How to Build Eco-Friendly Portfolios (Without Losing Returns)
(7 min read)
(7 min read)
• A new BIS working paper examines how decreasing carbon footprints impacts a passive investor’s portfolio performance.
• By excluding only the most-polluting firms from a portfolio of MSCI All Country World Index constituents, investors can massively reduce their carbon footprints.
• Most importantly, though, this emissions reduction does not sacrifice expected returns and can preserve existing sectoral and regional exposures.
This article is only available to Macro Hive subscribers. Sign-up to receive world-class macro analysis with a daily curated newsletter, podcast, original content from award-winning researchers, cross market strategy, equity insights, trade ideas, crypto flow frameworks, academic paper summaries, explanation and analysis of market-moving events, community investor chat room, and more.
We recently explained how decarbonisation selling pressure encourages highly carbon-intensive firms to reduce emissions. In this way, a critical mass of investors divesting from heavy polluters can help limit greenhouse gasses and mitigate climate change. But does reallocating investment towards greener corporates negatively affect financial performance?
No, according to a new BIS working paper. Investors can significantly reduce a portfolio’s carbon footprint by excluding just a small fraction of the worst polluters globally. They can then reallocate proceeds to retain the portfolio’s sectoral and regional exposures. Such a reallocation strategy would limit divestment from, say, EM countries, where corporates typically pollute more. Overall, the authors find:
Opinion varies on how environmental objectives impact portfolio performance. Theoretically, if the market is driven by investors using an environmental score only as an indicator of high future performance, high-environmental-score (HES) stocks should deliver higher expected returns. However, if investors enjoy holding HES stocks and are willing to pay a premium for them, expected returns will be lower.
Empirically, evidence of both exists. Investors are already demanding compensation for their exposure to carbon emission risk. But research also shows reducing the carbon footprint of a portfolio is associated with stronger future profitability and positive stock returns in a global universe of stocks. Another paper documents that a long carbon-efficient, short carbon-inefficient strategy earns positive abnormal profits.
A popular investment strategy among institutional investors that includes environmental objectives is exclusionary screening. Exclusion restrictions only require the omission of a few highly polluting firms from portfolios and can easily hedge against climate risk. The seminal work of Andersson et al. (2016) shows how such strategies can have little to no damage to portfolio performance. The BIS paper focuses specifically on this type of strategy.
The authors collect carbon emissions and emission intensity (emissions per million dollars of revenue) data from S&P Trucost on all firms that constitute the MSCI All Country World Index (ACWI). As of 2019, this amounts to 3,000 firms, classifiable into four large zones: emerging countries, Europe, North America, and the Pacific. They also collect financial and accounting data on all those firms from Refinitiv.
For context, Chart 1 gives the carbon intensity (in tCO2e/m$) of all firms listed on the ACWI. It is extremely skewed to the right, with a few firms generating very high carbon intensity levels. For example, the top 25% of firms contribute 74.4% of total carbon intensity. The top 1% contribute 16.8% and, on average, emit 12,470 tCO2e/m$ per year. This rose 48% between 2012 and 2019, against an overall reduction of 26% in the whole sample.
The authors compute three metrics of a portfolio’s carbon quality. The first is the weighted-average carbon intensity, which measures the portfolio’s exposure to carbon-intensive companies. It is simply the sum of the portfolio weights (dollars invested in firm i divided by dollar value of portfolio), multiplied by the carbon intensity of firm i (carbon emissions of firm i divided by revenue of firm i).
The second metric is the portfolio’s carbon intensity. It includes the same carbon intensity measure as the first metric, but rather than weights, it is adjusted for company size. Therefore, it is a measure of how carbon efficient the portfolio is in generating revenue through its investee companies. It is dollars invested in firm i divided by the market cap of firm i.
The third metric is the portfolio’s carbon footprint. It measures a portfolio’s total carbon emissions normalised by its market value. It expresses the amount of annual carbon emissions that can be allocated to the investor per million dollars invested in the portfolio. Note, all three metrics can be compared with those of the MSCI index benchmark.
The authors determine the exclusion restrictions via a value-weighted approach. Each year, they sort all firms by increasing carbon intensity and sum their relative market caps. They then exclude firms with a carbon intensity above a certain threshold, so that this group of firms represents 1%, 10% and 25% of the portfolio’s market value.
Next, they consider three different strategies for reinvesting the proceeds of selling the excluded firms. The first, which they call proportionate reinvestment, reinvests the proceeds in all the remaining firms proportionately to their weight in the benchmark portfolio. They call this portfolio the business-as-usual (BAU) benchmark.
The second, symmetric reinvestment, reallocates the proceeds by investing in firms with the lowest overall carbon intensity in the BAU benchmark. That is, they double the weight of these firms in the new portfolio. Lastly, regional/sectoral reinvestment reinvests in the least-polluting firms in the region/sector of the excluded firms.
The last reinvesting strategy works as follows. If the most-polluting European utilities represent 1% of the portfolio value, this 1% is reinvested in the least-polluting firms among European utilities that add up to 1% of the original portfolio’s market value. This means investors keep the same regional and sectoral exposure, and firms from emerging countries are not penalised as an asset class.
When the authors exclude the most-polluting firms until they represent 1% of the market value of the portfolio and use the proportionate reinvestment strategy, the portfolio’s weighted-average carbon intensity, carbon intensity and carbon footprint fall 16.9%, 20.3% and 15.6%, respectively (Chart 2, Panels A-C). Furthermore, the carbon footprint falls 48.1% and 70% if the threshold is increased to 10% and 25% of the market value.
To calculate financial performance, the authors assume investment occurs at the end of each year based on carbon emitted in that year. They measure performance in the subsequent year, including dividend payments. Excluding the worst polluters, whether to a market value of 1%, 10% or 25% of the portfolio, does not reduce the ex-post performance relative to the ‘do-nothing’ benchmark (Chart 2, Panel D).
If instead the authors reinvest the proceeds according to symmetric reinvestment, the portfolio’s improvement in carbon quality is even greater. The financial performance is also barely affected. This, according to the authors, is because most excluded firms are utilities and materials firms in emerging countries, which have had relatively low performance on average.
Lastly, regional and sectoral reinvestment has a similar carbon effect and financial performance to the first reinvestment strategy. However, the authors emphasise the benefits of using this third strategy. Without maintaining the same regional and sectoral exposures, the portfolio will pivot significantly towards less carbon-intensive sectors and more developed economies, with no or little risk-adjusted return benefits.
Finally, the authors build a benchmark portfolio that converges to net zero emissions. This requires that CO2 emissions decrease between 9% and 10% per annum on average over the next 30 years. So, they assume a passive investor decarbonises their portfolio by 10% per year by excluding the most-polluting firms. They reinvest proceeds across regions and sectors so that the portfolio’s overall structure remains as close as possible to the BAU benchmark, but with a lower carbon footprint.
The authors then evaluate the impact of this approach on the last 10 years of the sample (2010-2019). They find a 10% per year reduction in emissions would reduce the total carbon emissions by firms in the portfolio by 64%. The impact of this reduction on returns is nil, so the only cost of this approach is in the tracking error, which remains very low.
Decarbonising portfolios can encourage highly polluting firms to reduce emissions if enough investors do it. And this research shows divesting from such firms need not impact the financial performance of portfolios. A passive investor in 2010 could have reduced their portfolio’s carbon footprint by 64% over the last 10 years with no impact on overall returns. All it requires is removing a small fraction of the highest-polluting firms in a portfolio to other lower-polluting ones!
Already have a Macro Hive Prime account? Log in