EDHEC-Risk Concept Industry Analysis Featured Analysis Latest EDHEC-Risk Surveys Features Interviews Indexes and Benchmarking FTSE EDHEC-Risk Efficient Index Series FTSE EDHEC-Risk ERAFP SRI Index EDHEC-Risk Alternative Indexes EDHEC IEIF Quarterly Commercial Property Index (France) Hedge Fund Index Research Equity Index Research Amundi "ETF, Indexing and Smart Beta Investment Strategies" Research Chair Rothschild & Cie "Active Allocation to Smart Factor Indices" Research Chair Index Regulation and Transparency ERI Scientific Beta Performance and Risk Reporting Hedge Fund Performance Performance Measurement for Traditional Investment CACEIS "New Frontiers in Risk Assessment and Performance Reporting" Research Chair Asset Allocation and Alternative Diversification Real Assets Meridiam Infrastructure/Campbell Lutyens "Infrastructure Equity Investment Management and Benchmarking" Research Chair Natixis "Investment and Governance Characteristics of Infrastructure Debt Instruments" Research Chair Société Générale Prime Services (Newedge) "Advanced Modelling for Alternative Investments" Research Chair CME Group "Exploring the Commodity Futures Risk Premium: Implications for Asset Allocation and Regulation" Strategic Research Project Asset Allocation and Derivative Instruments Volatility Research Eurex "The Benefits of Volatility Derivatives in Equity Portfolio Management" Strategic Research Project SGCIB "Structured Investment Strategies" Research ALM and Asset Allocation Solutions ALM and Private Wealth Management AXA Investment Managers "Regulation and Institutional Investment" Research Chair BNP Paribas Investment Partners "ALM and Institutional Investment Management" Research Chair Deutsche Bank "Asset-Liability Management Techniques for Sovereign Wealth Fund Management" Research Chair Lyxor "Risk Allocation Solutions" Research Chair Merrill Lynch Wealth Management "Risk Allocation Framework for Goal-Driven Investing Strategies" Research Chair Ontario Teachers' Pension Plan "Advanced Investment Solutions for Liability Hedging for Inflation Risk" Research Chair Non-Financial Risks, Regulation and Innovations Risk and Regulation in the European Fund Management Industry Index Regulation and Transparency Best Execution: MiFID and TCA Mitigating Hedge Funds Operational Risks FBF "Innovations and Regulations in Investment Banking" Research Chair EDHEC-Risk Publications All EDHEC-Risk Publications EDHEC-Risk Position Papers IPE EDHEC-Risk Institute Research Insights AsianInvestor EDHEC-Risk Institute Research Insights P&I EDHEC-Risk Institute Research for Institutional Money Management Books EDHEC-Risk Newsletter Events Events organised by EDHEC-Risk Institute EDHEC-Risk Smart Beta Day Amsterdam 2017, Amsterdam, 21 November, 2017 EDHEC-Risk Smart Beta Day North America 2017, New York, 6 December, 2017 Events involving EDHEC-Risk Institute's participation EDHEC-Risk Institute Presentation Research Programmes Research Chairs and Strategic and Private Research Projects Partnership International Advisory Board Team EDHEC-Risk News EDHEC-Risk Newsletter EDHEC-Risk Press Releases EDHEC-Risk in the Press Careers EDHEC Risk Institute-Asia EDHEC Business School EDHEC-Risk Executive Education EDHEC-Risk Advances in Asset Allocation Blended Learning Programme 2017-2018 Yale School of Management - EDHEC-Risk Institute Certificate in Risk and Investment Management Yale SOM-EDHEC-Risk Harvesting Risk Premia in Alternative Asset Classes and Investment Strategies Seminar, New Haven, 5-7 February, 2018 Investment Management Seminars Contact EDHEC-Risk Executive Education Contact Us ERI Scientific Beta EDHEC PhD in Finance
Indices & Benchmarks - December 16, 2014

Passive investment can add much more value than it does today - an interview with Noël Amenc

In this month's interview, we speak with Noël Amenc, Professor of Finance at EDHEC Business School, Director of EDHEC-Risk Institute and CEO of ERI Scientific Beta, about passive investment in general and smart beta investing in particular.


