎜

⎞

⎠⎟ σ '( ) −1

σB

Consider the fraction of wealth At allocated


2 - See Martellini (2006), who in the context of a liability-driven benchmark uses the convex duality (or martingale) technique for solving the optimal portfolio allocation problem.




to the performance-seeking, or satellite, portfolio. It is given by

Hence, it appears that the fraction of wealth allocated to the satellite is equal to a constant multiple m of the cushion, i.e., the difference between the asset value and the floor defined as At - kLt. This strategy, which was first introduced by Amenc, Martellini, and Malaise (2004) under the name of dynamic core-satellite investment, is reminiscent of constant proportion portfolio insurance (CPPI) strategies, which it extends to a relative risk management context. While CPPI strategies are designed to prevent final terminal wealth from falling below a specific threshold, extended CPPI strategies (or dynamic core-satellite strategies) are designed to keep asset value from falling below a pre-specified fraction of some benchmark value.

1' σσ '( ) −1μ − r1( )

γAt −

At kFt

⎛

⎝⎜⎞

⎠⎟=

1' σσ '( ) −1μ − r1( )

γAt −kBt( )

References




• Acharya, V., and L. Pedersen, 2005, “Asset pricing with liquidity risk”, Journal of Financial Economics, vol.77, p. 375-410.

• Amenc, N., P. Malaise, and L. Martellini, 2004, “The benefits of bond ETFs for institutional investors. The natural vehicle for a core-satellite approach”, working paper, EDHEC.

• —, 2004, “Revisiting core-satellite investing—A dynamic model of relative risk management“, Journal of Portfolio Management, vol. 31, n°1, fall, p. 64-75.

• Black, F., and R. Jones, 1987, "Simplifying portfolio insurance", Journal of Portfolio Management, vol. 14, fall, p. 48-51.

• Black, F., and A. Perold, 1992, Theory of constant proportion portfolio insurance, Journal of Economic Dynamics and Control, vol. 16, p. 403-426.

• Cvitanic, J., and I. Karatzas, 1995, "On portfolio optimization under drawdown constraints". IMA Volumes in Mathematics and its Applications 65, p. 35-46.

• Cochrane, J., 2001, Asset Pricing, Princeton University Press, Princeton, New Jersey.

• Comer, G., E. Elton, M. Gruber, and K. Li., 2002, "Spiders: Where are the bugs?" Journal of Business, vol. 75, n°3.

• Demaine, J., 2002, “Exchange-traded funds for the sophisticated investor”, Derivatives Use, Trading & Regulation, vol. 7, n°4, p. 354-361.

• Engle, Robert, 2002, "Dynamic conditional correlation—A simple class of multivariate GARCH models", Journal of Business and Economic Statistics, p. 20(3).

• Engle, R., and Sarkar, D., 2006, “Premiums-discounts and exchange traded funds”, Journal of Derivatives, vol. 13, n°4, p. 27-45.

• Estep, T., and M. Kritzman, 1988, "TIPP: insurance without complexity", Journal of Portfolio Management, vol.14, summer, p.38-42.

• Fama, E., and K. French, 1992, The cross-section of expected stock returns, Journal of Finance, p. 442-465.

• Gastineau, G., 2001, “Exchange Traded Funds: An Introduction”, Journal of Portfolio Management, vol. 27, spring, p. 88-96.

• Goltz, F., L. Martellini, and M. Vaissié, 2007, “Hedge Fund Indices: Reconciling Investability and Representativity”, European Financial Management, vol.13, p. 257–286.

• Grinold and Kahn, 2000, Active Portfolio Management: Quantitative Theory and Applications, McGraw-Hill.

• Grossman, S. J., and Z. Zhou, 1993, Optimal investment strategies for controlling drawdowns, Mathematical Finance, vol. 3 n°3, p. 241-276.

• Hamelink F., H. Harasty, and P. Hillion, 2001, “Country sector or style: what matters most when constructing global equity portfolios? An empirical investigation from 1990-2001”, working paper, FAME.

References



• Jagannathan R., and Wang Zhenyu, "The conditional CAPM and the cross-section of expected returns", Journal of Finance, vol. 51, 1, March 1996, p. 3-53.

• Ibbotson Associates, 2002, “Stocks, bonds, bills, and inflation”, 2002 Yearbook, Chicago.

• Kostovetsky, L., 2003, “Index mutual funds and exchange-traded funds”, Journal of Portfolio Management, vol. 29, summer, p. 80-92.

• Lakonishok, J., Shleifer, A., and R. Vishny, 1994, “Contrarian investment, extrapolation and risk“, Journal of Finance, 5, 1541-1578

• Lhabitant, F.-S., D. Mirlesse, and M. Chardon, 2006, “Absolute returns in wealth management: Implementing risk controlled strategies”, The Journal of Financial Transformation.

