The Research
To better understand the likely impact of adopting an ethical screening process we undertook a review of the already significant body of research on the subject.
Across more than 60 peer-reviewed studies covering 47 years and 3,000 SRI-screened Funds, we found that applying an ethical filter to portfolio selection processes created an average drag on returns of approximately 0.37%.
This includes close to 40% of reports that found no material difference in returns; excluding these reports the average cost was 0.79% per annum. We also found evidence of a significant negative skew, though portfolio volatility tended to be lower. This provides a somewhat false impression that SRI strategies perform roughly the same as unconstrained strategies on a simple risk-adjusted basis.
To better understand the likely impact of adopting an ethical screening process we undertook a review of the already significant body of research on the subject.
Across more than 60 peer-reviewed studies covering 47 years and 3,000 SRI-screened Funds, we found that applying an ethical filter to portfolio selection processes created an average drag on returns of approximately 0.37%.
This includes close to 40% of reports that found no material difference in returns; excluding these reports the average cost was 0.79% per annum. We also found evidence of a significant negative skew, though portfolio volatility tended to be lower. This provides a somewhat false impression that SRI strategies perform roughly the same as unconstrained strategies on a simple risk-adjusted basis.
As to be expected from such a broad review, we found many conflicting results and different methodologies, especially from academic research that focused on smaller samples and shorter periods of time. Of research that found SRI enhanced returns, the empirical data used for each review took in an average of 26% fewer Funds and the time period analysed was 46% shorter. This is not surprising as we found SRI strategies tended to be characterized by long periods of underperformance, followed by short bursts of outperformance in times of market crisis.
On a related note, we also observed that the relatively few research papers that found SRI enhanced return tend to be cited much more often than the papers showing a link between SRI and reduced returns.
This is, of course, to be expected given that many of the authors and firms citing or reproducing this data are themselves in the business of manufacturing or marketing "ethical" investment products. This point only warrants attention to help illustrate that the facts, as they are presented to the casual observer, are likely to be manipulated to some extent by the author. For a truly fair and unbiased review, we need to treat all data with a healthy dose of skepticism.
Despite a conspicuous lack of consensus in defining what is (and isn't) an ‘ethical' investment, we did nevertheless extract a number of key findings of the way in which ethical screening processes affect returns:
This is, of course, to be expected given that many of the authors and firms citing or reproducing this data are themselves in the business of manufacturing or marketing "ethical" investment products. This point only warrants attention to help illustrate that the facts, as they are presented to the casual observer, are likely to be manipulated to some extent by the author. For a truly fair and unbiased review, we need to treat all data with a healthy dose of skepticism.
Despite a conspicuous lack of consensus in defining what is (and isn't) an ‘ethical' investment, we did nevertheless extract a number of key findings of the way in which ethical screening processes affect returns:
- The strictness of the SRI-screening process and parameters has a direct bearing on the costs incurred; the stricter the requirements the higher the cost.
- SRI portfolios slightly outperform the market benchmark in times of market crisis, however, we should point out that this is characteristic of actively managed strategies, more generally.
- Despite reducing the number of assets available for investment, over the long run SRI portfolios tend to exhibit lower overall volatility with slight negative skew.
- The risk-adjusted return of SRI portfolios is approximately the same as normal, unconstrained portfolios. This does not consider correlation with other strategies.
- There is evidence to suggest that companies with the poorest governance practices (ESG) perform the worst in all market conditions.
Comparing the returns from investment managers offering the same investment strategy with or without a SRI filter, we find that the SRI funds report small losses much more frequently. These losses are partly made up for through periods of outperformance, but now with enough consistency to generate alpha long term.
A final point worth mentioning is that back-testing data on the basis of SRI – particular where SRI incorporates ESG considerations - is fraught with inaccuracies. One of the problems here is that significant events such as negative company announcements or environmental breaches can often only be identified after they occur. When they do we often see a fall in shareholder's confidence in management (plus the expectation of fines, delayed production etc.) reflected in share prices. For investors and analysts comparing performance, there are practical limitations to categorizing companies based upon their corporate behavior and operations for each moment in time.