Noël Amenc

The recent report from the Financial Services User Group, which advises the European Commission in the preparation of legislation or policy initiatives which affect the users of financial services, reveals that European investment management does not create much value, that few active managers outperform indices, and that even fewer persist in their outperformance. Do you think that this observation could be the death knell for the active asset management industry, which is continually losing market share to passive investment?

Noël Amenc: This study is not a scoop—its results are perfectly consistent with those of all the academic studies produced on mutual fund performance measurement in both the US and Europe. The vast majority of the studies show that if one takes account of the premia associated with risk factors that are well known in the financial literature to be rewarded over the long term (the market factor, and also the value, size, momentum and, more recently, volatility, profitability and investment factors), then it is indeed possible to capture most of the performance of any active manager that attempts to beat a market index. Where the results of the studies published by the Financial Services User Group are more surprising is in the area of multi-class investment management, where it is shown that the flexible investment management offered by the professionals delivers worse performance than a fixed benchmark. Even though dynamic allocation has been the subject of extensive research1, which has demonstrated its benefits for both institutional and private wealth management, it appears that professionals do not know how to implement it and prefer to dream, or have others dream, about the benefits of tactical allocation rather than putting the real and robust benefits of dynamic risk allocation into practice2.

Ultimately asset managers invest a lot of resources and spend a lot of time on second-order questions, such as stock picking and tactical allocation, which give results of limited significance and robustness over the long term, while neglecting first-order topics such as strategic allocation.

Passive investment has a bright future because it provides a response to this issue. Passive investment contributes to the efficiency of the investment process because it avoids wasting risk budgets and resources on the false promises of alpha.

But does that ultimately confirm John Bogle’s premise that the first necessary, and often sufficient, step in any investment process is to invest passively by replicating market indices?

Noël Amenc: I think that one should differentiate between efficient passive investment and traditional index investment.

The origin of this distinction is in the very definition of active investment management. If one defines active investment management as the capacity to make decisions that are based on forecasts of the future returns of assets, then yes, this form of active investment is called into question by many academic studies that show that the alpha from active investment is rarely significant and even more rarely persistent. This observation forms the basis for passive investment, which uses the information provided by the market rather than forecast future returns.

That does not mean however that one should give up on making any kind of decision and slavishly replicate market indices. There is the possibility of performing passive investment, i.e. without forecasting future asset returns, on the basis of systematic decisions taken with regard to the observation of market conditions.

While it is relatively pointless to try to estimate expected returns with past information, quite simply because, as Robert Merton (1980) underlined, this estimator is not convergent, it has been shown that the risk estimation converges towards its true value in continuous time. In other words, more data on asset returns does not improve the chance I have of my past estimation being right for the future, but on the other hand, the more data I have to estimate the risks of my assets, the better chance I have of improving my estimation.

All the efforts by researchers and practitioners in the past 50 years have therefore logically concerned asset risk estimation. Using not only econometric and statistical techniques, but also factor models, these efforts involved meeting the challenge of a robust estimation in discrete time of the risk parameters that could be used in portfolio construction and asset allocation models.

It is this positioning on risk that is the source of both smart beta and risk allocation offerings.

We can no longer speak of active investment in the traditional sense of the term since the promise is limited to good estimation of risks, but nor does it involve cap-weighted index replication that would trust the market to naturally offer an efficient portfolio or allocation.

We can speak of "efficient" or "smart" passive investment. This investment management does not aim to build portfolio performance by forecasting future asset returns or by detecting anomalies that are liable to create arbitrage profits, but by taking the information available on the market into account in order to construct efficient portfolios.

The promise is less ambitious. It does not involve aiming for absolute performance, which pure alpha managers or tactical allocation specialists claim to be able to achieve, but simply capturing the market’s risk premia in an efficient way and allocating the investor’s risk budget in an equally efficient way between the various market factors that are rewarded over the long term.

The value added by this smart or efficient passive investment is actually based on the principle of diversification to which EDHEC-Risk Institute is particularly attached3.

The diversification of smart factor indices4 involves being able to achieve the best risk-adjusted returns for each of the factors rewarded over the long term, which constitute natural candidates for single-class or multi-class asset allocation.

