Allocation to factors has become increasingly popular in recent years, but practical implementation remains a puzzle for many investors. Avoiding unintended factor biases often ranks amongst their top concerns.
Analyzing and attributing the origin of past portfolio returns is an essential part of the investment process. This is obvious for investors using a traditional fundamental country and business sector selection approach, but it is also the case for those allocating systematically to factor premiums.
However, while this kind of analysis appears relatively easy to perform for fundamental investors, accessing a portfolio’s exposure to factor premiums and calculating the performance of individual factors is much less straight forward, especially for newcomers to the factor investing arena.
This has far-reaching implications as undetected and unintended factor biases can seriously affect performance. A FTSE Russell survey carried out in 2016 actually suggested that avoiding unintended factor biases ranked second among investors’ concerns, when considering factor allocation.
As we indicated in the first article of this series dedicated to the major challenges investors face when considering factor investing, the fact that it is virtually impossible to effectively time the different premiums makes a strong case for broad diversification. Unfortunately, this is easier said than done because factors can also clash with each other.
These potential clashes are one of the reasons why products based on common smart beta indices often prove inefficient when it comes to harvesting factor premiums. For example, most generic value strategies do not avoid stocks that are cheap for a reason, such as those of financially distressed companies. This is a typical case of the value factor clashing with quality.
‘Investors need to be able to access their exposure to different factors’
As a result, an investor using this kind generic value investment product will also often end up being negatively exposed to quality, without realizing it. In a 2015 white paper David Blitz, head of Quantitative Equity Research at Robeco, and Matthias Hanauer, Quant Equity Selection Researcher, argued that the performance of a particular strategy explicitly targeting one specific factor depended heavily on the implicit exposures towards other premiums.
They also found that the return difference between ‘bad’ and ‘good’ strategies could amount to as much as 5%-7%, depending on secondary exposure to other factors. To measure this, they simulated four generic global strategies targeting value and momentum – two of which were good and two of which were bad. The good strategies specifically avoided factor clashes, while the bad strategies had strong negative exposure to other risk factors.
At Robeco, we exploit four factors that have proven their long-term performance potential: value, momentum, low volatility and quality. And while combinations of generic single factor products often result in opposing premium exposures that partly or totally neutralize each other, we make sure that all our strategies efficiently combine factors to avoid unintended biases. This is true both for equities and fixed income(1).
However, it is also important to note that existing portfolios may already be heavily factor-biased without investors realizing this. When considering a new factor-oriented strategy, investors must therefore make sure they are fully aware of their exact exposure to different premiums. For example, an existing significant value tilt in a client’s holdings may mean that their optimal factor strategy needs to target a lower explicit weight for this specific factor.
When considering allocation to factors, the first task investors face is determining which premiums they intend to exploit and defining them precisely. Then, they also need to be able to evaluate their exposure to each individual factor and the corresponding performance.
To make sure our clients know exactly which factor premiums they are exposed to and how these have performed in the past, we have developed an innovative performance attribution model, the Factor Exposure Monitor(2). This comprehensive tool fully allocates performance to the set of selected factors used in our quantitative equity strategies.
Of course, there are many other tools available in the market for quantitative factor performance attribution, and the body of academic literature on the subject is extensive. Our proprietary approach, however, stands out thanks to its relative simplicity and the fact that the applied factor definitions are consistent with those used in our own quantitative investment processes.
Many non-Robeco models that are used for a broad range of different strategies are rather complex, something that can also make the results more difficult to interpret and understand. Furthermore, the factor definitions used are not completely in line with those we apply in our own strategies.
Our Factor Exposure Monitor provides a concise overview of the contribution of each factor to the portfolio’s return. It can be used in conjunction with other tools in order, for example, to evaluate allocation and selection effects resulting from sector or country positions.
(1) Read our previously published article about the way we efficiently combine factors for bonds.
(2) Read more about our Factor Exposure Monitor in our case studies book.
BY CLICKING ON “I AGREE”, I DECLARE I AM A WHOLESALE CLIENT AS DEFINED IN THE CORPORATIONS ACT 2001.
What is a Wholesale Client?
A person or entity is a “wholesale client” if they satisfy the requirements of section 761G of the Corporations Act.
This commonly includes a person or entity: