The last few months have seen a number of press releases and environmental finance panel discussions around the issue of embedded carbon and water risk in portfolios. While the ESG community is advocating the integration of management and carbon or water intensity metrics in investment decisions, the financial community (i.e. the portfolio managers) argue that this needs to be done based on financial performance metrics. As some participants in a recent environmental finance roundtable argued: “If we divest out of carbon-intensive portfolios or companies, where do we put the money? There aren’t enough asset alternatives out there”.
The same goes for the business water risk discussion. Investor meetings in the water space argue that asset managers have to take into account water risk in their decisions. The question is: What should be the basis for these decisions? To date, ESG data are largely based on water risk exposure and management metrics, or shadow risk pricing, and not on the metrics that reward portfolio managers: financial risk and returns of their assets. Pricing water risk in the capital markets is daunting.
The recent SEC troubles of S&P, which argued that they provide opinions – not financial advice – and therefore should not be regulated, highlighted how opinions influence markets. Opinions are supported by data. However, do they reflect causation of risk? More importantly: Does it matter given how beliefs and imperfect information influence the rational investor?
As I argued in previous blogs on the development of water value-at-risk (VaR) models, water use, water access, and other physical water risk are not necessarily an indication of financial risk. In fact, the correlation between companies that have high financial intensities such as electric utilities (figure below), mining and O&G companies and their financial volatility risk (measured based on VaR metrics) is at best imperfect.
The waterVaR is based on firm productivity theory to explain the impact of scarce resources (such as water) on economic output of the company in terms of revenue and goods, and relates this to stock volatility risk from environmental exposures. The data show that physical and financial performance metrics don’t correlate. Neither does water use by the company, and financial intensity (data not shown). Yet, a significant focus of integrating environmental metrics in investment decision-making is based on physical data.
As an environmental engineer, I am aware of the operational risks companies face in regards to water, from a policy and from a technology perspective. As a finance expert, I understand that these risks are reflected in increased COGS, capex outlay for PP&E, and accelerated depreciation of assets. However, the translation of these costs in asset risk valuations by the capital markets is unclear at best.
Is there a sound theoretical basis to argue for a relationship between portfolio allocations based on physical water data and financial performance?
The prevailing argument in ESG circles is that good management will also translate into responsible environmental behavior. This has become part of the belief system of the rational investor. While this may or may not be true, both the practice and academic analysis shows unclear relationships between ESG data and the financial performance of S&P500 companies. In addition, MSCI and HSBC data on carbon-balanced portfolios shows a statistical dead heat between ‘carbon-light’ and ‘carbon heavy’ portfolios in terms of return:risk, volatility and tracking performance.
What does this mean? Is there no financial cost for going ‘light’, or does current risk accounting not drive financial performance unless assets become stranded?
At this time, there is no theoretical basis in finance from which to argue that there should be a relationship between physical environmental risk accounting and financial performance. There is a commodity price on carbon, but water (shadow) pricing is largely based on natural capital theory. Can arguments be made from firm productivity or asset pricing theory that resource constraints will result in inefficiencies, risk and cost? And that these should or could impact financial asset pricing?
At Equarius Risk Analytics, we have developed leading indicators for water risk that are based on (and extensions of) financial theories. We have further integrated these waterVaR risk metrics in smart beta designs for price-based portfolio allocations. Shown below are the financial performance results of a portfolio comprised of 28 utilities, allocated both using waterVaR-adjusted and unadjusted price-based strategies. The smart beta was risk-normalized for companies with a waterVaR exceeding 3% of VaR, viewed as a trigger for material risk.
The figure shows that the volatility and return on the portfolio are highly dependent on the allocation of companies with high stock volatility due to water risk, and that there is a window where risk-based portfolios outperform unadjusted portfolios between 55%-70% high waterVaR companies. Below this value, the returns diminish; above this value, the portfolio volatility is above that of price alone. The figure also shows the point of minimal VaR, where the returns on the portfolio are 3.4% below those of price-based allocations alone.
This shows that there is a material financial cost to not incorporate financial risk due to water stress. However, accounting for water risk needs to be based on financial – not physical – metrics, to allow for understanding the risk in equities and portfolios.
The author: Peter Adriaens is CEO of Equarius Risk Analytics and Professor of Environmental Engineering, Entrepreneurship and Finance at the University of Michigan. He can be reached at firstname.lastname@example.org