In this regard, diversification plays the role that Harry Markowitz and William Sharpe first gave it, i.e. reduce the non-rewarded or specific risks to obtain the best return possible for a given quantity of risk (beta).

Diversification will also be the core of the idea of efficient allocation between risk factors. Here, it no longer involves removing non-rewarded risks but genuinely diversifying the exposure to rewarded risk factors. This diversification can rely on many risk allocation techniques in the context of absolute or relative risk management and constraints. With the support of Amundi ETF & Indexing, we recently published5 the conclusions of applied research on advances in the area of smart beta allocation which allows us to observe that risk allocation between smart beta indices enables absolute risk budgets, such as factor risk parity or minimising the total volatility of the portfolio, to be respected, but also, in a relative context, to substantially reduce the tracking error or the risk of maximum drawdown (max relative drawdown) with respect to an index. As such, two results deserve to be highlighted.

The first is that in relative terms, it is possible to sharply reduce the tracking error and the relative drawdown of the smart beta investment with robust risk allocation techniques in a portfolio of smart beta indices. As such, a Relative Equal Risk Contribution (ERC) or Relative Global Minimum Variance (R-GMV) approach for a Developed World universe gives tracking error of around 2.5% with max relative drawdown of 5%, even though the max relative drawdown of smart beta indices can often exceed 30%.

The second is that in absolute terms and as part of a long-only allocation, even though investable smart beta indices are never pure in the long-only space, it is possible to respect factor risk parity constraints. This result means that it is not necessary to turn to long/short or highly concentrated factor indices that present investability problems and are particularly poorly diversified when achieving objectives on controlled exposure to risk factors.

This passive investment is not in any way static. Whether it involves diversifying the indices that are representative of the rewarded risk factors or allocating risk budgets between these indices, there are indeed systematic decisions on buying and selling securities and rebalancing the portfolio or the allocation.

These movements are not the fruit of a discretionary decision on the part of the manager but of the application of a risk management model.

It is curious moreover to observe today that on the pretext that smart beta is delivered in index form, many professionals are reticent towards this movement. They would like to benefit from dynamic allocation without turnover on the pretext that turnover is costly! What is most costly today is not so much the turnover of the indices or the turnover resulting from risk allocation between the indices, which remains limited (with figures varying from 25% to 40% of one-way annual turnover), but that these indices are poorly designed. For example, the annual difference in performance over the last 10 years for the developed universe between good minimum volatility indices such as those produced by FTSE or Scientific Beta, and a poorly designed minimum volatility index, such as MSCI’s, is respectively 2.74% and 2.16%6.

In the same way, between well diversified factor indices, which we can refer to as "smart," and cap-weighted factor indices, we observe differences of 68% in Sharpe ratio over the long term7.

Ultimately the true cost of smart beta is rarely the index replication fees, which are limited, or even the transaction fees, which are also limited for a broad mid-cap index in the developed universe, but the choice of index, which can perform poorly because it was poorly designed.

How would you characterise that poor design?

Noël Amenc: The poor design of a smart beta index actually relates to the different dimensions of smart beta.

The first is the fact that smart beta is representative of risk factors that are different from those that are contained in the cap-weighted indices.

This exposure to risk factors can be controlled to a greater or lesser degree. In 20128, EDHEC-Risk Institute made a clear distinction between first-generation smart beta indices, where exposure to factors was implicit and uncontrolled, and second-generation indices, where this factor exposure is explicit, controlled, and even optimised.

Moreover, since the Smart Beta 2.0 approaches make a clear difference between desired risk exposures (because they are rewarded) and undesired risk exposures (specific risks and unrewarded factors), they constitute a consistent methodological and conceptual framework which enables the investor to diversify, or even neutralise, non-diversified factor exposures, which are often the source of irregular performance and underperformance that is not explained solely by the factor exposure that is displayed or claimed.

As such, so-called "fundamental" indices, which are used today as investment proxies for the value factor, due to their lack of diversification and their weighting criteria which maximise the in-sample performance of the track records by avoiding technology stocks, lead to sector biases that are not controlled by investors and have turned out to deliver volatile and contrasted performance in recent years.

The second is that, since they often wish to create proprietary approaches that would justify original promotion on the market, or their fees, smart beta promoters confuse scientific research with marketing innovation.

