Active management was based on the idea that superior investment skill can deliver returns above the market return.   Investors would select which market they wanted exposure to and then allocate funds to active managers they believed could augment this market return with alpha. 

Mandates were given and mutual funds compared within peer groups classified by geographies after CAPM (1976), and further sub-divided by style betas after Fama-French (1992, 1997).  The former assumed that only systematic risk mattered, everything else being alpha, and the latter found that active strategies based on the traditional style exposures of value, growth and size, in the long term, consistently provided abnormal returns above the market (beta) return—too systematically, in fact, to be called alpha.

In the early 1990s, Morningstar implemented a very popular example of this allocation strategy with its three-by-three style boxes for mutual fund classification.  The benchmark industry followed suit, delivering indices designed to capture exposure to these style betas.  Active managers were then pigeon-holed into mandates representing their stated investment style, such as large-cap value, or small-cap growth, to name two of the most popular.  No longer could under-performers hide their lack of stock-picking skills behind the returns of a certain style beta. 

With both the asset and style allocation decisions now being made by the investor, the only apparent avenue left for active managers to add value was through stock selection. 

Or so we thought.

The search for ways to generate more consistent returns brought quantitative investment processes front and centre.  Armed with multi-factor models, a new generation of portfolio managers used computers to uncover new sources of systematic risk, along with a process to efficiently realign their portfolios along those dimensions—all while staying inside the box allocated to them by the investor.  They not only created a new source of beta returns, but a repeatable way to capture those returns.

These so-called smart betas include characteristics such as momentum, leverage, volatility, and liquidity.  The turnover associated with harvesting returns linked to these dimensions is much higher than it is for the traditional ones of value, growth, or size, with which these new signals were often combined.  Smart betas require a more sophisticated, scalable, and disciplined investment process. The ensuing success of quant management is well documented, leading the media to dub the last decade the Quant Era.  BGI, the largest of the quant firms, made the cover of the Bloomberg Markets magazine in February 2007, for an article titled Empire of the Quant, when they attained the top spot on the manager rankings by assets under management.

While quants became a victim of their own success soon after this article was published, the sources of return they uncovered are here to stay.  Most investors demand that they be accounted for as a systematic sources rather than stock-selection skill when evaluating managers. 

Three recent trends have combined to drive demand for better ways to benchmark and track smart beta returns. First, asset-to-asset correlations have been steadily increasing in all equity markets in recent years. When correlations are high, common characteristics, including smart betas, across all assets in the market explain an even larger fraction of returns.

Second, the current environment of rapidly changing volatility has become the new norm and managers need new tools to hedge these esoteric dimensions if they are to successfully navigate in this new reality. Third, advances in portfolio construction technology now make it possible to steer away from a traditional cap-weighting methodology toward an optimized weighted scheme when constructing an index.  This new optimization-based methodology allows benchmark providers to design and maintain smart beta indices with the following essential properties:

    * Return purity via direct management of the index’s tracking error to the return performance of the targeted characteristic

    * Low cost replication via constraints on turnover, transaction size, and the number of names held by the index portfolio

    * Neutrality via constraints on exposure to non-targeted characteristics

Smart beta indices can be used to capture the above-market performance of any smart beta  theme or characteristic such as leverage, volatility, liquidity, or momentum.

During times of above-average market volatility, for example, whether caused by a systemic shock such as the global financial crisis or a transient event such as the recent earthquake in Japan, investors may want to seek protection from a large drawdown caused by panic or speculative short-selling.  Both liquidity and volatility would be characteristics to hedge in such circumstances. Investors could hedge their liquidity and volatility exposures via a smart beta ETF launched on the smart beta volatility or liquidity index.

Conversely, these may be characteristics a hedge fund may want to leverage during such events.  For those speculators, an ETF tracking a net (long/short) smart beta index would provide a leveraged exposure to volatility or liquidity, while avoiding unnecessary exposure to all other characteristics.  The purity of smart beta indices is one of the key value propositions here. The decomposition of total return into beta plus smart beta plus alpha can help investors analyze market conditions to decide which type of strategy will do well in the current environment.

Take emerging markets as a recent example.  In late 2010 and early 2011, risk models showed that returns in emerging markets were being driven by company-specific news; alpha one might say.  And, indeed, January and February were good months for stock pickers in emerging markets.   Strategies based on stock picking were excellent performers. When the Japan crisis hit, everything flipped back to an environment where beta and smart betas were dominating the returns.  In this environment the search for alpha must take a backseat to efficient beta and smart beta management.

The above two examples highlight yet another use for smart betas.  Just like beta became a source of return for those managers prophesising to be able to time the market, style rotation strategies is predicated on the notion of smart beta timing. 

Multi-factor models represented an innovation from the one-factor model world of CAPM where everything above the market return was alpha.  In a multi-factor world, you have beta, smart beta, and alpha.  Smart beta indices and their related ETFs represent a new frontier in passive investing, offering growth opportunities for ETF providers.  As a hybrid of passive and active investing, smart beta ETFs are likely to attract a new generation of investors eager to take greater control of their investments in a cost-effective manner.

Olivier d”Assier is Axioma’s managing director for Europe and Asia