In addition, we note that much of the research employs a theoretical attribution analysis, comparing the relative performance of a market portfolio with- and without-SRI-excluded sectors. Given that commercial applications of SRI are still relatively new this certainly appears to be a more robust method for considering the impact of SRI on return characteristics. We do, however, note that fees and taxes are not considered. With SRI more likely to trigger portfolio rebalances, plus the additional costs associated with monitoring and managing this strategy, it is likely that the true difference in return is slightly greater than indicated by the research.
In summary, we find the evidence supports the view that Socially Responsible Investing (SRI) weighs on investment returns, though tends to outperform in times of market.
Applying these findings to The Jack Brockhoff Foundation’s portfolio, we estimate that adopting an ethical screen through SRI specialist managers could reduce gross returns by as much as $268,000 per annum.
As both Deutsche Bank and Macquarie Private Portfolio Management have indicated there would be no additional cost to the Foundation, the costs would likely be much lower at around $82,000 per annum.
Opportunity Costs
A key tenet of investment is that limiting the breadth of investment opportunities invariably restricts an investor's ability to achieve optimal returns. While this somewhat oversimplified[1], we are able to say with confidence that limiting the number of potential investment opportunities cannot, in and of itself, enhance returns.
Previously we have reviewed the empirical research on ethical filters and noted that several reports point to SRI enhancing returns while reducing risk, and many others that suggest no material difference in return. While these results somewhat contradict Modern Portfolio Theory, to err on the side of caution and to keep an open mind we give equal attention (and skepticism) to all studies.
What we found interesting was that many of the reports that found some benefit, or at least no disadvantage, included at least one market crisis. This may not seem significant at first, but we consider the performance of sectors is inextricably linked to the broader economic environment it stands to reason that highly cyclical, discretionary products and services are first to be cut in times of crisis. For example, in the recent Global Financial Crisis tobacco and gaming shares were hit particularly hard. We also see moments of crisis less directly linked to the consumer, such as the recent oil glut/price crash, which wiped trillions off the value of oil company shares.
It’s these infrequent bouts of devastation that provide SRI strategies with a chance to shine. Left holding higher allocation to slow and stable mature businesses (like grocery stores, agriculture, toll roads), their assets tend to also have “stickier” revenue streams thus making them more resilient and more attractive for investors spooked by plunging share prices.
For a more rounded view of SRI strategies, we must try to determine what level of return is attributable simply to the period of time being studied, and how much is actually due to the SRI screening methods.
To do this it helps if we start by looking at the companies excluded by SRI screening. We can then measure their performance and market-weight to determine their influence on the overall result. Having performed this exercise against some of the largest tobacco, gaming and oil companies in the S&P500 (United States) we find many examples of abnormally higher returns from companies that would normally be banned from SRI portfolios.
Tobacco, for example, has a volatile yet remarkably strong record over recent decades, delivering shareholders an annualized return roughly twice that of the broader share market (to put this in perspective, $1 million invested 20 years ago in the S&P500 would now be valued at around $3.7 million. The same $1 million invested in British American Tobacco would be valued at approximately $10.5 million). Similar, less spectacular examples of outperformance can also be observed with defence contractors, gaming venues (casinos etc.) and alcohol manufacturers and distributors.
Though this disparity in returns suggests that companies excluded by SRI criteria outperform, we find, somewhat surprisingly, that the evidence to support the idea that non-SRI companies outperform is very weak. Of the few investment managers that do specifically and purposefully target companies excluded by common SRI guidelines, we find that performance after fees is very similar to that of the broader market, with slightly higher levels of volatility.
This brings us to the conclusion that the true cost of “ethical investing” is not simply a result of “ethical” companies generating a lower return for shareholders than “unethical” companies. Rather the cost arises indirectly as a consequence from missed opportunities, the full impact of which requires consideration of the number of investment opportunities excluded by any particular filter, as well as the investment manager's skill in working with these constraints while allocating capital in the most efficient manner possible.
A key tenet of investment is that limiting the breadth of investment opportunities invariably restricts an investor's ability to achieve optimal returns. While this somewhat oversimplified[1], we are able to say with confidence that limiting the number of potential investment opportunities cannot, in and of itself, enhance returns.
