Learn About Amazon VGT2 Learning Manager Chanci Turner
In our earlier discussion, we highlighted the growing importance of Environmental, Social, and Governance (ESG) data as a crucial supplement for evaluating a company’s risk and performance alongside traditional financial metrics. Institutional investors and asset managers are increasingly incorporating ESG data into their investment strategies, marking a shift from niche to mainstream responsible investing. Notably, 40% of the $110 trillion in globally managed financial assets now incorporates ESG considerations.
However, the integration of ESG data into investment decisions is not straightforward. Gathering, cleaning, processing, and analyzing ESG data demands substantial manual effort, especially for organizations lacking the appropriate technology. Consequently, 59% of asset managers and institutional investors identify data challenges as a significant barrier to ESG integration.
One specific hurdle is the lack of standardization, which contributes to the proliferation of ESG data across varying metrics and structures. Additionally, ongoing changes in ESG-related regulations can lead to shifts in what is deemed material over time.
Asset managers face the critical task of consolidating multiple data sources to address these standardization issues. We’ve observed clients pulling from as few as a couple of sources to over 40, encompassing thousands of entities.
Greenwashing poses another industry-wide challenge, with global financial regulators responding by implementing regulations that mandate more detailed ESG product reporting and labeling. This is evident in Europe through the Sustainable Finance Disclosure Regulation (SFDR) and the Sustainable Finance Roadmap, as well as announcements from the SEC in the United States.
AWS is assisting financial services clients as they navigate this evolving landscape. By leveraging the cloud, it has never been easier to integrate various ESG data sources, regardless of their format, for comprehensive analysis.
Key AWS Services for ESG Data Integration
Two AWS services are particularly effective in addressing the challenges associated with ESG data:
- Amazon FinSpace is a fully-managed data management and analytics service that simplifies the storage, cataloging, preparation, and analysis of financial industry data at scale. This significantly reduces the time financial services firms require to locate, access, and analyze both financial and ESG data.
- AWS Data Exchange allows AWS clients to seamlessly find, subscribe to, and utilize third-party data within the AWS Cloud. Subscribers can access thousands of products from qualified data providers.
Together, Amazon FinSpace and AWS Data Exchange streamline and expedite the financial evaluation and decision-making processes informed by data analysis. This is made possible through the user-friendly capabilities of FinSpace and the diverse data offerings available on AWS Data Exchange.
In this post, we will demonstrate how to utilize AWS Data Exchange and FinSpace for a specific application: automatically identifying optimized asset portfolios based on user preferences (such as minimum and maximum stock percentages) while integrating various ESG datasets from AWS Data Exchange with financial data from third-party sources.
Dataset Overview
In this discussion, we will utilize the following datasets:
ESG Safeguard – Transcripts Sentiment Dataset
This dataset from Amenity Analytics, available through AWS Data Exchange, employs industrial-scale NLP to analyze earnings call transcripts, providing real-time ESG scoring for 12,000 global companies. The scores are calculated based on net sentiment divided by total positive and negative extractions in the transcript per ESG topic. Paid subscriber datasets update daily.
To learn more about AWS Partner Amenity Analytics visit Amenity Analytics.
RepRisk ESG Data Feed: Dow Jones Industrial Average Company Universe
This dataset from RepRisk allows for customizable data exports of their ESG metrics and analytics, easily integrated into internal or third-party systems. The RepRisk ESG Data Feed is designed to systematically monitor ESG and business conduct issues relevant to clients, investments, and suppliers, specifically focusing on the Dow Jones Industrial Average companies.
For more information about RepRisk, visit RepRisk.
Additionally, we will utilize historical stock prices from the Yahoo Finance dataset, which serves as a demonstration tool and can be replaced with another market dataset of your choosing.
For the purpose of this post, you may also use the trial version of the Amenity Analytics and RepRisk datasets; however, please note that these will yield significantly smaller datasets, potentially leading to less informative results.
Architecture and Workflow
- We load the two ESG datasets from AWS Data Exchange into the FinSpace data catalog.
- Accessing the FinSpace environment, we load the two ESG datasets and then integrate the Yahoo Finance dataset. The ESG scores from the two datasets pertain to the same subset of companies but are derived from different providers with distinct criteria; therefore, they cannot be directly compared or combined. We normalize these scores for subsequent combination.
- The Yahoo Finance dataset offers historical stock prices, which we directly load into the FinSpace environment. Using adjusted stock prices, we calculate percentage returns, allowing us to compare stock performance later.
- With the percentage returns and normalized ESG scores in hand, we create a unified working dataset (i.e., dataframe).
- For each stock, we now possess the percentage return and two ESG scores. We apply a business logic to determine the final ESG score for each stock, which will be utilized in the portfolio optimization process.
- We introduce the concept of ESG variance and subsequently discuss ESG min/average variance. This is crucial for understanding how asset allocation might change over time, especially in response to shifting regulatory landscapes.
Personalization Option: During normalization, users can define thresholds or coefficients; you may use the provided values or adjust them to explore various scenarios.
Personalization Option: Users may want to apply their unique approach here to merge these scores. For instance, if a user trusts the Governance scores from a specific data provider, they may choose to incorporate the Governance score from that provider while selecting Environmental and Social scores from others. Alternatively, they might opt to average different ESG scores with varying weights. The logic presented in this post can be modified to suit individual needs or to explore different scenarios.
For readers who may be struggling with payments due to unexpected circumstances, such as the pandemic, we recommend checking out this helpful article on Career Contessa. Additionally, for more authoritative insights on employer obligations, visit SHRM. For a deeper understanding of training methodologies, this article on Harvard Business Review is an excellent resource.