Factor investing: good in theory, difficult in practice

Investors are always looking for higher returns at lower cost. This has never been truer than in today’s market environment. Not surprisingly many have responded by embracing “factor investing” – the idea that active managers actually earn most of their excess returns from a common set of risk factors. Constructing a portfolio that systematically captures these factors can potentially deliver a return above that of the broader equity market at a much lower cost than traditional active strategies.

Robertarnott
Robert Arnott, Research Affiliates

High returns at a lower cost – sounds like the perfect solution! But is factor investing really the panacea that equity investors have been waiting for?

We agree that structural sources of excess return do exist and that capturing these can provide a powerful boost to a portfolio. Investors need to be careful, however. Poor implementation can erode a factor’s return benefits and lead to sub-par (and often unexpected) outcomes. Investors should answer three crucial questions:

How rorbust are the factors being targeted?

Since the idea of multi- dimensional risk factors was introduced in the mid-1970s by Ross, Fama, and others, and then extensively analysed by Fama and French in the early 1990’s, the factor world has come a long way. Over the last few years, hundreds of new factors have been “discovered”. A recent study of 600 factors from academic and practitioner research found that 49% produced zero to negative premia out of sample.

Research Affiliates has conducted similar research, testing a wide array of factors to determine how they performed over different time periods, worked across geographies, and stood up to minor definitional variations. The conclusion? Only a handful survived and added value. Among the most reliable sources of long- term outperformance was value. Low volatility, momentum and illiquidity were also significant. Other asserted “factors” were found to be insignificant, including the ever-popular quality.

What does this mean for investors? A lot of so-called factors being marketed today do not stand up to detailed scrutiny.

Are current valuations attractive?

Factor strategies have attracted enormous flows over the last few years and much of this has been based on strong – often “paper” or backtested – historical performance. This sort of performance-chasing behaviour is a trap many investors fall into, and it rarely ends well.

Investors should consider an added nuance when it comes to factor investing: Not all types of outperformance are equal. Value- add can be structural (hence, plausibly a source of future excess return) or situational (a consequence of rising enthusiasm for, and valuation of, the selected factor).

Research Affiliates has found that many commonly used factors delivered excess returns only because they had grown more expensive. In other words, these strategies do not deliver structural excess return, but rather “alpha” is a result of fortuitous timing.

This presents a problem for investors. Even if current valuation levels hold, an expensive starting point reduces future return prospects. Worse still, many factors may be cyclical, and if these overvalued factors mean revert to historic valuation norms investors better watch out!

The takeaway is that, whether buying stocks or factors, price matters! Performance chasing does not pay. In fact, the opposite is usually true: factors that have underperformed in the past are cheaper today and more likely to provide excess returns in the future.

Can the strategy be implemented efficiently?

Even if a factor strategy looks good on paper, implementation is a crucial consideration. Transaction costs can sometimes erode a factor’s return premia. Momentum is a good example of this. With a simple passive implementation, the momentum premium is more than offset by trading costs. This doesn’t mean that momentum should be avoided altogether, but rather investors must realise that skilled trading can sometimes be as important as the factor itself.

Beyond returns, other traits investors should value in their core equity allocations include liquidity, transparency, diversification and stability. Potentially even more important, strategies should be intuitive and understandable – and easy to explain to investment boards. Many factor strategies fall short on these counts, being too often opaque and driven by complex quant models, which results in concentrated, unstable portfolios.

Given these concerns, what should investors take away from factor investing?

We believe structural sources of excess return are available in equity markets, but we are sceptical of many newly “discovered” factors, preferring to focus on those which have been thoroughly researched and debated. This short list includes value, momentum and low volatility. The current high valuation level of many commonly touted factors is also a concern; in particular, low volatility and quality stand out as being expensive. By contrast, value is enjoying its most attractive valuation level since the Tech crash, providing strong future return potential for investors willing to take the slightly uncomfortable position of investing in what has recently underperformed.

Although pure factor strategies can be useful tactically, we are cautious about using them as a core equity allocation. Portfolio construction methods are either inefficient when factors are naively combined in equal measure or overly complicated when quant models and optimisation techniques are employed. The former involves unnecessary implementation costs and inefficiencies as a result of combining factor sleeves. The latter complicated approaches are difficult to understand and even harder to explain to investment boards. Should these approaches really form your core equity portfolio?

PIMCO and Research Affiliates collaborate on a strategy that takes a different approach. Like other factor strategies, we share a goal of capturing structural sources of return in a low-cost, systematic way. Rather than simply combining different factors, however, PIMCO RAE Fundamental embeds value as a key consideration. The core of the strategy is an alternative portfolio-weighting methodology (i.e. fundamental rather than market-cap weighting) that skews toward attractively priced stocks.

Additionally, a number of insights are incorporated based on research into sources of structural return. The outcome is a dynamic exposure to a range of equity factors – value being the most significant, but with quality, momentum and size also playing a role. In our approach, the factor exposures are not the building blocks, but the outcome of the methodology.

We believe our approach provides investors the benefits of factor strategies, but does so in a more intuitive way.

1 Yaron Levi and Ivo Welch. 2014. “Long-Term Capital Budgeting,” SSRN (March 29).
2 Jason Hsu and Vitali Kalesnik. 2014. “Finding Smart Beta in the Factor Zoo,” Research Affiliates Fundamentals (July).
3 Robert D. Arnott, Noah Beck, Vitali Kalesnik, and John West. 2016. “How Can “Smart Beta” Go Horribly Wrong?” Research Affiliates Fundamentals (February).
4 Robert Novy-Marx and Mihail Velikov. 2014. “A Taxonomy of Anomalies and Their Trading Costs,” NBER Working Paper No. 220721 (December).
5 Arnott et al. 2016.

Source: Pimco