Previously we have reviewed the empirical research on ethical filters and noted that several reports point to SRI enhancing returns while reducing risk, and many others that suggest no material difference in return. While these results somewhat contradict Modern Portfolio Theory, to err on the side of caution and to keep an open mind we give equal attention (and skepticism) to all studies.
What we found interesting was that many of the reports that found some benefit, or at least no disadvantage, included at least one market crisis. This may not seem significant at first, but we consider the performance of sectors is inextricably linked to the broader economic environment it stands to reason that highly cyclical, discretionary products and services are first to be cut in times of crisis. For example, in the recent Global Financial Crisis tobacco and gaming shares were hit particularly hard. We also see moments of crisis less directly linked to the consumer, such as the recent oil glut/price crash, which wiped trillions off the value of oil company shares.
It’s these infrequent bouts of devastation that provide SRI strategies with a chance to shine. Left holding higher allocation to slow and stable mature businesses (like grocery stores, agriculture, toll roads), their assets tend to also have “stickier” revenue streams thus making them more resilient and more attractive for investors spooked by plunging share prices.
For a more rounded view of SRI strategies, we must try to determine what level of return is attributable simply to the period of time being studied, and how much is actually due to the SRI screening methods.
To do this it helps if we start by looking at the companies excluded by SRI screening. We can then measure their performance and market-weight to determine their influence on the overall result. Having performed this exercise against some of the largest tobacco, gaming and oil companies in the S&P500 (United States) we find many examples of abnormally higher returns from companies that would normally be banned from SRI portfolios.
Tobacco, for example, has a volatile yet remarkably strong record over recent decades, delivering shareholders an annualized return roughly twice that of the broader share market (to put this in perspective, $1 million invested 20 years ago in the S&P500 would now be valued at around $3.7 million. The same $1 million invested in British American Tobacco would be valued at approximately $10.5 million). Similar, less spectacular examples of outperformance can also be observed with defence contractors, gaming venues (casinos etc.) and alcohol manufacturers and distributors.
Though this disparity in returns suggests that companies excluded by SRI criteria outperform, we find, somewhat surprisingly, that the evidence to support the idea that non-SRI companies outperform is very weak. Of the few investment managers that do specifically and purposefully target companies excluded by common SRI guidelines, we find that performance after fees is very similar to that of the broader market, with slightly higher levels of volatility.
This brings us to the conclusion that the true cost of “ethical investing” is not simply a result of “ethical” companies generating a lower return for shareholders than “unethical” companies. Rather the cost arises indirectly as a consequence from missed opportunities, the full impact of which requires consideration of the number of investment opportunities excluded by any particular filter, as well as the investment manager's skill in working with these constraints while allocating capital in the most efficient manner possible.
Filters & Missed Opportunities: ASX300
To get some feel as to how this affects prospective returns in a real-world environment it helps to see how an ethical filter actually affects the breadth of investment opportunities available to us.
Given The Foundation's bias toward Australian equities and the relatively small and concentrated nature of our market, we decided to take a closer look at the 300 largest companies on the Australian share market and consider how our investment opportunities might be restricted by an ethical screen.
To get some feel as to how this affects prospective returns in a real-world environment it helps to see how an ethical filter actually affects the breadth of investment opportunities available to us.
Given The Foundation's bias toward Australian equities and the relatively small and concentrated nature of our market, we decided to take a closer look at the 300 largest companies on the Australian share market and consider how our investment opportunities might be restricted by an ethical screen.
Level 1 SRI Screen
We found that by applying a standard “full” SRI screen[2] only excluded about $219 billion (~14%) of the market. Three of these companies (BHP, RIO, and Woolworths) account for nearly $116 billion of this.
We found that by applying a standard “full” SRI screen[2] only excluded about $219 billion (~14%) of the market. Three of these companies (BHP, RIO, and Woolworths) account for nearly $116 billion of this.
Level 2 SRI Screen
Loosening up the filter a little, we then sorted through companies on the basis of their primary business interests and activities[3]. This reduced the number of excluded companies down to $71 billion (less than 5% of ASX300).
Loosening up the filter a little, we then sorted through companies on the basis of their primary business interests and activities[3]. This reduced the number of excluded companies down to $71 billion (less than 5% of ASX300).