• Longin F., and B. Solnik, 1995, "Is the correlation in international equity returns constant?", Journal of International Money and Finance, vol. 14, p. 3-26.

• Martellini, L., 2006, "Intertemporal asset allocation decisions in the presence of liability constraints", working paper, EDHEC Risk and Asset Management Research Centre.

• Merton, Robert C., 1971, "Optimum Consumption and Portfolio Rules in a Continuous-Time Model", Journal of Economic Theory, vol. 3, n°4, p. 373-413.

• —., 1973, "An intertemporal capital asset pricing model", Econometrica, 41, p. 867-888.

• Petkova, R., and L. Zhang, 2002, "Is value riskier than growth?" working paper, William E. Simon Graduate School of Business Administration, University of Rochester.

• Poterba, J. M. and J. B. Shoven, 2002, "Exchange-traded funds: A new investment option for taxable investors", American Economic Review, vol. 92, n°2, p. 422-427.

• Ross, S., 1976, The arbitrage theory of capital asset pricing, Journal of Economic Theory, December, p. 341-360.

• Rubinstein, M., 1989, “Market basket alternatives”, Financial Analysts Journal, vol. 45, September/October, p. 20-29.

• Scherer, B., 2002, Portfolio construction and risk budgeting, Riskbooks.

• Sharpe, W., 1964, “Capital asset prices: a theory of market equilibrium under conditions of risk“, Journal of Finance, 19, 425-442.

• Stock, H. J., 2006, “The ETF Explosion”, Bank Investment Consultant, vol. 14, n°5, p. 22-27.

References



The choice of asset allocationThe EDHEC Risk and Asset Management Research Centre structures all of its research work around asset allocation. This issue corresponds to a genuine expectation from the market. On the one hand, the prevailing stock market situation in recent years has shown the limitations of active management based solely on stock picking as a source of performance.

On the other, the appearance of new asset classes (hedge funds, private equity), with risk profiles that are very different from those of the traditional investment universe, constitutes a new opportunity in both conceptual and operational terms. This strategic choice is applied to all of the Centre's research programmes, whether they involve proposing new methods of strategic allocation, which integrate the alternative class; measuring the performance of funds while taking the tactical allocation dimension of the alphas into account; taking extreme risks into account in the allocation; or studying the usefulness of derivatives in constructing the portfolio.

An applied research approachIn an attempt to ensure that the research it carries out is truly applicable, EDHEC has implemented a dual validation system for the work of the EDHEC Risk and Asset Management Research Centre. All research work must be part of a research programme, the relevance

and goals of which have been validated from both an academic and a business viewpoint by the Centre's advisory board. This board is made up of both internationally recognised researchers and the Centre's business partners. The management of the research programmes respects a rigorous validation process, which guarantees both the scientific quality and the operational usefulness of the programmes.

To date, the Centre has implemented six research programmes:Asset Allocation and Alternative Diversification Sponsored by SG Asset Management and Newedge

The research carried out focuses on the benefits, risks and integration methods of the alternative class in asset allocation. From that perspective, EDHEC is making a significant contribution to the research conducted in the area of multi-style/multi-class portfolio construction.

Performance and Style AnalysisPart of a business partnership with EuroPerformance (Member of Telekurs Financial)The scientific goal of the research is to adapt the portfolio performance and style analysis models and methods to tactical allocation. The results of the research carried out by EDHEC thereby allow portfolio alpha to be measured not only for stock picking but also for style timing.

Indices and BenchmarkingSponsored by Af2i, Barclays Global Investors, BNP Paribas Investment Partners, NYSE Euronext, Lyxor Asset Management, and UBS Global Asset ManagementThis research programme has given rise to extensive research on the subject of indices and benchmarks in both the hedge fund universe and more traditional investment classes. Its main focus is on analysing

About the EDHEC Risk and Asset Management Research Centre

40% Strategic Asset Allocation

3.5% Fees

11% Stock Picking

45.5 Tactical Asset Allocation

Percentage of variation between funds

Source: EDHEC (2002) and Ibbotson, Kaplan (2000)

EDHEC is one of the top five business schools in France.

Its reputation is built on the high quality of its faculty (104

professors and researchers from France and abroad) and

the privileged relationship with professionals that the school has been developing since its

establishment in 1906. EDHEC Business School has decided

to draw on its extensive knowledge of the professional

environment and has therefore focused its research on themes

that satisfy the needs of professionals. EDHEC is

also one of the few business schools in Europe to have received the triple

international accreditation: AACSB (US-Global), Equis

(Europe-Global) andAssociation of MBAs

(UK-Global).EDHEC pursues an active

research policy in the field of finance. The EDHEC Risk and Asset Management Research Centre carries out numerous research programmes in the areas of asset allocation and

risk management in both the traditional and alternative

investment universes.




the quality of indices and the criteria for choosing indices for institutional investors. EDHEC also proposes an original proprietary style index construction methodology for both the traditional and alternative universes. These indices are intended to be a response to the critiques relating to the lack of representativeness of the style indices that are available on the market. In 2003, EDHEC launched the first composite hedge fund strategy indices.