The academic community is fairly conservative when it comes to unveiling a new factor. For example, Fama and French waited more than 20 years before introducing two additional factors to explain equity returns. This conservatism is highly correlated with a concern to publish robust results. It aims to avoid a new factor actually being a proxy for a more fundamental existing factor in order to preserve the parsimony of the model, which is the primary condition for robustness. It also aims to check the statistical significance of the risk premia associated with these factors so as to avoid the displayed performance being a short-term illusion, and above all to ensure that these new factors are not statistical artefacts or the product of factor fishing.

On the practitioner side, the principles of truth and robustness are often outweighed by business realities. Managers’ and investors’ intuition has little regard for scientific precautions. In recent years, most smart beta strategies have relied on marketing innovation that has had nothing to do with academic consensus. This represents a genuine danger for this new form of investment that is supposed to allow the investor to capture the risk premia available on the market and would not have to depend on the opinion of a manager to introduce stock-picking-type approaches into the methodologies.

As such, in a recent study on the robustness of smart beta we were given to observe that none of the traditional accounting criteria used in the indices termed "fundamental" enabled significant positive risk premia to be obtained over the long term9. This non-significance contrasts with the results obtained using traditional Book-to-Market or Earnings-to-Price ratios.

In fact, by wishing to avoid any reference to the market, the fundamental approaches have an accounting defect; they no longer convey relevant information for constructing a portfolio. In a concern to seduce investors who have been addicted to stock picking for so many years and therefore in search of anomalies, they have one believe, without any serious academic basis, that there are structural anomalies in the market and overvalued stocks that the observation of highly backward-looking data, which accounting averages are, would enable one to capture.

In fact, if cap-weighted indices are not necessarily efficient benchmarks, it is not only because they are overly concentrated, but also because they are exposed to factors that are poorly rewarded over the long term, such as large-cap or growth. These large-cap or growth stocks are not overvalued but simply, in accordance with the asset pricing theory consensus10 marginally, i.e. over a short period, less risky than value or small-cap stocks. It is totally inconsistent, as we see too frequently, to begin a factor investing presentation with a critique of cap-weighted indices for being speculative or overvalued and then to refer to academic research, which rejects any idea of overvaluation or undervaluation over the long term, to justify factor investment proposals.

In the same way, and the marketing nomenclature contributes to this confusion, having one believe that so-called ‘quality’ indices contain stocks that are less risky and more profitable over the long-term is scientifically counterfactual. The scientific basis for the existence of risk premia on stocks that are present in quality factors is their higher marginal short-term risk. An investor with a long-term horizon will be able to accept these short-term risks and therefore benefit from the risk premia that are actually associated with two factors (profitability and low investment) and are explained by a higher cost of capital, which in turn is related to a higher appreciation of the risks of the firm or of its investment projects. This type of marketing confusion often leads to quality indices that are based on data mining with composite quality scores that mix very different factor proxies and lead to the construction of indices that are no longer truly exposed to the low investment factor or the high-profitability factor11.

Ultimately, wanting to sell smart beta as a search for anomalies using accounting data is inconsistent with smart beta’s starting point, which was to avail of a simple, systematic strategy to capture these risk premia in an efficient way within the market, not to beat it by using accounting data! If investors wish to continue to hope to capture anomalies, they should get as far away as possible from passive investment, whether traditional or smart.

The third and final dimension of smart beta is not linked to good risk factor exposure but to diversification.

Here too, anecdotal evidence and false innovations make concepts that are clearly documented in the financial literature confusing. There is no contradiction between investing in factors and being well diversified between these factors, and above all lowering the specific risks associated with these factors. In the long-only world there are no pure factors. A stock is exposed to multiple risks, whether they are rewarded or not. The cost of purity is the cost both of the short and of the level of concentration of the index, which in the end is no longer truly investable and above all very poorly diversified in terms of specific risks. The pure factor index approaches are statistical artefacts that can of course remove a large number of risk factors through optimisation but leave the investor with illiquid indices that have considerable unrewarded specific risks. If investors wish to guarantee that they control their portfolio’s exposure to different risk factors in their allocation, they can do it with smart factor indices that are not pure. It simply involves imposing absolute or relative risk budgets with respect to the cap-weighted index in their allocation to these different smart factor indices, which should remain simple and parsimonious in their design12.

