Investors often ask us about the expected impact of certain socially responsible (SRI/ESG) constraints on the risk and performance of a portfolio. In our earlier work, we explored several different approaches for constructing a custom SRI/ESG portfolio to help frame that discussion. Here we take a closer look at short- and long-horizon impacts of a popular industry exclusion when using an optimizer to construct portfolios.
Aperio's investment process seeks to create a diversified equity portfolio that looks and behaves like its benchmark while still respecting SRI/ESG objectives. As a result, portfolios that exclude a specific industry will tend to redeploy that capital in other industries whose fundamental and economic risks are similar to the divested industry’s.
In our earlier work, we referred to this approach as the “Optimized Exclusion” strategy in order to contrast it clearly with a naïve reweighting approach (“Cap-Weighted Exclusion”). The latter approach is simple: it does not require a risk model nor an optimizer. While we expect this approach to outperform if the excluded securities underperform, it also may lead to excessive forecast tracking error.
In contrast, we expect our optimized approach to “cushion” the negative performance impact that could occur when an industry that has been excluded outperforms. However, because other industries are not perfect substitutes for the excluded industry, we expect the customized portfolio to have some amount of variation around the benchmark’s performance. We use forecast tracking error as the primary gauge of how much a portfolio’s performance could vary from its benchmark.
For example, investors who excluded the Oil, Gas & Consumable Fuels (OGCF) industry generally experienced underperformance relative to standard (market-capitalization weighted) benchmarks such as the MSCI ACWI in Q2 2018. To illustrate this, we simulated the historical performance of a nontaxable portfolio with an OGCF exclusion.
If we look at the next quarter’s performance (Q3 2018), we see that the situation has reversed and the custom portfolio outperformed its benchmark, erasing all of the underperformance in Q2.
A quarter is a very short time period to evaluate the performance of a strategy. (Research* has shown that checking performance too often increases the likelihood of making suboptimal investment decisions.) So we examined the performance of the same custom portfolio as far back as 2008. The following chart shows the correlation of the active returns of the custom portfolio against the active returns of the OGCF industry.
While there is a fluctuation in the magnitude of the historical correlation, it is pretty clear that a portfolio that excludes the OGCF industry will generally underperform when the OGCF industry outperforms (the correlation is generally negative).
Aperio’s approach to portfolio construction tries to minimize the impact on performance (relative to a benchmark) from an exclusion, but it cannot completely eliminate it as a driver of risk and return. This is reflected in summary risk statistics such as forecast tracking error, which provides a general guide for the expected magnitude of deviation around a benchmark but does not indicate whether the potential deviations will be positive or negative.
Send questions or comments to firstname.lastname@example.org.