ESG Analytics
Digital Bucket Company
The purpose of our ESG tool was established after discovering the important role Environmental, Social, and Governance (ESG) indexes are benchmarking a companies corporate, social responsibility practices. Added to that, is the growing regulatory taxonomy frameworks that need to provide clear measurements to the level of ESG impact.
The financial services there is common consensus future investment need to be centred around the ESG framework. However, with any financial investment with high liquidity that are normally the backbone to pension funds or sovereign funds the level of risk to diversifying to alternative products is critical. Our research indicates there are three risks that are essential to understand through various data modelling processes and analysed in ESG investments are:
1. Environmental: Examines how the company has an impact on the ecosystem, the planet, and human health.
2. Social: Ensures that wages are fair, health and safety considerations are acted upon, and that human rights are enforced.
3. Governance: it is an assessment of management and leadership practices, i.e avoid excessive boardroom pay. U.S. assets under management using ESG strategies grew to $17.1 trillion at the beginning of 2020. A 42% increase from $12 trillion at the beginning of 2018 (United Kingdom Economy, 2021). Research shows that in 2020 funds that embraced the ESG criteria out-performed their competitors, with the growth averaging around 27%.
The dynamic and often varying ESG standards set around the world requires a data platform that can adopt and analyse data according to the different regulatory and market environments. We took the opportunity as a data consultancy to explore whether a tool could be developed that provides the answers to the ESG investment market.
Developing ESG Investment Strategy
A single collation of all ESG index funds.
There is currently no single collation of all ESG index funds. There is the need for a software artefact that gathers index fund data and manages it under a singular system. This poses a few challenges:
● The data will need to be sourced from an estimated 800 indices. This data will vary in format for each fund and subsequently will need harmonising.
● There will be large volumes of dynamic data which will require being processed on-the-fly. To solve this, A scalable data pipeline will need to be developed which will feed into an analytics engine capable of processing big data in order to produce real time results.
These two challenges are not unprecedented in the general domain of big data, thus, there are existing software products to utilise to reduce the development workload.
Specification Functional Requirements
1. Data Ingestion. Data will need to be collected from an estimated 800 index funds at regular intervals.
2. Clean data. Perform data profiling so that only relevant data is stored. This will reduce the computational workload.
3. Harmonize Data. The data should be stored under a standard format in order for comparisons to be made
4. Store Data. The data should be stored in a central repository, allowing for the examination of historic data.
5. Execute Financial Model. The harmonized data should be inputted to a financial model with the results saved to a central repository.
6. Data Visualization for Financial Analysis. A visualization of the results of the financial model should be displayed to the user on a web page.
7. Authenticate Users. Only authorized users should be allowed to view the data at any stage of the process.
Non-Functional Requirements
1. Scalability: The data pipeline should be scalable so that any number of additional index funds can be added without the architecture crumbling.
2. Real-Time: The data should be collected and analyzed in real time.
3. Security: The data should be secure at all stages within the software artifact.
Evaluation Criteria
1. Data ingestion should be automated and occur for the specified interval. Data should be able to be ingested from any reasonable source.
2. Data cleaning should eliminate all unnecessary data so that the same fields are collected from every API, or all fields are derivable from the cleaned data.
3. Data harmonization should ensure that all data is stored in a consistent standardized format. There should be no disparity in the way data is stored between indexes.
4. Data storage should store data together, store historical data, and should not be erased.
5. Financial analysis should be performed which allows a user to draw comparisons between index funds. 6. Visualization of the financial analytics should be considered intuitive and insightful when surveyed on users.
7. Authorization in the form of username and password needs to be given to Users to access the web client.
8. Scalability should be paramount so that additional APIs can be added without an architectural redesign.
9. Real-time analytics results will be shown within five minutes of the data being ingested. 10. Security should ensure only authorized users have access to the data within the pipeline.
Financial Dashboard To turn the results of the financial modeling into easily digestible information, a user friendly dashboard was created using Tableau. The software offered by Tableau enabled the quick and easy development of an interactive data visualization tool.
The functionalities being offered by the dashboard are:
1) Users can filter through different portfolios and risk factors
2) Users can click on a specific portfolio on the dashboard and any corresponding information about the portfolio anywhere on the dashboard will also be highlighted.
3) Users can easily export the numerical data of the risk factors and portfolios
4) Users can easily access any documents on a specific portfolio by clicking the portfolio.
