Signatory type: Asset manager
HQ location: United States
Covered in this case study:
Asset class: Equities
AUM: $685bn in equities (Jan 2023)
The Invesco Quantitative Strategies team manages sustainable systematic equity strategies that aim to deliver superior long-term performance. The investment process is built on three key factors forming the multi-factor model: momentum, quality and value. Invesco is a member of the Net Zero Asset Managers initiative (NZAM) and is committed to supporting the global goal of reaching net-zero greenhouse gas (GHG) emissions by 2050 or sooner.
Why we pursue a net-zero approach
Following the 2021 Intergovernmental Panel on Climate Change report on the climate emergency, countries, companies and investors are increasingly acknowledging the importance of reducing GHG emissions. There is an undeniable momentum building around the need to manage climate-related risks. This, alongside the fact that Invesco takes a client-centric approach towards sustainable investing, means that we are also eager to offer solutions that are in high demand to our clients. On 12 April 2022, Invesco submitted its NZAM-related disclosures to the Institutional Investors Group on Climate Change, where we committed $195.3 billion of our assets to a net-zero approach.
Regulatory requirements, additional ESG objectives such as net zero, increasing data complexity and financial objectives (i.e., return targets, risk tolerance, etc) have made the construction of a true ESG portfolio with financial targets highly complex. A systematic approach that can handle multiple data sets and various constraints is well positioned to deliver an optimal solution, one that both controls unrewarded risks and ensures financial objectives are met.
In order to also make portfolios compliant with net-zero targets, the Invesco Quantitative Strategies team set out to develop a framework that evolves the existing process of harvesting returns from rewarded factors (i.e., those factors that have been associated with a long-term return premium) while controlling for risk.
How we developed a systematic approach
Invesco has developed an innovative two-step optimisation approach to deliver on non-financial targets, such as net-zero alignment, and financial objectives, such as relative return, volatility and tracking error.
In step one, we construct an anchor portfolio that aims for temperature alignment with the 1.5C° Paris Agreement goal using data from Vivid Economics. The raw temperature score for each issuer is ultimately a combination of Scope 1, 2 and 3 GHG emissions and the company’s decarbonisation pathway. Stocks for which no implied temperature rise (ITR) is available – on average, around 7% of the total global large- and mid-cap universe – are excluded from the universe.
Determining the temperature alignment of an asset is highly model dependent. Estimates of a sector’s and a company’s forward-looking, multi-year carbon trajectory ultimately results in heavy reliance on assumptions. The temperature alignment for each asset is aggregated at the portfolio level using a weighting approach that reflects both the size of the holding in the portfolio (weight) and the current emissions intensity (measured as GHG emissions relative to the companies’ revenues) of each asset. Therefore, a temperature reduction in low-emitting sectors does not reduce the portfolio temperature significantly. The process gives greater influence to emission reductions in high-impact sectors, such as energy or utilities.
We target a specific portfolio temperature alignment by formulating an investment objective that contains the temperature goal rather than defining it as one of many constraints in the portfolio optimisation. This approach ensures that temperature alignment is the focus of portfolio construction while all other risk factors are neutralised as much as possible.
With the first step resulting in a temperature-aligned anchor portfolio, the second step positions the portfolio towards the traditional investment style factors, such as quality, momentum and value. The team uses multiple underlying signals to harvest the factor premium. For value, a combination of cashflow, book value and earnings yield is utilised. For momentum, the team combines elements from fundamental momentum for earnings and price momentum – using customised versions of the traditional 12-1 month stock performance, i.e. the last 12 months’ performance minus the most recent month. Lastly, quality is comprised of signals measuring profitability, leverage and accounting integrity. Key inputs for the optimisation are the factor attractiveness for each company as well as a risk estimate for each asset. The optimisation then strikes a balance between additional factor exposure against an increase in risk. Harvesting premiums associated with each factor helps to deliver on financial targets while adhering to the temperature-alignment objective.
There are additional benefits of a tiered portfolio construction approach. Performance and risk attribution is much cleaner. It allows for a clear separation of the impact of the temperature alignment in step one and the impact of the factor overlay from step two.
A full description of the process can be found in our paper Factor Investing in Paris: Managing Climate Change Risk in Portfolio Construction.
The investment approach outlined above was applied to a global developed markets universe with more than 3,000 stocks. The ITR estimates in the universe range from 1.20°C to 4.80°C and the simple average ITR is 2.82°C. When weighting by market cap instead of emissions, the average ITR is slightly lower (2.58°C). The first step is to align the portfolio with the 1.5°C objective while controlling for tracking error relative to a broad equity market index – in this case the MSCI World. The portfolio is constructed using a mean-variance optimisation approach. To achieve a net-zero result, the expected-return input is the negative of the carbon-weighted temperature score. The ex-ante tracking error against the benchmark is capped at 1%. Constraints are also applied to the sector deviations. The resulting anchor portfolio holds around 630 stocks.
In step two, we fix the temperature path and do not allow the final portfolio to deviate from the 1.5°C trajectory. The portfolio inherits the constraints (such as sector and country) from the previous optimisation. Moreover, we tightly control active positions in sectors, industries and countries by applying constraints relative to the cap-weighted benchmark of ± 2%. A wider risk budget of 3% (against the temperature-aligned anchor portfolio) is used in the factor overlay optimisation. The final portfolio consists of close to 200 stocks and achieves its goals with significantly fewer stocks than the MSCI World, with around 1,560 holdings.
Illustrated below in Figure 1 are temperature alignments for four portfolios, as calculated by Invesco:
- MSCI World: 3.93°C when the strategy was constructed in September 2021 (based on portfolio aggregation that reflects the carbon footprint of the sector the stock belongs to)
- Anchor portfolio, step 1: 1.49°C
- Factor overlay, step 2: 1.49°C
- One-step portfolio construction: 1.49°C
To compare characteristics across factors and across companies, the tables under Figure 1 show standardised factor exposures (MOM = momentum, VAL = value, QAL = quality). Those factor exposures are constructed to follow a standard-normal distribution and denote the z-value, naturally ranging from -3 to +3. The figures show that Invesco’s two-step tiered construction is well suited to deliver on both goals, i.e., temperature alignment and establishing stronger and well-balanced factor exposures. The one-step process produces the same ITR but with lower value and quality factor exposures.
Figure 1: ITRs of the benchmark and the optimised portfolios.
Source: Invesco, Vivid Economics. For illustrative purposes only.
Given limited data history, our analysis was run for a single point in time. The absence of history for the ITRs in particular has made back-testing difficult. However, in order to manage such a strategy, we advise our institutional clients implementing this strategy with Invesco to rebalance the portfolio on a monthly basis to ensure the temperature alignment and to maintain tracking error budgets and a sensible factor profile.