While it might seem better to focus on current so-called ESG leaders, we believe there is untapped value in the underappreciated ESG improvers.
- ESG as a source of alpha: no longer just an element of risk-management, ESG is a growing source of risk-adjusted performance, especially among ESG improvers
- The real value is in identifying positive ESG trends: companies with an improving ESG profile relative to their industries tend to outperform those with static or deteriorating ESG credentials
- Machine Learning can help discover ESG momentum: Equity Machine Intelligence’s (EMI) models can glean signal from noise, quickly and effectively identifying ESG improvers and better monitor their changes through time than their human counterparts
In the television series “Mission Impossible,” the protagonist would receive instructions that began with the famous line: “Your mission, should you choose to accept it...”. When asked to integrate environmental, social and governance (ESG) factors into their investment process, portfolio managers, perhaps less suave than our erstwhile hero, could no doubt be forgiven for feeling a certain affinity for the difficulty of the mission. Indeed, ESG assets were on track to exceed $50 trillion by the end of 20211, and while fully integrating ESG into investment processes is easier said than done, it is even harder to do so in an authentic and value-additive way. Then again – why should it be easy?
Relying on exclusionary screens and headline ESG ratings is nowhere near enough to achieve the mission. Generating “value through values” — a principle that Voya embraces – requires doing things differently. Mission accepted.
ESG as a source of alpha
Thankfully, some investors are beginning to look through simplistic headline ESG ratings and asking important questions, such as: Can ESG data inputs add value to the stock selection process? Which issues are material to financial risk and performance (i.e., doing well)? Which ESG metrics are most linked to broad societal goals and a company’s license to operate (i.e., doing good)? Is it better to focus on current so-called ESG leaders, or is there untapped value in the underappreciated ESG improvers?
For starters, the answer to the first question is a resounding, yes. In a meta-analysis of over 2,000 studies on the impact of ESG considerations on equity returns, 63% of findings were found to be positive and only 8% were found to be negative. More specifically, research indicates that ESG can add alpha from both a tilt and momentum – or trend – perspective.2,3
When looking at the remaining questions, however, there is a big caveat: not all ESG metrics are created equal, and thus, outperformance depends on successfully identifying material ESG issues from the hundreds of available metrics and datapoints. Moreover, the importance of different sustainability issues varies systematically across firms and industries.4 In short, there is no free ESG lunch. So how, then, to address the twin challenges of materiality and conditionality?
The long-standing approach of Voya’s Equity Machine Intelligence (EMI) group is to use machine learning to tease out the most material ESG metrics – i.e., those linked to positive financial performance. This takes ESG investment process integration to the next level.5
Real value is in identifying positive ESG trends
For many, ESG investing has long focused on the current rating of a company: “if it’s high then buy, if it’s low then avoid”. This simplistic snapshot approach neglects crucial nuances, such as a company’s ESG characteristics tending to evolve over time, rather than remaining static. Voya’s EMI team has found that companies with an improving ESG profile relative to their industries outperform those with deteriorating ESG credentials.
We call this “ESG momentum” and calculate it as the improvement in a company’s material ESG features through time relative to the relevant industry group. For instance, when Hormel Foods Corporation (ticker: HRL) was identified in January 2013 by our EMI models as a potential long-term outperformer, its ESG rating was only modest, according to published MSCI ESG data. However, the model identified the company as underappreciated and saw that key fundamental and ESG metrics were beginning to improve. Not only has Hormel made substantial improvements since then, the company’s market capitalization also doubled from around $10 billion to over $20 billion. As of the end of 2021, Hormel was among the leaders in its industry for ESG integration and goal setting, having set 20 qualitative and quantitative goals to achieve by the end of 2030 and receiving recognition across the ESG spectrum.
While we use proprietary inhouse ESG models to determine momentum, publicly available data and research support this approach on an asset class level.2,3 A March 2020 research paper from AON6 used MSCI ESG scores to show that high ESG momentum companies outperform low ones (Figure 1 is based on this study).
Past performance is no guarantee of future results. Cumulative performance differential of the top and bottom ESG momentum quintiles for developed market companies (MSCI World). ESG momentum is defined as the prior 12 month change in ESG score. Performance is constructed from a hypothetical long-short indexed portfolio, going long for the sector neutral upper ESG momentum quintile of MSCI World Index, while the bottom sector neutral quintile goes short. Data is from December 2007 to December 2021. Source: Voya Investment Management, Factset, MSCI.
That ESG momentum is linked to stock outperformance speaks to a logical economic intuition: A company with weak ESG credentials (i.e., high ESG risk) that moves in the right direction on material ESG factors should become lower risk, and hence, may benefit from improved valuations and a corresponding fall in its cost of capital. This is analogous to credit risk, where a company’s bonds benefit as it moves from a credit rating of, say, BBB to A.
This makes financial sense, too, as ESG improvements link to cash flow in five important ways:7
- facilitating top-line growth
- reducing costs
- minimizing regulatory and legal interventions
- increasing employee productivity, and
- optimizing investment and capital expenditures
Furthermore, recent research points to another advantage of ESG momentum: Companies with positive ESG momentum tend to have a bigger impact on the environment and society in general, such as by reducing their waste or pollution footprint or by supporting their local communities. Recent research by Global AI Corp8 found that an ESG momentum portfolio has historically outperformed the MSCI US index and has a relatively better SDG (U.N.’s Sustainable Development Goals) footprint than that of the broader index. According to the report, “We establish a positive contemporaneous connection between the portfolio’s ESG ratings momentum and its SDG footprint. Thus, a positive linkage exists between ESG, alpha, and the SDGs.”
Finally, improving ESG metrics may lead to a company’s inclusion in ESG screens or indices, and attract fund flows as investors become more and more ESG-aware.
Machine learning can help discover ESG momentum
Active ESG investing has the potential to drive outperformance by identifying companies where ESG risks are mispriced. This “ESG Alpha” can be achieved by integrating material ESG factor analysis deep into the investment process and by unearthing companies which are demonstrating positive momentum in addressing ESG risks and opportunities (ESG improvers). Both paths are complementary, recognizing that no company’s ESG profile is set in stone – rather it evolves over time, as is the case for credit risk.
Better yet, the overweighting of ESG improvers helps recognize companies that are changing their characteristics for the better (by lowering their cost of capital versus competitors over time). At a societal level, there is scope for this to be a force for good.
However, ESG investors can’t have their (sugar-free) cake and eat it too. To outperform, they must outcompete others in gleaning signal from noise. Here is where Equity Machine Intelligence can help. Not only do machine learning systems typically process mountains of corporate ESG data more quickly, they can do so more effectively, while monitoring changes through time. This means ESG integration can go well beyond the well-known limitations of headline ratings. Furthermore, ensuring ESG portfolios are adding societal value requires a shift towards measuring the true impact that companies have on their communities – that is, to bring transparency and authenticity to the process. At Voya, we see this as our mission, and we embrace it wholeheartedly.