In the same way as the legitimate desire to control the factor exposure of one’s portfolio should not remove the concern to soundly diversify the indices that are the ingredients for this diversification, one should not overplay the benefits of this diversification to investors either.

In particular, we have to reject some of the arguments put forward by the defenders of first-generation smart beta indices that are based on the sole objective of diversifying the portfolio and not having to deal with factors explicitly. Claiming that this type of smart beta will eliminate unrewarded risk and leave you with just the right mix of rewarded risk is both trivial and untrue.

The trivial aspect should first be explained. In our academic dreams, there is an ideal portfolio that is extremely well diversified. It is the true market portfolio.

It is clear that the true market portfolio from modern portfolio theory, as formalised by William Sharpe, is perfectly diversified, i.e. it is an equilibrium Max Sharpe Ratio portfolio and as such only contains perfect exposures to rewarded factors.

The problem is that this portfolio cannot be found by itself and relies on equilibrium conditions that are not realistic—notably that all investors have the same aversion to risk, that they all have the same investment horizons, and that it is possible to invest in all goods, including human capital.

That is why all financial researchers and practitioners in the past 50 years have been trying to construct proxies for this extremely well diversified portfolio with more realistic assumptions, notably based on arbitrage theory, which relaxes the assumptions on the universality of investment opportunities and homogeneity of behaviours that underlie the equilibrium portfolio.

There are in fact two ways to think of this Max Sharpe Ratio portfolio research.

The first is to replace market forces with an ex ante optimisation that will attempt to construct a portfolio that one hopes will be close to the Max Sharpe Ratio ex post.

This research, which does not pertain to any factor investment approach, presupposes that the Sharpe ratio portfolio obtained will be so well diversified that only rewarded factors will be available, in quantities that guarantee the best Sharpe ratio. The problem is that this diversification always relies on an unrealistic idea, namely, that one can invest in all the opportunities available on the market. In fact, a Max Sharpe Ratio portfolio is optimal if and only if the optimisation that enabled the Max Sharpe Ratio portfolio to be constructed is conducted on all investment opportunities available. This is never the case. As soon as one carries out this optimisation on a limited number of stocks, the maximum in the case of an index being the index universe, the Max Sharpe Ratio obtained, on condition that it is still the Max Sharpe Ratio ex post, does not represent the best factor diversification, but only a partial optimal portfolio constrained by the selection of stocks contained in the index. For example the true Max Sharpe Ratio portfolio formed on the basis of large-cap growth stocks will have a wrong (that is, negative) exposure to the value premium and the size premium even in the hypothetical absence of parameter uncertainty.

It is in this context that the notions of rewarded risk factors were put forward, notably through the work of Richard Roll and Stephen Ross (1976) and "microeconomic" applications of Arbitrage Pricing Theory (APT) on the basis of the work of Eugene Fama and Kenneth French (1993). The factor approach removes the assumption of completeness of investment opportunities. The idea behind the risk factor approach is that stocks that are exposed to risk factors such as value, small cap, and momentum are riskier than others, in the sense that over a short period there can be extremely negative conditions for the economy (and therefore for the markets) that mean that these stocks will lose more value than others and as such will be less useful in protecting consumption capacity at the time when it is most needed. It is therefore because value, small cap and to some extent momentum stocks tend to do particularly poorly in bad times, when marginal utility of consumption is high, that these stocks are more risky, less in demand and therefore command a higher expected return.

More recently (cf. papers by Hou, Xue and Zhang (2014a, 2014b)), new factors have been introduced such as high profitability and low investment, which also reflect greater company risk. The low volatility factor is also subject to a risk-based explanation (cf. paper by Frazzini and Pedersen( 2014)) which distances it from a behaviourist explanation (anomaly). Indeed, it is this rational explanation that led us to integrate the low volatility factor in our flagship multi-beta multi-strategy index.

As we have already highlighted, these premia based on rational explanations (they are associated with higher marginal risk) are therefore persistent and can be captured by long-term investors whose liabilities/investment horizon enables them to withstand these high marginal (short-term) risks.

