Assessing carbon risk in portfolios is a difficult but necessary task for investors. Joop Huij, Head of Indices and Disruptive Innovation, and Laurens Swinkels and Dries Laurs, quantitative researchers, discuss novel insights from their research.
Reducing greenhouse gas (GHG) emissions is essential to meet the goals of the Paris Agreement to limit global warming to well below 2.0° Celsius by 2100. The financial industry has a significant role to play in facilitating the transition to a low-carbon economy, not least because asset returns are expected to be hit hard by the impact that climate change has on the real economy.
Institutional investors around the world are therefore evaluating their investment policies with regard to climate risks and opportunities. Climate models such as those developed by Nobel Prize winning economist William D. Nordhaus rely on the introduction of carbon pricing policies – either in the form of emission trading systems or carbon taxes – to eliminate excessive climate risks.
The exact magnitude of the economic impact of a business-as-usual approach to climate change is hard to predict. But simulations suggest that in aggregate, the impact of climate change on the value of financial markets could be as large as 16.9%. However, when GHG emissions are reduced in line with the maximum 2°C temperature increase, this potentially worst outcome drops to just 7.7% of global asset values.
An important question for asset allocators is which risks are rewarded in financial markets. The existence of a risk premium can be determined theoretically or empirically. For carbon risks, the academic literature has put forward widely varying estimates of both the sign and magnitude of the associated risk premia.
Most of the negative climate impacts are only beginning to be observed. Climate risks are therefore not well represented by historical data, as it is likely that markets have only started to price carbon exposure in recent years. As a result, traditional asset pricing methods are less suited for pricing carbon and other climate change-related risks.
Investors concerned with climate change can make a number of possible strategic asset allocation decisions. They can decarbonize their investment portfolios by divesting the largest GHG emitters, for example, by creating a fossil fuel-free portfolio.
However, divestment comes down to a transfer of ownership to other – potentially less sustainable – investors. Alternatively, investors could explicitly target investments in companies that help the transition to a low or zero-carbon society. This could, for example, be achieved through allocations to thematic funds, clean-tech private equity, or green bonds.
All this has consequences for portfolios. Fossil fuel-free investments add considerable relative risk to a capitalization-weighted market portfolio. In addition, radical carbon risk reduction may unintendedly alter factor-based equity portfolio strategies.
To get an idea of how markets are exposed to carbon risks, it is helpful to consider the potential financial effects of a universal carbon tax. While the cash flows of some companies would be directly hit by such a tax, the impact for others would be lessened by their ability to raise prices or substitute their current emissions with low-emission alternatives.
In addition, companies without significant direct emissions might be negatively affected as well, as their input products would become more expensive. In the case of banks, some of their most carbon-intensive borrowers might not be able to repay all of their debt. Such scenario analysis can be helpful in examining the financial impact of climate change on an investor’s portfolio.
The majority of investors use corporate emissions data to manage such risks. But reliable data has been mostly lacking in the past. As a result, a commonly accepted framework on how to account for and report on emission activities was introduced in the form of the Greenhouse Gas Protocol in 1998.
Since then, increasingly more data providers have started to publish corporate emissions data. Most of these providers offer data on Scope 1 and 2 emissions, which are direct emissions from corporate activities, and indirect emissions from the purchase of electricity, respectively. Scope 3 emissions are all other indirect emissions resulting from a company’s upstream and downstream value chain. Yet Scope 3 emissions rely heavily on simplifying assumptions and are only scarcely available. Besides, they are not fully reflective of carbon risk, as they overlook future aspects such as the feasibility of a company’s transition strategy.
We propose a market-based measure to estimate carbon risk in a complementary way. If carbon risk represents a systematic risk factor that partially drives returns, a multi-factor asset pricing model will be able to uncover asset-specific exposures to this risk factor.
Key to this analysis is finding a suitable proxy for systematic carbon risk. A portfolio with long/short exposures to assets with roughly opposing footprint characteristics, or the price development of carbon allowances as traded in Emission Trading Schemes, might be suitable candidates for creating a carbon risk factor.
Regressing the return series of any financial asset on the return series of the carbon risk factor while controlling for other exposures to traditional factors allows us to estimate an asset’s carbon risk sensitivity. An asset with high (low) carbon sensitivity generally rises (drops) in value when the carbon risk factor rises in value. Hence, it is highly exposed to carbon risks, even if it does not report on carbon emissions at all.
As long as a suitable carbon risk factor can be found, estimating an asset’s carbon sensitivity only requires the availability of the asset’s return series. It also means that the analysis can be used for other asset classes besides equities.
Assets for which emissions data is not yet available or to which it is not relevant can be assessed in a similar manner. For commodities, private equity or real estate, such insights might prove valuable.
This article is an excerpt of a special topic in our five-year outlook.