In recent years a new category of investment products has emerged. Companies offering these products claim that they are able to systemically outperform a market index in the long run – a promise that has rarely been fulfilled within the financial industry to date. Marketed as “smart beta” funds, they combine active and passive fund management, allegedly delivering the benefits of the two approaches. They are based “on active strategies which are not implemented by portfolio managers, but systematically implemented according to clearly defined rules“. Such rules can either relate to the criteria for selecting securities or to the methods for weighting them. Examples of them are: equal weighting, minimum variance, small cap, high dividends, low P/E ratio, CROCI, etc.
Many of these strategies, however, are not at all new. “Smart Beta” is ultimately a relabelled version of so-called “factor investing”. In the past, this was the collective term used to designate quantitative strategies based on numerically assessable success factors for the stock markets, which were identified and empirically determined by analysts. They were systematically used to put together securities portfolios.
Factor investing was first proposed in an article by Fama and French in 1992. However, it soon met with criticism; already in 1993, Fischer Black argued that its alleged results were very closely linked to “data mining” (Dr. Norbert Häring provided an appropriate definition of “data mining” in the German newspaper Handelsblatt some years ago: Scientists understand “data mining” as “…the analysis of datasets without any theoretical underpinning, with the objective of revealing potentially interesting or useful statistical relationships. Often, however, such correlations occur randomly and do not reflect a causal relationship“). This strategy therefore remained on the periphery until people started marketing it as “smart beta”. Changing its name, however, has not solved its conceptual problems.
William F. Sharpe explains why smart beta will never be able to work in the long run
“When I hear smart beta, it makes me sick,” explained Nobel laureate William F. Sharpe at the CFA Institute Conference last year. The co-inventor of the term “beta” as it is applied in the models of capital market theory rejects “smart beta” for several reasons:
- The term is ambiguous and confusing. Smart beta is actually meant to generate alpha – a risk-adjusted outperformance – by selecting securities on the basis of superior selection criteria. This has nothing to do with beta, which is a measure of risk and therefore only measures outperformance related to taking more risks.
- The assumption behind smart beta is that there are significant inefficiencies on the market, which do not change on a long-term basis. But financial markets learn quickly and this assumption is therefore completely absurd.
- Smart beta can only ever work for a minority of investors. Once a large number of investors start to invest in a smart beta approach, a market process is activated that eliminates the underlying inefficiency.
- Empirical evidence for the functioning of factor investing is largely limited to time periods of several years and to specific markets. There are no reliable scientific findings showing that specific factors function in the long run at various periods on different markets.
Alongside pointing out a rather strange usage of terminology, Sharpe’s considerations reveal that the concept of outperforming the market with smart beta is ultimately grounded in incorrect generalisations about isolated phenomena. In fact, a long-term investment strategy cannot be based on such a concept.
Quantitative investing is not as simple as it might appear at first glance
Although quantitative investment strategies have a tremendous advantage, they are also characterised by three major limitations.
Their main benefit is that investment decisions can be made without emotions. Indeed, emotions cause investment managers to repeatedly make flawed investment decisions, often implementing their own strategies inconsistently, because they are perhaps opportunistic, greedy or excessively nervous. In addition, they tend to make superfluous transactions in order to convince themselves and others that they are sufficiently active.
“Emotional discipline” is therefore a decisive success factor for active investment decisions, as put forward by Benjamin Graham, the father of modern financial analysis, already 80 years ago. Investors lacking in such discipline can be tempted to make bad decisions, which in turn may reduce the value of assets so significantly that the investor would have been better off putting his money into a passive, purely index-oriented product. In reality, there are only very few investors – for example Warren Buffett – who are able to exercise sufficient emotional discipline and outperform market indices in the long term.
Quantitative strategies such as smart beta seem to be a solution to this problem. Like active investors, they claim to identify and exploit market inefficiencies, but they are not affected by errors resulting from emotional decisions. In reality, however, it is not quite as straightforward as it sounds. Indeed, they have three basic sources of weakness which need to be considered.
- They are either based on past experience, and therefore implicitly assume that patterns from the past will be repeated in the future, or on forecasts.
- They rely on the analysis of extensive data, with the quality of data as a critical factor, or on the accuracy of forecasts.
- They only function to a limited extent with criteria that are not general knowledge.
Quant investors who have been successful in the long run, such as James Simons from the Renaissance Technologies hedge fund, have therefore set up extensive teams of analysts tasked with a) permanently monitoring market trends and, if necessary, adapting portfolio parameters at very short notice as well as b) ensuring the quality of data. Furthermore, they maintain the highest standards of confidentiality regarding the criteria they identify. Smart beta, on the other hand, does not permanently review its strategy and is based on transparent criteria. This seems very naive to me.
The term “smart beta” is a typical example of reversification
The British author John Lanchester published the book “How to Speak Money” in 2014. In this publication, he explained the language of the financial world in a way that is understood by the general public. While doing so, he came across a phenomenon that makes many financial terms particularly difficult to understand: so-called “reversification”. The names of financial instruments sometimes go beyond merely embellishing their characteristics – they occasionally mean the exact opposite to their true sense. Lanchester gives the term “hedge fund” as an example. The word “hedging” is normally used to describe a strategy that provides protection. Based on this, one would therefore expect hedge funds to be very safe products. Due to their sometimes highly speculative strategies and legal forms, however, hedge funds are particularly unsafe in practice.
Smart beta is a blatant example of reversification. Firstly, because the product is not about beta, but about alpha. And, secondly, because it is not particularly smart to believe that a one-time success can be repeated on an ongoing basis by implementing relatively simple factor investing strategies. On the contrary, it is even very stupid to assume that the stock or bond market indices can be sustainably outperformed, thus generating alpha, by pursuing relatively simple quantitative rules. The real world does not work quite as easily as that, particularly not in the financial markets.
Smart beta is the latest excuse for pro-cyclical behaviour
Factor investing does not have to be bad. If it is used, for instance, to put together a portfolio that has a different risk-return profile to a market index and, as a result, better meets specific investor requirements, then it is a useful strategy. But one should not expect it to outperform a representative market index in the long term. However, that is exactly what the proponents of smart beta are promising. Longitudinal empirical studies show that smart beta strategies work very well during specific phases of the stock market cycle and fail during other periods. In particular, when investors pursue a strategy that has resulted in very good returns in recent years, it is a recipe for failure. However, many consultants are currently recommending this approach and an alarmingly high number of investors are implementing it, above all in the institutional sphere. Smart beta is therefore a new excuse for following the herd and making pro-cyclical investment decisions.
Our complex world is subject to different economic cycles and is changing permanently; financial markets adapt rapidly. Successful quantitative investors know this, keep their strategies secret as long as possible and adjust their models on an ongoing basis. Smart beta, on the other hand, ignores the relatively banal insight that, if the world is changing, quantifiable success factors are also permanently changing, in particular when they become public knowledge. Anybody who pursues outdated success criteria will end up unwisely destroying the value of their assets in the long run – in other words, “very stupid no-alpha”.
GOod post. All valid points about backtesting, and practicalities of implementing quant strats, but as any investor knows, nothing in the market is for certain, but a well established factor in low cost, transparent fund, is the best bet we can make at beating the market.