The Unintended Consequences of Factor-Based Investing

New research shows that some funds that use a factor-based construction process may have over- or under-exposure to industries, sectors, countries and other attributes relative to a market-cap-weighted index.

Factor investing (attributes, such as value, that explain how an asset contributes to the expected return and risk of a portfolio) often results in unintentional exposures to industries, countries and even other factors. Some incidental exposures are persistent (such as value’s negative exposure to momentum), while some vary widely over time. Since such unintentional exposures can lead to significant differences in tracking variance and can also enhance or detract from the performance of factor-based investment strategies, investors should understand how neutralizing those incidental exposures will impact performance. For example, while sector neutrality typically reduces the factor’s exposure relative to the total market cap-weighted index, it also reduces tracking variance.

Today, some indexes, such as the Russell Indexes, do not seek sector neutrality, while others, such as the MSCI value weighted Indexes, do. As an example of how factor-based strategies can lead to unintentional exposures to industries, as of August 16, 2021, the Russell 1000 ETF (IWB) had 25% exposure to the technology sector and 14% exposure to financial services, while the Russell 1000 Value ETF (IWD) had exposure of just 10% to technology and 21% to financial services.

John Scruggs contributes to the factor-based investment literature with his study, “Does Neutralizing Style Factors Help or Hurt?,” published in the August 2021 issue of The Journal of Investing, in which he disentangled the effects of global style, industry and region factors. His data sample included the universe of global, large-capitalization stocks (the FTSE All-World Index) over the period 1995-2019. Industry neutralization was accomplished by decomposing the style factor exposures into interindustry components (the result of tilting a portfolio’s holdings toward industries with high factor exposures whereby the stocks within each industry have benchmark weights) and intra-industry components (the result of tilting a portfolio’s holdings toward stocks with high factor exposures while remaining industry neutral). The 23 countries in the investment universe were mapped into one of four regions (North America, Europe, Asia-Pacific ex-Japan, and Japan), and five types of style factors were analyzed (size, momentum, reversal, value and profitability). For the value factor, four metrics were analyzed: earnings-to-price (E/P), book-to-price (B/P), cash flow-to-price (C/P) and forward earnings-to-price (forward E/P). For profitability, ROE and operating profitability were analyzed.