Asset Allocation and DerivativesSponsored by Eurex, SGCIB and the French Banking FederationThis research programme focuses on the usefulness of employing derivative instruments in the area of portfolio construction, whether it involves implementing active portfolio allocation or replicating indices. “Passive” replication of “active” hedge fund indices through portfolios of derivative instruments is a key area in the research carried out by EDHEC. This programme includes the “Structured Products and Derivatives Instruments” research chair sponsored by the French Banking Federation.

Best Execution and Operational PerformanceSponsored by CACEIS, NYSE Euronext, and SunGard This research programme deals with two topics: best execution and, more generally, the issue of operational risk. The goal of the research programme is to develop a complete framework for measuring transaction costs: EBEX (“Estimated Best Execution”) but also to develop the existing framework for specific situations (constrained orders, listed derivatives, etc.). Research also focuses on risk-adjusted performance measurement of execution strategies, analysis of market

impact and opportunity costs on listed derivatives order books, the impact of explicit and implicit transaction costs on portfolio performances, and the impact of market fragmentation resulting from MiFID on the quality of execution in European listed securities markets. This programme includes the “MiFID and Best Execution” research chair, sponsored by CACEIS, NYSE Euronext, and SunGard.

ALM and Asset ManagementSponsored by BNP Paribas Investment Partners and AXA Investment ManagersThis research programme concentrates on the application of recent research in the area of asset-liability management for pension plans and insurance companies. The research centre is working on the idea that improving asset management techniques and particularly strategic allocation techniques has a positive impact on the performance of asset-liability management programmes. The programme includes research on the benefits of alternative investments, such as hedge funds, in long-term portfolio management. Particular attention is given to the institutional context of ALM and notably the integration of the impact of the IFRS standards and the Solvency II directive project. It also aims to develop an ALM approach addressing the particular needs, constraints, and objectives of the private banking clientele. This programme includes the “Regulation and Institutional Investment” research chair, sponsored by AXA Investment Managers, and the “Asset Liability Management and Institutional Investment Management” research chair, sponsored by BNP Paribas Investment Partners.




Five Research Chairs have been endowed:Regulation and Institutional InvestmentIn partnership with AXA Investment ManagersThe chair investigates the interaction between regulation and institutional investment management on a European scale and highlights the challenges of regulatory developments for institutional investment managers.

Asset-Liability Management and Institutional Investment ManagementIn partnership with BNP Paribas Investment PartnersThe chair examines advanced Asset-Liability Management topics such as dynamic allocation strategies, rational pricing of liability schemes, and formulation of an ALM model integrating the financial circumstances of pension plan sponsors.

MiFID and Best ExecutionIn partnership with NYSE Euronext, SunGard, and CACEIS Investor ServicesThe chair looks at two crucial issues linked to the Markets in Financial Instruments Directive: building a complete framework for transaction cost analysis and analysing the consequences of market fragmentation.

Structured Products and Derivative InstrumentsSponsored by the French Banking Federation (FBF) The chair investigates the optimal design of structured products in an ALM context and studies structured products and derivatives on relatively illiquid underlying instruments.

Financial Engineering and Global Alternative Portfolios for Institutional InvestorsSponsored by Morgan Stanley Investment Management The chair adapts risk budgeting and risk management concepts and techniques to the specificities of alternative investments, both in the context of asset management and asset-liability management.

The EDHEC PhD in FinanceThe PhD in Finance programme at EDHEC Business School is designed for professionals who aspire to higher intellectual levels and aim to redefine the investment banking and asset management industries

It is uniquely offered in two tracks: a ‘residential track‘ for high-potential graduate students who will hold part-time positions at EDHEC Business School, and an ‘executive track‘ for high-level practitioners who will keep their full-time jobs.

Drawing its faculty from the world’s best universities and enjoying the support of the research centre with the most impact on the European financial industry, the EDHEC PhD in Finance creates an extraordinary platform for professional development and industry innovation




Industry surveys: comparing research advances with industry best practices EDHEC regularly conducts surveys on the state of the European asset management industry. They look at the application of recent research advances within investment management companies and at best practices in the industry. Survey results receive considerable attention from professionals and are extensively reported by the international financial media.

