Well before factor investing became popular in the late 2000s, many investors were already exploiting individual factor premiums. Value strategies are a good example. The value effect is the empirically-documented tendency of inexpensive securities to achieve above-market returns relative to their intrinsic value as measured for example by the book-to-price ratio of a company.

For decades, prominent investors have advocated buying securities trading below their intrinsic value and many traditional active managers have been offering so-called value strategies. As early as the 1930s,¹ Benjamin Graham and David Dodd of Columbia Business School advocated investing in undervalued stocks. Later on, Warren Buffett became famous for his very successful investment philosophy largely based on this same principle.

In this context, many investors have recently turned to factor investing as a systematic and cost-efficient way to achieve exposure to a particular factor premium, such as value or momentum, or to a specific set of factors. In fact, a recent FTSE Russell survey of asset owners found that getting specific factor exposure ranked fifth among the top investment goals that led them to consider factor-based strategies.

As discussed in a previous article in this series, decades of academic research have shown that strategies focusing on a handful of well-vetted factor premiums deliver statistically and economically significant abnormal returns. These factor premiums are distinct phenomena, which exist beside one another, and have been identified across markets and asset classes.²

Targeting different factor premiums will therefore normally lead to different investment outcomes. Figure 1 provides an illustration of this. It shows the Sharpe and information ratios generated with four generic single-factor strategies – based on popular equity indices like the MSCI World Value Weighted index, the MSCI World Momentum index, the MSCI World Minimum Volatility index or the MSCI World Quality index – investing global equity markets, over the period from June 1988 to December 2015.

The differences between different single-factor strategies are not only visible over long periods of time. In the shorter term, for instance, factor premiums can experience periods of underperformance or outperformance, relative to the market as well as other factor premiums. Such periods can continue uninterrupted for several years.

All these findings illustrate how different factors perform independently over time. It therefore makes sense to consider exposures individually and to allocate to each individual factor, depending on the needs and priorities of each investor, using single-factor strategies. For example, investors that have a clear preference for income, can allocate more to value or low volatility strategies, which typically deliver high dividends. Other investors may prefer to limit turnover, and therefore choose not to allocate to momentum which tends to lead to higher portfolio rotation.

But while monitoring and adjusting individual factor exposures depending on the strategic interests of each investor may look like a simple task on paper, there is much more to it in practice. The body of academic literature on the subject is extensive and there are many products available in the market for quantitative factor exposure measurement and performance attribution, including Robeco’s own tool.

Nevertheless, accurately measuring exposures to these factors often remains a challenge, in particular for the less sophisticated investors, who typically lack the necessary resources. Many academics and practitioners have warned about the dangers of poorly designed or inappropriate models.³

Without the relevant measurement tools, investors may, for example, confuse systematic exposure to one particular factor with alpha generated by an active portfolio manager. This explains why asset managers increasingly offer (multi-factor) solutions that provide exposure to a preset blend of factors.

Another important pitfall for those looking for specific factor exposures has to do with the way factors interact and in some cases clash with each other. Generic single-factor strategies usually ignore these interactions and therefore provide suboptimal factor exposures, resulting in, e.g. a value strategy that has very negative momentum exposures.⁴ This underscores the need for efficient factor strategies that use enhanced factor definitions that prevent negative exposures to other proven factors.