From our point of view, this is the key element in the superiority of the performance of smart beta compared to cap-weighted indices, which are exposed to poorly rewarded factors.

Naturally, choosing factor exposures is a necessary but insufficient first step, because one must then ensure that the non-rewarded risks are well diversified (reduced) in order for the risk-adjusted performance associated with this factor to be as good as possible. That is why we differentiate between factors and smart factors. We offer the latter, because they associate an explicit choice of factor with good diversification of non-rewarded risks.

Of course, once these smart factors are available, the investment can allocate them according to different objectives, including within a Max Sharpe Ratio framework. One is no longer dependent on the set of stocks available but, on condition that it is indeed the right set of factors, can construct an allocation that is a robust and efficient proxy of an optimal portfolio.

This choice of allocation is essential. From a pragmatic point of view, there is no universally right or wrong level of factor exposure. Factor indices empower investors to make factor allocation decisions based on their risk tolerance and investment context. For example, a value tilt may be undesirable for an investor whose labour income depends heavily on the economic cycle, because value stocks tend to become more risky during recessions. By contrast, investors with smooth labour income may be willing to tilt towards value stocks and harvest the associated risk premium. Using factor indices is like taking well-established prescription medicine, where effectiveness and side effects are well documented. With this information, one can weigh the benefits and the risks, and decide whether it is appropriate for a particular case.

The untrue aspect of the premise that factors do not matter deserves an illustration.

For the reasons mentioned above, and while we are convinced that diversification adds value by maximising the Sharpe ratio associated with factors, we cannot accept the argument that the implementation of the diversification method of a portfolio or index dispenses with the need to address and control the exposures to rewarded risk factors.

Unless one thinks that an optimisation of the diversification of a portfolio carried out ex ante on a limited number of stocks leads ex post to a generalised Max Sharpe Ratio portfolio, then there is a scientific untruth in the idea that diversification can replace the factor allocation decision.

You speak of the added value of passive investment, but in their positioning on smart beta a large number of active asset managers are criticising the smart beta index approaches. They consider that these approaches are often too systematic and therefore blinded, and also that they can be subject to front running and arbitrage because of their transparency.

Noël Amenc: Indeed—because of its success smart beta is a tempting target and the subject of competition between the various players in the market. A recent editorial by Jacobs and Levy, which appeared in the Financial Analysts Journal13 summarises the arguments used by the defenders of this approach fairly well.

In short, smart beta indices are alleged to suffer from three main evils, namely transparency, which allows front running; the simple and replicable nature of the factor construction criteria, which allows arbitrage profits and would disappear through a phenomenon termed "factor crowding;" and the smart beta index’s exposure to a single dimension of risk.

Objectively, one can understand that the emergence of smart beta in passive investment poses problems for many "systematic" active managers. Most of their performance came de facto from capturing risk premia that are packaged today in indices that are accessible at much lower prices than those of active management. However, we think that the right response from active managers to this concern lies in their capacity to add value in the process of allocation between these factors rather than the caricature of smart beta and the anecdotal evidence without any genuine scientific basis that are the core of Jacobs and Levy or other active managers’ affirmations.

First of all, we find it inconsistent, to say the least, or to be honest, disingenuous, to criticise the systematic and transparent character of smart beta indices, which are the only guarantees of the accuracy and credibility of the simulated long-term track records that are published, while at the same time, as an active manager, wanting to explain, on the basis of non-transparent techniques and leaving room for discretionary decisions, that one can do much better than these simulated track records. How credible is a statement of that kind when the non-transparent and discretionary nature of the active strategies proposed is precisely the reason why the use of simulated track records is forbidden? One might also add that when one compares the best active smart beta strategies with the best indices that are representative of those same strategies, the results are not favourable for active investment for the comparable live periods.

On the subject of the risk of anomalies disappearing due to crowding, one should recall that what characterises smart factor investing is the search for risk premia that have rational or behavioural explanations, and that as such are persistent and not, as is the case in traditional active investment management, a search for anomalies or mispricing. Risk premia exist not because few people are aware of them but simply because they are supported by the logic of the market or of the behaviour of the actors in the market. More prosaically, smart beta index rebalancing varies greatly between providers and strategies and concerns a huge number of stocks, which makes the crowding, or even front running, effects hard to believe.

On the more specific subject of front running, which is allegedly the result of the excessive transparency of smart beta indices14, one should first recall that smart beta indices do not give the market information on future rebalancings, which relate not only to foreseeable additions and deletions in the universe, but also to changes in weights relating to applications of the weighting methods. Transparency, when it exists, is generally provided ex post. Moreover, here too the increasing number of smart beta strategies and rebalancing techniques make any anticipation of the rebalancing fairly difficult15. A manager who uses one or more smart beta indices as a reference is perfectly capable of taking some leeway, which will be expressed as tracking error, to guarantee best execution and notably to cope with the "blind" rebalancing which presupposes application of the index’s management and construction methodology. This chosen tracking error reconciles for example the integration of an exceptional event such as a sharp rise in volatility which means that the parameter estimation in the calibration period (in-sample) no longer corresponds to the reality of the market at the time of rebalancing and makes the out-of-sample application of the weighting proposed by the model less relevant and therefore less robust, e.g. a particular event on a stock (announced delisting or exceptional corporate action announced at the time of the rebalancing). This optimal execution tracking error reconciles the search for best execution with transparency, and therefore the reliability of long-term simulations used as part of the construction of allocation to smart factor indices.

Again, in a concern for honesty, or at least for relevance, it would be good if criticism such as that from Jacobs and Levy did not have one believe that the multidimensional approach to risk is not possible with smart beta indices. Within the framework of multi-factor approaches many index providers give investors the means to allocate to several risk factors in an explicit and controlled manner. In addition, this allocation can be made using a global benchmark that consolidates the rebalancing relating to each sub-index or factor and thereby limits the transaction costs and the risks of front running.

Ultimately, we feel that the proposal by active smart beta managers to lead investors towards opaque factors that are supposedly the fruit of discretionary decisions and proprietary models that are not documented in the academic literature, entails both a lack of robustness and almost a desire to have investors seek out gurus that are liable to exploit the anomalies in the markets, rather than towards a systematic and rational approach to capturing the risk premia associated with the risk factors identified on the market in an efficient (smart) way.

For many years investors have been paying a high price, in terms of both management fees and poor performance, for the seductive but completely ineffective anecdotes from traditional active investment management. Thanks to its systematic character and its academic foundations, smart beta is a response to this disappointment. It would be a shame if these same active managers distorted smart beta by transforming it into discretionary management. Smart beta deserves better than a fight for the opacity and data mining or factor fishing of proprietary techniques.

We are convinced that smart factor indices, when they are well designed and correspond to academic consensus, can provide genuinely high-performance investment solutions not only for investors but also for active managers. The latter can rationalise their costs and their efforts to add real value in the core part of the investment management process, which is asset allocation. The way to be active in smart beta is not by replicating stock-picking practices that have failed to prove their effectiveness, and even less so when they rely on information that is as backward-looking as accounting data. It is instead by working on risk management and efficient allocation to smart beta indices.



Footnotes

  1. EDHEC-Risk Institute has conducted extensive high-level research in this area. A detailed summary of our work on the contribution of dynamic investment in a retail investment context was published in Amenc, N., R. Deguest, L. Martellini and V. Milhau. July 2012. Long-Term Investing Strategies in Private Wealth Management, EDHEC-Risk Institute Working Paper. The same type of detailed summary on the dynamic approach in Liability-Driven Investing solutions is available in Badaoui S., R. Deguest, L. Martellini and V. Milhau. February 2014. Dynamic Liability-Driven Investing Strategies: The Emergence of a New Investment Paradigm for Pension Funds? EDHEC-Risk Institute Publication supported by BNP Paribas Investment Partners.

  2. A recent study conducted by EDHEC-Risk Institute with the support of BNP Paribas Investment Partners confirms that the vast majority of institutional investors do not implement a risk management and allocation process that is appropriate for their objectives and liabilities. Badaoui S., R. Deguest, L. Martellini and V. Milhau. February 2014. Dynamic Liability-Driven Investing Strategies: The Emergence of a New Investment Paradigm for Pension Funds? EDHEC-Risk Institute Publication.