Recent industry surveys conducted by the EDHEC Risk and Asset Management Research Centre1/ The EDHEC European Investment Practices Survey 2008 sponsored by Newedge

2/ The Impact of IFRS and Solvency II on Asset-Liability Management and Asset

Management in Insurance Compagnies sponsored by AXA Investment Managers

3/ EDHEC European Real Estate Investment and Risk Management Survey sponsored

by Aberdeen Property Investors and Groupe UFG

EuroPerformance-EDHEC Style Ratings and Alpha League TableThe business partnership between France’s leading fund rating agency and the EDHEC Risk and Asset Management Research Centre led to the 2004 launch of the EuroPerformance-EDHEC Style Ratings, a free rating service for funds distributed in Europe which addresses market demand by delivering a true picture of the alphas, accounting for potential extreme loss, and measuring performance persistence. The risk-adjusted performance of individual funds is used to build the Alpha League Table, the first ranking of European asset management companies based on their ability to deliver value on their equity management.www.stylerating.com

EDHEC-Risk websiteThe EDHEC Risk and Asset Management Research Centre’s website makes EDHEC’s analyses and expertise in the field of asset management and ALM available to professionals. The site examines the latest academic research from a business perspective, and provides a critical look at the most recent industry news.www.edhec-risk.com

Research for BusinessTo optimise exchanges between the academic and business worlds, the EDHEC Risk and Asset Management Research Centre maintains a website devoted to asset management research for the industry: www.edhec-risk.com, circulates a monthly newsletter to over 170,000 practitioners, conducts regular industry surveys and consultations, and organises annual conferences for the benefit of institutional investors and asset managers. The Centre’s activities have also given rise to the business offshoots EDHEC Investment Research and EDHEC Asset Management Education.

EDHEC Investment Research supports institutional investors and asset managers in the implementation of the centre’s research results and proposes asset allocation services in the context of a core-satellite approach encompassing alternative investments.

EDHEC Asset Management Education helps investment professionals to upgrade their skills with advanced risk and asset management training across traditional and alternative classes.

1- Morgan Stanley report, 2007 review2 - Barclays Global Investors, May 2008



About iShares

iShares Exchange-Traded Funds (ETFs), managed by Barclays Global Investors, are efficient and innovative investment vehicles which offer excellent means to get instant exposure to different countries, regions, asset classes and much more.

iShares global overview• world’s No 1 ETF provider1

• manage over 320 ETFs• manage more than €275 billion in ETF assets• listings in 14 countries2

Behind iShares is the world’s largest fund managerIn the world of investment, experience breeds innovation. At Barclays Global Investors (BGI) it also takes a highly disciplined approach to breed new products. More than 30 years ago BGI developed the first index strategy. Since then, BGI has continued evolving its index expertise through in-depth research and risk analysis, leading to the launch of the iShares funds.

iShares are powerful investmentsiShares combine the best features of stocks and mutual funds. As flexible as common stock, iShares ETFs can be bought and sold during trading hours through a broker or a financial advisor. Similarly to mutual funds, iShares ETFs seek to reflect the performance of a chosen index, through holding a diversified underlying basket of assets with, for example, a domestic or an international focus.

iShares funds are listed on exchanges across Europe• Borsa Italiana• Eurolist by Euronext Amsterdam• Eurolist by Euronext Paris• Frankfurt Stock Exchange• London Stock Exchange• SWX Swiss Exchange• SWX Europe• Chi-x Europe LtdiShares powerful combination of easy market access, diversification and cost efficiency, along with global reach, appeals to several types of investors, allowing them to select and implement a desired style or strategy. From growth or value, small-medium and large-caps, to even dividend or sub-sector driven products, iShares offer exposure to the indices that our investors need.

iShares expanded beyond traditional equity tracker portfoliosiShares cover the world’s largest markets and the most demanded assets. Their expansion is mostly driven by their clients’ needs. iShares diverse offering allows investors to instantly build tailor-made investment portfolios, adjust their risk profile and benefit from new market trends, in full transparency and at very competitive rates.

Contact iSharesTo assess fund performance and holdings, index information, web-based investment tools, trading strategies using iShares and much more, please visit www.ishares.eu or call:UK: 0845 357 7000France: 0800 940 299Germany: +49 (0) 89 42729 5858Italy: 800 898085Netherlands: 0800 0233 466Switzerland: 0848 44 3300

EDHEC Risk and Asset ManagementResearch Centre393-400 promenade des AnglaisBP 311606202 Nice Cedex 3 - FranceTel.: +33 (0)4 93 18 78 24Fax: +33 (0)4 93 18 78 41E-mail: [email protected]: www.edhec-risk.com

The EDHEC European ETF Survey 2008€¦ · portfolio performance analysis, and active asset...

Documents

Transcript of The EDHEC European ETF Survey 2008€¦ · portfolio performance analysis, and active asset...