Climate scenario analysis and SAA

Earlier this year we formally integrated climate scenario analysis into our long-term forecasts for expected returns. In ‘strategic asset allocation’ (SAA), when we talk about expected returns, ‘expected’ is used in the technical sense of ‘probability-weighted mean across scenarios’.

We use a number of structural macro-economic scenarios for this purpose, which we call ‘paradigms’. We have a ‘high inflation’ and a ‘deflation’ scenario, for example. We have also added a range of climate scenarios in our expected return forecasts.

Our climate scenario modelling platform contains 16 different climate scenarios, covering a range of assumptions shaped by the stringency of climate policy in different regions and sectors, as well as assumptions about technological progress.

Each climate scenario has an associated carbon and global temperature trajectory, together with trajectories for energy demand, oil prices and other key economic factors.

Bespoke climate platform

Climate scenarios stress testing is now required by financial regulators around the world and also in a number of reporting frameworks, for example by the Taskforce on Climate-related Financial Disclosures.

Our models go beyond the standard scenario approach in two main ways.

Firstly, we have tried to be more realistic in our assumptions. The standard Network for Greening the Financial System scenarios used by central banks assume a single global carbon price¹. We think it’s more realistic to assume different degrees of climate policy stringency in different regions, resulting in better insights into development paths and how these affect companies indifferent places.

Secondly, we assign a probability to each scenario to generate a probability distribution of potential climate outcomes. We describe our approach to scenario specification in our detailed white paper².

What we’re expecting

Our scenario analysis tool allows us to generate forecasts on the effects on fair value for more than 20,000 individual stocks and bonds for each climate scenario. These can be aggregated up to the level of sectors or regional indices.

The following chart shows the latest results for 12 sectors within the MSCI ACWI global equity index, based on current assumptions embedded in our probability-weighted mean, or ‘most likely’, climate scenario.

Chart 1 - Differing sectoral impairments under climate scenarios

Note: Sectors are categorised using GICS classifications. Valuation impairment figures are all relative to market prices as of 31/06/2021. We update this analysis annually, with the next iteration expected September 2022.

For some sectors the impact of the climate transition is negligible – healthcare, information technology, and financials, for example. In general, these sectors have a low-carbon intensity, so this should come as no surprise.

The biggest effects will be felt in the energy sector, which contains the big fossil-fuel producers. The fair value impairment for this sector is 28%.

This means that, at the time the scenario analysis was run (June 2021), the estimate of the net present value of the sector’s future cash flows was 28% lower than the value assigned by the market price. Essentially, this is because we forecast a quicker decline in global oil demand than that assumed by the markets (in our probability-weighted mean scenario).

Conversely, utility is the one sector that stands to gain – with a 10% uplift in fair value compared to market prices. In our mean scenario, we expect more demand for electricity in a decarbonising world, creating potential for faster growth overall in this sector.

What does this mean for SAA?

Our approach is similar to the paradigms model we use more generally in SAA. The focus is on calculating the ‘expected’ impact on annualised returns in the probability-weighted mean across scenarios.

To calculate this we aggregate the individual company impairments for each of our regional equity indices. We then make two important assumptions:

1. The market mispricing implied by our impairment calculation is ‘corrected’ within five years. It is reasonable to assume that the mispricing will be corrected, but five years is an arbitrary time period. It’s hard to know what will trigger the correction or when it will occur. This assumption means that if a company has a -10% impairment in the climate scenario analysis, this translates into approximately a -2% per annum impact on returns in each of the next five years.

2. Climate risk is ‘orthogonal’ to macro-economic paradigms. We assume there’s no relationship between economic paradigms and climate transition assumptions (‘orthogonal’ in the technical jargon). This isn’t entirely realistic – carbon emissions growth will likely be slower in a ‘Deflationary Slump’ paradigm (i.e. low or negative real economic growth with no inflation) than in a ‘Back to the New Normal’ paradigm (i.e. elevated corporate activity generating strong growth) that results in higher emissions. However, at this stage, it’s not practical to run combined economic and climate scenarios across the 16 x 6 combined scenarios we explore. The upshot of this assumption is that we can simply add the expected climate impairment impact to the expected paradigm return.

The following chart shows results for the main regional equity indices:

Chart 2 – Regional equity indices have differing climate impairments

Source: Planetrics, abrdn.

Note: indices are calculated using standard MSCI Benchmarks. Valuation impairment figures are all relative to market prices as of 31/06/2021. We update this analysis annually, with the next iteration expected September 2022.

Investment lowdown

Based on the assumptions in our probability-weighted mean, or ‘most likely’, scenario:

• Impact on expected returns is quite modest – lower than 1% per annum in most cases. There are two main reasons for this. First, the most negatively-impacted sectors (oil and gas and mining) have a relatively small weight in most indices – oil is now less than 5% of the MSCI ACWI. The biggest sectors – technology, financials, and healthcare – are little affected. Second, winners tend to offset losers. In the chart above, the positive impairment of the utility sector somewhat offsets the negative impairment for the energy sector, so that at the index aggregate level the net effect is small.

• All regions are negatively impacted. This might come as a surprise but the climate transition is, in effect, a quasi-tax on companies. Companies will need to pay in order to emit greenhouse gases and this will eat into corporate profits.
  
  Of course, if investors wish to make more thematic strategic asset allocations, for example as part of a ‘net-zero’ strategy, the positive or negative effects on returns can be materially larger – adding 1-2% per annum to the annualised expected returns of a climate solutions strategy.

• Impact on expected returns for standard credit benchmarks is negligible. This is for two reasons. First, credit is senior in the capital structure to equity. If a company’s earnings collapse because of a decline in demand for oil or diesel cars, the value of its shares may collapse, but its creditors may still be repaid. Second, credit risk is much more time dependent than equities. An equity’s value can be thought of as a discounted sum of its future expected cash flows in perpetuity. However, a 10-year bond is redeemed a decade after issuance so its price is unaffected by climate change in the years after its maturity. The main corporate bond indices have an average maturity of under 10 years.

¹ Regulators often make use of the scenarios developed by the Network for Greening the Financial System - https://www.ngfs.net/ngfs-scenarios-portal/.
² abrdn (2021) Climate Scenario Analysis: A rigorous framework.