  3. For a short summary of EDHEC-Risk Institute’s position on the subject, please refer to Amenc N. and L. Martellini, In Diversification We Trust? Journal of Portfolio Management, Vol. 37, No. 2, Winter 2011.

  4. For more information on the smart factor concept, please refer to Amenc N., F. Goltz, A. Lodh and L. Martellini, Towards Smart Equity Factor Indices: Harvesting Risk Premia without Taking Unrewarded Risks, Journal of Portfolio Management, Vol. 40, No. 4, Summer 2014.

  5. Amenc N., R. Deguest, F. Goltz, A. Lodh, E. Shirbini. Risk Allocation, Factor Investing and Smart Beta: Reconciling Innovations in Equity Portfolio Construction. July 2014. EDHEC-Risk Publication supported by Amundi ETF & Indexing.

  6. The performance of the FTSE Developed Minimum Variance Index (USD), the Scientific Beta Developed Efficient Minimum Volatility Index (USD), and the MSCI World Minimum Volatility Index (USD) is calculated using daily total returns over the period 31-December-2003 to 31-December-2013 (10 years).

  7. Average of the differences in Sharpe ratio observed between 31-Dec-1973 and 31-Dec-2013 (40 years) for all long-term track record multi-strategy factor indices and their cap-weighted factor equivalent calculated on a universe of the 500 largest capitalisation US stocks. All the details on the calculations and the indices are available on the www.scientificbeta.com website.

  8. Amenc N., F. Goltz and A. Lodh. Choose Your Betas: Benchmarking Alternative Equity Index Stratégies. Journal of Portfolio Management, Vol. 39, No. 1, Fall 2012. Amenc N. and F. Goltz. Smart Beta 2.0. Journal of Index Investing, Vol. 4, No. 3, Winter 2013.

  9. Amenc N., F. Goltz, A. Lodh and S. Sivasubramanian. October 2014. Robustness of Smart Beta Stratégies. ERI Scientific Beta Publication. This study shows that long/short portfolios based on accounting variables like book value, sales, dividends and cash flow do not show any significant risk premium over the long term (40 years).

  10. For a summary of the economic rationales behind the traditional rewarded factors, one can refer to Amenc N., F. Goltz and A. Lodh, Principles of Equity Factor Investing, in the EDHEC-Risk Institute Research Insights supplement to Investment & Pensions Europe, Spring 2014 (p. 17).

  11. For example, both the Russell 1000 High Efficiency Quality Index and the Russell Developed Large-Cap High Efficiency Quality Index are only exposed to the High Profitability factor and show no significant exposure to the Low Investment factor. Their methodology involves computing composite Quality scores using scores from three variables – Return on Assets, Debt to Equity, and Earnings Variability. More details can be found in ERI Scientific Beta’s forthcoming white paper – “Low Investment and High Profitability Smart Factor Indices.”

  12. A summary of this factor risk budgeting constraint approach in smart beta allocation is available in Amenc, N., R. Deguest, F. Goltz and A. Lodh, Risk Allocation with Smart Factor Indices: a Case Study with Factor Exposure Constraints, in the EDHEC-Risk Institute Research Insights supplement to Investment & Pensions Europe, Autumn 2014 (p. 5).

  13. Jacobs, B. I., and K. N. Levy, Investing in a Multidimensional Market, Guest Editorial, Financial Analysts Journal, November/December 2014.

  14. Unfortunately this transparency is not so evident for many indices. In March of this year, EDHEC-Risk Institute published a study on index transparency which found a lack of transparency for many indices that were nonetheless popular with investors: Amenc N., and F. Ducoulombier. March 2014. Index Transparency— A Survey of European Investors’ Perceptions, Needs and Expectations. EDHEC-Risk Institute Publication.

  15. In addition to the technical difficulty, we might also consider the usefulness of front running for whoever performs it. Strategies that seek to benefit from corporate events (here it would be the change in the weight of a stock in a smart beta index at the rebalancing date) generally need a considerable number of very regular events in order to make a large number of bets and thereby generate profits without too much risk. The rebalancing of a smart beta index, which is often a quarterly, half-yearly or annual event, is not particularly appropriate for this type of activity. If one is unfortunate enough to make a loss on a rebalancing, one has to wait several months to have a chance to make up for it.



References