In deciding whether to divest assets in carbon industries, investors face compelling arguments on both sides. From the advocates of divestment, investors hear about the serious environmental damage already incurred and the benefit from taking a public stance on a critical ethical issue.
This publication has carried a number of guest feature commentaries around subjects such as impact investing, and investing with a view to bring about social, environmental and other outcomes that are not always easily measurable in dollars. There remains controversy around the subject - can impact or ESG-based investing match or beat returns from more conventional approaches? This publication does not want to get into the arguments about whether the theories that drive certain ideas are correct in all respects, such as human-caused global warming, but it must acknowledge that many wealthy investors do worry about issues such as global warming, and the need to reduce carbon dioxide emissions and other "Greenhouse gasses".
This article is by Aperio. This is an investment management firm managing approximately $20 Billion in public equity portfolios for ultra-high net-worth investors, foundations, and endowments through financial intermediaries such as independent advisors and family offices. The firm’s major product lines include active tax management and socially responsive indexing/ESG. The firm’s specialty is designing and managing “bespoke” portfolios that can track index benchmarks or deliver targeted risk, factor, geographic, or industry exposures customized to a client’s specific tax situation, values, and/or desired economic exposure.
The article is written by Patrick Geddes; Lisa Goldberg, Robert Tymoczko and Michael Branch.
The editors of this publication want to thank the firm for its article and invite readers to respond. This news service does not necessarily endorse all the views expressed. Email firstname.lastname@example.org
Since we published our last analysis of a US carbon-free tracking portfolio in 2015, the fossil fuel divestment movement has continued to develop across the globe. This version includes updated data and also new analysis of markets outside the United States, including Australia, Canada, and global markets. Within the US, the divestment campaign remains active among campuses, foundations, endowment board rooms, and family offices. More than 100 foundations and family funds have signed a pledge to divest in as part of “Divest-Invest Philanthropy,” (1) while many more foundations are considering their options. These include doing nothing, divesting from just the coal industry, avoiding entire fossil fuel industries, or becoming active shareholders to influence corporate behavior (especially around disclosure).
When the idea of fossil fuel screening is raised, the first thing an endowment committee, foundation board, or private investor wants to know is whether screening will impose a penalty. While there is no definitive answer, the often-presumed assumption of a return penalty is not consistently borne out by research. In fact, results from a wide range of studies on social and environmental screening do not provide a consensus on whether there has been a return penalty or benefit from carbon screening. (2)
Looking forward, there are compelling scenarios investors can imagine that lead to outperformance of carbon industries and others that lead to underperformance. As an example of the former, large-scale divestments could lower demand for securities in carbon industries, artificially lowering prices. According to an article in the Journal of Financial Economics published in 2009, investors willing to own carbon industry securities could benefit. (3) In the other direction, government-imposed carbon emission controls could lead to stranded assets, permanently eroding profits of companies of carbon-centric companies. (4) The data does not really support either supposition in any clear way, although as humans we naturally cite the version of the future that best serves our goals and prejudices.
Lacking a consistent story about the future return impact of divesting in carbon, we shift the focus onto the impact of carbon screening on portfolio risk. Specifically, we look at the tracking error, or variability of the return difference between an index and a screened portfolio, to measure the impact of exclusion. (5) A lower tracking error means that the screened portfolio replicates the index returns more faithfully (i.e., there is less risk of deviating from the benchmark). Lower deviation implies that the return of the tracking portfolio is unlikely to vary dramatically (positively or negatively) from the return to the index. (6) We examine the magnitude of this deviation later in this paper.
Our study focuses on hypothetical equity portfolios obtained by excluding carbon industries from standard market indices in Australia, Canada, and the US, as well as a global index. In each market, we exclude from the universe the Oil, Gas and Consumable Fuels industry (7) from the broadest available index (“Tracking Portfolios”). (8) Then hypothetical portfolios are created using an optimizer to track that respective index as tightly as possible, subject to those industry exclusions, and, of course, the accuracy of the model.
The indices used to create hypothetical carbon-free Tracking Portfolios in each market are shown in Table 1.
