“Just when I learned the dance, they changed the music.”Cover to August 1927 Saturday Evening Post
Download the PDF of this note here.
Most of our research is devoted to uncovering and analyzing narratives in financial and political markets. Investment industry conventions and practices, however, are susceptible to abstraction and common knowledge effects as well. The most glaringly obvious example? Socially responsible, sustainable or ESG investing.
Putting aside our personal perspectives on the underlying premises of ESG investing (or those put forward by its practitioners), it is easy to see why it lives so easily in the land of narrative and abstraction. First, ESG doesn’t have a single objective definition, which means that it can be modified and tilted to reflect individual predispositions. Beyond what it actually is, however,the idea of responsible or sustainable investing is inextricably attached to all sorts of powerful social ideas, like political identity, moral value, finding meaning in professional work, national security, justice, climate change, and the roles of labor and capital. When we talk about ESG investing, sometimes we are talking about…ESG investing. Far more often, we are really talking about a combination of these powerful social ideas. That layer of abstraction makes ESG a ripe target for narrative and missionary behavior.
For the True Believer who happily incorporates social values (other than a belief in the power of markets and efficient allocation of capital, that is) into their investment philosophies and processes, this isn’t a big deal. Sure, as common knowledge about what it means to invest responsibly shifts, it may influence implementation. We’ve observed that, for example, with a general shift from negative screen-based approaches to more positive, proactive definitions. ‘Impact investing’ seems to be the most current term of art. But the big questions posed by SRI/ESG narratives are for advisers, allocators and other investors who may be ambivalent or even cynical about the whole thing. If common knowledge about ethical and responsible investing changes, it will influence the products available to us. It will also influence the pressures placed on us by clients, boards, legislatures and other constituencies. It will influence our reputations and livelihoods as investors. We can believe whatever we want about ESG, but we cannot escape the influence of a powerful narrative, or an industry which has come to believe that everyone knows that everyone knows that ethical investing means following this set of rules or another.
From time to time, the issue comes up in force, usually in a featured piece arguing that “ESG’s Time Has Finally Come”, or something to that effect. I’ve heard from a lot of readers and subscribers that they feel this is happening now. We have observed a great many conversations that came out of Cliff Asness’s piece last year, for example. It seems to us that it has been present on more institutional investment conference agendas, too. But more conferences and some extra articles do not a narrative make. So let’s put a finer point on it. Is there a persistent and growing ESG narrative? Has it finally become part of the Zeitgeist, a long-term feature of our investing culture?
It’s a perfect question to run through the Narrative Machine. We did. And we think the answer is no.
ESG and The Narrative Machine
Using our usual set of broadly distributed financial media sources from Quid, we added Pensions & Investments and Institutional Investor to supplement discussions of the practices of asset owners. We explored ESG/SRI stories on a quarterly basis from the beginning of our dataset in September 2013 through the end of February 2019.
To answer this kind of question, the focus of our analysis will be our Cohesion measure. For any network graph of related articles, Cohesion measures the share of articles for which the mean normalized harmonic distance between a node and all other nodes is very near. That’s a very fancy way of saying that it measures what percentage of articles are saying very similar things using very similar language. It is not our only measure of narrative influence, but it is the one that best reflects the strongest sine qua non for common knowledge: the forceful crowding out of interpretations of a topic which don’t fit the dominant narrative.
The exhibit below displays our narrative cohesion measure for ESG/SRI since Q4 2013.
The scale for cohesion falls between 0 and 1, so the scale of shifts experienced over the last 5+ years is dramatic. These are not rolling measures, but reflect only the period in question. The cyclicality, then, is a real feature of the data, and not some artifact of smoothing or averaging. In short, the cohesion of ESG/SRI narratives appears to rise and fall in cycles. It is useful, then, to understand what a low cohesion ESG/SRI network looks like, and what a high-cohesion ESG/SRI network looks like.
In the low cohesion example – here from Q3 2015 – a few things are apparent. First, even within clusters, the NLP program finds fewer adjacencies, which in the visualization leaves fewer lines and more “white space.” Second, articles about different entity types separate into clusters with less connectivity; for example, articles about ESG/SRI in context of asset managers, asset owners and corporations all use distinct enough language to produce separate clusters at greater distance from one another. Third, articles relating to financial market results (e.g. those I have marked using the far-right circle) are distant from and loosely connected to strategic or ‘philosophical’ pieces about responsible investing.
In the aggregate, even at low ebb this is still a topic with reasonably high cohesion in comparison to others we research. While comparisons across topics are difficult, both visualization-based and quantitative measures of cohesion for ESG/SRI are consistent from what we would expect from a network with active narratives and a mature taxonomy for those narratives. Note, as usual, that the squares reflect clusters produced by Quid, while the circles drawn reflect centers of gravity within the network graph that I have identified based on distance and connectivity across clusters.
Narrative Cohesion – Q3 2015
The exhibit below presents the network graph at its peak value two years later, in Q3 2017. There are very few clusters or even nodes that do not fall within tightly defined clusters, and nearly every cluster overlaps somewhat with other clusters. The plain-English explanation of this clustering is that the language used in discussions about ESG and SRI (both descriptive language as well as language that communicated value judgments) was far more similar. The existence of a strong counter-narrative or differential view would likely have resulted in more distance between clusters and nodes, and almost certainly more isolated clusters at much further average distance. It didn’t. This is what common knowledge looks like visualized.
Narrative Cohesion – Q3 2017
We think the data suggest two questions:
- What is it that causes the strength of ESG/SRI narratives to ebb and flow like it has over the last 5+ ears?
- What is it that makes Epsilon Theory believe that ESG/SRI principles are not part of the common knowledge of a new, long-term investing Zeitgeist?
Fortunately, we think that both questions have the same answer. That answer lies below:
Our belief – and it is just that, a belief – is that ESG/SRI investing remains a bull market philosophy, a luxury of periods in which returns have been good. The absence of challenging financial market returns has created a more fertile ground for these strategies to grow and thrive, but in the end, we believe that the same dynamics which appear to clearly drive cyclicality in the shorter-term narrative strength around ESG/SRI are likely to prevent its emergence as a true Zeitgeist-level change that would force investors into the same kind of polarized positions on ‘responsible investing’ that we observe in political markets.
Said another way, there are no atheists in foxholes, and there are no fiduciaries who won’t walk away from their big ESG/SRI/Impact Investing plans when risky assets are sinking. We don’t think it means that ESG/SRI won’t grow. It probably will. But we also think it means that investors who are mostly concerned about whether they need to build ESG/SRI capabilities beyond what their constituencies demand can safely decline to do so.