Paper: arXiv 2311.12450
Authors: Michele Azzone, Maria Chiara Pocelli, Davide Stocco
Abstract
Sustainable investing refers to the integration of environmental and social aspects in investors’ decisions. We propose a novel methodology based on the Triangulated Maximally Filtered Graph and node2vec algorithms to construct an hedging portfolio for climate risk, represented by various risk factors, among which the CO2 and the ESG ones. The CO2 factor is strongly correlated consistently over time with the Utility sector, which is the most carbon intensive in the S&P 500 index. Conversely, identifying a group of sectors linked to the ESG factor proves challenging. As a consequence, while it is possible to obtain an efficient hedging portfolio strategy with our methodology for the carbon factor, the same cannot be achieved for the ESG one. The ESG scores appears to be an indicator too broadly defined for market applications. These results support the idea that bank capital requirements should take into account carbon risk.
Complexity vs Empirical Score
- Math Complexity: 7.5/10
- Empirical Rigor: 6.0/10 (capped from 8.0: transaction costs not modelled; no out-of-sample test; see penalty flags)
- Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 8.0)
Why this score: The paper employs advanced network theory concepts (TMFG, node2vec) and statistical methods, indicating high math complexity, while it provides a specific, backtest-ready methodology with empirical results on S&P 500 data and risk metrics, demonstrating strong empirical rigor.
Research Flowchart
flowchart TD A["Research Goal: Hedge Climate Risk in Equities (S&P 500)"] --> B["Input Data: Historical Price Data & Risk Factors<br>(CO2 Intensity, ESG Scores)"] B --> C["Core Methodology: Network Approach"] C --> D["Step 1: Triangulated Maximally Filtered Graph<br>(Construct Asset Network)"] C --> E["Step 2: node2vec Algorithm<br>(Extract Structural Embeddings)"] D & E --> F["Compute Correlations &<br/=> Build Hedging Portfolios"] F --> G["Key Findings & Outcomes"] G --> H["CO2 Factor:<br>Successfully Hedged<br>(Strong link to Utility Sector)"] G --> I["ESG Factor:<br>Hedge Failed<br>(Indicator too broad for market application)"] G --> J["Policy Implication:<br>Bank Capital Requirements should<br>incorporate Carbon Risk"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a novel methodology using a Triangulated Maximally Filtered Graph and node2vec algorithms to build hedging portfolios for climate risk factors such as CO2 and ESG.
Source sentence
We propose a novel methodology based on the Triangulated Maximally Filtered Graph and node2vec algorithms to construct an hedging portfolio for climate risk, represented by various risk factors, among which the CO2 and the ESG ones.
Assumes 6
- All portfolios are equally weighted. execution
Source sentence
All the portfolios are equally weighted.
- Daily logarithmic returns are used to evaluate portfolio performance. statistical
Source sentence
We calculate their daily logarithmic returns and study their exposure to the FF5 factors and to the additional sustainability factor.
- Only stocks that remained in the S&P 500 throughout the entire six‑year period are included. data
Source sentence
We select only stocks that remained part of the S&P 500 Index throughout this entire six-year period to construct our financial network.
- The CO2 factor is strongly correlated with the Utility sector over time. market
Source sentence
The CO2 factor is strongly correlated consistently over time with the Utility sector, which is the most carbon intensive in the S&P 500 index.
Show 2 more
- The node2vec embedding preserves relationships among nodes with minimal information loss. model
Source sentence
By applying the node2vec algorithm to our filtered network we end-up with a lower dimensional representation of the relationships between nodes while minimizing the information loss.
- Regression analyses use Newey‑West standard errors and HAC robust standard errors. statistical
Source sentence
In Table 2, we report the results of the six regressions, considering Newey‑West standard errors.
Data
- Refinitiv Eikon Reuters dataset market-prices commercial
Source sentence
... The resulting dataset is composed of 470 stocks, each of them described by a vector of daily price observations.
Costs
Not modeled
Finds 5
- An efficient hedging portfolio can be built for the CO2 factor, but not for the ESG factor. negative
Source sentence
As a consequence, while it is possible to obtain an efficient hedging portfolio strategy with our methodology for the carbon factor, the same cannot be achieved for the ESG one.
- Including the CO2 factor in regressions increases adjusted R‑squared and is significant for the close and far‑close portfolios. statistical
Source sentence
When we include the CO2 factor in the regressions, the adjusted R-squared always increases. The CO2 factor is significant at 5% for the close portfolio and the far-close portfolio.
- The close and far portfolios outperform the far‑close portfolio on Sharpe, Sortino, Omega ratios, with the far portfolio being the most efficient. performance
Source sentence
The close and the far portfolio outperforms the far-close portfolio in terms of all ratios (the far portfolio being the most efficient).
- The CO2 factor is consistently close to the Utility sector in the embedded network, while ESG factors show no specific sector proximity. theoretical
Source sentence
The CO2 factor is consistently close to the Utility sector while the three ESG factor are not close to a specific sector in different time-windows.
- Adding the ESG (promised or realized) factor to FF3 regressions yields a consistent improvement in R‑squared. statistical
Source sentence
When adding the corresponding risk factor to the ESG promised and ESG realized regressions we observe a consistent improvement.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 23 related papers). Coverage is partial, so this is neither replication nor refutation.
1 paper in the archive
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (27)
- Tilting the Wrong Firms? How Inflated ESG Ratings Negate Socially Responsible Investing under Information Asymmetries (2022)
- The rise in investors’ awareness of climate risks after the Paris Agreement and the clean energy-oil-technology prices nexus (2022)
- The impact of carbon disclosure and carbon emissions intensity on firms’ idiosyncratic volatility ({2023} publisher={Elsevier})
- The environmental pillar of ESG and financial performance: A portfolio analysis (2023)
- Reformulating taxes for an energy transition (2019)
- Green investment and asset stranding under transition scenario uncertainty (2023)
- The sharpe ratio (1998)
- Carbon beta: A market-based measure of climate risk (2022)
and 19 more
- Inside the ESG ratings: ( D is)agreement and performance (2021)
- Chasing the ESG factor (2022)
- ESG controversy as a potential asset-pricing factor (2023)
- A factor approach to the performance of ESG leaders and laggards (2022)
- A five-factor asset pricing model (2015)
- Common risk factors in the returns on stocks and bonds (1993)
- Carbon risk (2020)
- Do investors care about carbon risk? (2021)
- Pearson correlation coefficient (2009)
- Network filtering for big data: T riangulated M aximally F iltered G raph (2016)
- node2vec: Scalable feature learning for networks (2016)
- ESG rating disagreement and stock returns (2021)
- Aggregate confusion: The divergence of ESG ratings (2022)
- Does carbon risk matter? evidence of carbon premium in eu energy-intensive companies (2021)
- Establishing ESG as risk premia (2018)
- ESG preferences, risk and return (2021)
- The relationship between air pollution, investor attention and stock prices: Evidence from new energy and polluting sectors (2021)
- Addressing COP21 using a stock and oil market integration index (2018)
- Have consumption-based CO2 emissions in developed countries peaked? (2024)
0 of 27 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.