Source: MSCI Indices are trademarks of MSCI Inc.; Russell Investments is the source and owner of the trademarks, service marks, and copyrights related to the Russell Indexes; Standard & Poor's S&P Indices are registered trademarks of Standard & Poor's Financial Services LLC.
Tracking error on hypothetical portfolios
Table 2 shows the tracking errors for the hypothetical carbon-free Tracking Portfolios against their respective indices in the four markets over back-tested periods ranging from 11 to 25 years. (9) The Tracking Portfolios were rebalanced quarterly. (10) In the Australian, US, and global markets, the tracking errors were less than 100 basis points. The larger tracking error of 2.91 per cent in Canada reflects the substantial exposure to Oil, Gas and Consumable Fuels in that country’s stock market. That exposure averaged 20.60 per cent over the 13-year study period. By comparison, the average exposures in the Australian, US, and global markets were 5.10 per cent, 6.80 per cent, and 8.00 per cent over their respective study periods.
Table 2: Market Indices
Source: Aperio Group LLC. Past performance is not a guarantee of future returns. Please refer to important disclosures at the end of this paper.
To put the tracking errors of the carbon-free portfolios into perspective, consider the active risk of 5 per cent taken by the typical institutional investor. (11) That dwarfs the tracking errors of the carbon-free portfolios, even in Canada.
Building the tracking portfolios
It may come as a surprise that the returns to a carbon-free portfolio can closely mimic the returns to a broad market index. The explanation lies in the two-step process used to build a carbon-free Tracking Portfolio. In the initial step, Oil, Gas and Consumable Fuels industry stocks are excluded from the index. In the second step, the remaining stocks are re-weighted so that the portfolio can track the index as closely as possible.
The re-weighting process takes into account the fundamental risk characteristics of the excluded assets, such as their size, valuation ratios, leverage, and liquidity. (12) A quantitative optimization is used to match the risk characteristics of the Tracking Portfolio as closely as possible to the risk characteristics of the index. (13)
Summary statistics for the four market indices and their Tracking Portfolios are shown in Table 3. Note that the annualized returns are higher for the carbon-free Tracking Portfolios than for the indices. The difference is 0.85 per cent in Canada, although it is much smaller in the other markets. Turning from return to risk, the realized total volatility (annualized standard deviation) of the Tracking Portfolios is in line with the broad market indices.
Table 3: Summary Statistics*
Source: Aperio Group LLC. The statistics above are shown for illustrative purposes only. They are based entirely on back-tested portfolios and are hypothetical. Each strategy is for the time period indicated, based on data availability. Performance figures shown reflect the reinvestment of dividends and other earnings, are gross of fees, and do not include transaction costs. Past performance is not a guarantee of future returns. Please refer to important disclosures at the end of this paper and refer to Appendix II for hypothetical results including management fees and transaction costs over several historical periods.
The hypothetical returns for Tracking Portfolios should in no way be construed to imply that divestment leads to better performance. They show only that over the time periods analyzed, this version of divestment just happened to play out that way. While doomsayers claiming a return penalty to divestment may resent the fact that there was no such penalty over the period, advocates of divestment may want to avoid promising any grand return benefit in the future based on the stranded-asset hypothesis that supposes carbon assets to be overvalued. An inability to predict returns leaves investors with managing risk, something we can all control to a much greater extent.
Active sector weights
As funds normally allocated to carbon industries need to be re-allocated in the carbon-free Tracking Portfolios, it is inevitable that some sectors become overweighted. Table 4 shows the active weights of the carbon-free Tracking Portfolios relative to their indices. On average over the study periods, the carbon-free Tracking Portfolios were overweight in the Utility sector by 1.26 per cent in Australia, 3.67 per cent in Canada, 3.08 per cent in the US, and 2.75 per cent in the global market. In other words, the investment shifted from carbon-heavy energy producers to energy consumers. For some investors concerned with climate change, this transfer may not be acceptable from a values perspective. The effect can be mitigated or eliminated by constraining utility stocks too, but the likely result would be an increase in tracking error.