Level 3 SRI Screen
Finally, we then concentrated our filter to only exclude companies that derive the majority of their income from alcohol, tobacco, armaments and gambling. This resulted in $46 billion of excluded securities (less than 3% of the ASX300).
Finally, we then concentrated our filter to only exclude companies that derive the majority of their income from alcohol, tobacco, armaments and gambling. This resulted in $46 billion of excluded securities (less than 3% of the ASX300).
Performance
Upon comparing the relative performance of each of these hypothetical portfolios we find that SRI strategies that avoided the heavy losses of the past 8 years sustained by the energy sector outperformed over one, three and five years (to 28 July 2016).
Most of the losses were attributable to just four companies: BHP (BHP), Santos (STO), Origin Energy (ORG) and Woolworths (WOW).
Upon comparing the relative performance of each of these hypothetical portfolios we find that SRI strategies that avoided the heavy losses of the past 8 years sustained by the energy sector outperformed over one, three and five years (to 28 July 2016).
Most of the losses were attributable to just four companies: BHP (BHP), Santos (STO), Origin Energy (ORG) and Woolworths (WOW).
Administration: implementation, management, compliance
In conducting this review we felt if prudent to bring attention some of the practical aspects involved with mandates that involve restrictions or special considerations.
Ensuring compliance with investment allocation objectives and restrictions is paramount. In the case of ethically screened investment portfolios, this can present challenges, particularly as many large businesses operate multiple lines of business across different jurisdictions.
To manage these potential conflicts many investment managers have dedicated ESG teams or advisers, or rely on external research from organizations such as CAER, Sustainalytics etc.
Naturally, this comes at a cost. For markets in which the Portfolio Manager is selecting individual securities, their reaction time may slow. For markets in which they use external managers (unit trusts etc.), indirect management costs will likely increase as SRI-specialist managers are employed.
There is also the question as to who takes responsibility for The Foundation’s overall compliance with their investment strategy, particularly when a portion of assets are managed in-house, by an external (non-SRI) unit trust, or via a passively managed strategy.
Following discussions with Macquarie PPM and Deutsche Bank, the explicit costs of implementing an ethical screen appear to be minimal, though we should expect – and indeed accept - that indirect costs will increase.
Against The Foundation’s current portfolio we would expect overall administration and management costs to increase by approximately $14,000 per annum.
In conducting this review we felt if prudent to bring attention some of the practical aspects involved with mandates that involve restrictions or special considerations.
Ensuring compliance with investment allocation objectives and restrictions is paramount. In the case of ethically screened investment portfolios, this can present challenges, particularly as many large businesses operate multiple lines of business across different jurisdictions.
To manage these potential conflicts many investment managers have dedicated ESG teams or advisers, or rely on external research from organizations such as CAER, Sustainalytics etc.
Naturally, this comes at a cost. For markets in which the Portfolio Manager is selecting individual securities, their reaction time may slow. For markets in which they use external managers (unit trusts etc.), indirect management costs will likely increase as SRI-specialist managers are employed.
There is also the question as to who takes responsibility for The Foundation’s overall compliance with their investment strategy, particularly when a portion of assets are managed in-house, by an external (non-SRI) unit trust, or via a passively managed strategy.
Following discussions with Macquarie PPM and Deutsche Bank, the explicit costs of implementing an ethical screen appear to be minimal, though we should expect – and indeed accept - that indirect costs will increase.
Against The Foundation’s current portfolio we would expect overall administration and management costs to increase by approximately $14,000 per annum.
[1] In practice this does not always hold true. For example, the costs associated with analyzing every investment may eclipse the additional return available from such an exercise.
[2] Excluded if more than 5% of revenues directly involved in Alcohol & Tobacco, Armaments, Gambling/Gaming, Uranium Mining, Human Rights, CSG, Shale & Thermal Coal, Exploitative Lending Practices.
[3] More than 50% gross revenues.
[2] Excluded if more than 5% of revenues directly involved in Alcohol & Tobacco, Armaments, Gambling/Gaming, Uranium Mining, Human Rights, CSG, Shale & Thermal Coal, Exploitative Lending Practices.
[3] More than 50% gross revenues.