There are around 100-150 odd Carbon/Climate related ESG Metrics that are on-boarded via different vendor sources by ESG Tech, some of them calculated internally, available at issuer level. These data points are very scattered, not very useful to PMs/Analysts in the raw format, or at the company/issuer level. PMs/Research analysts want to view/compare these Metrics at the portfolio level, or benchmark level.

ESG reporting via Power BI – ESG Data model built in Power BI to consolidate all the 800 odd ESG metrics

• Carbon/Climate Report - One stop shop to all the Carbon/Climate related ESG Metrics, showcased at the security/portfolio level, for PMs/ESG Research Analysts/Credit Analysts to refer to. For example, one of the comparison in the report showcases the comparison of Weighted Carbon Intensity(which is the average of the issuer’s greenhouse emissions compared to its overall revenue - calculated in Power BI) at account vs. it’s corresponding benchmark, or for that matter any other benchmark(the PMs want to view)

Similarly this report showcases multiple comparisons utilizing these 100 odd Carbon/Climate ESG Metrics , and greatly assist PMs/Credit Analysts to perform Scenario Analysis(buying/selling securities, and the impact on the overall portfolio on say selling a security with High Carbon Emissions, replacing it with a security with comparatively low Carbon Emissions/Footprint

One of the bigger challenges was how to handle the large data size in PowerBI, bringing together data for 600K odd securities across 800 odd ESG Metrics on a day in day basis. As per the Business ask, Data is run on a daily basis for 4 dates(Last Business Day, Last Month End, Last Quarter End & Year End).

Rather than fetching this huge chunks of data from On-PREM database sources, ultimately impacting the PowerBI performance, we adopted the approach to generate the data for the 4 dates through Python scripts. Four different files are created on a daily basis through these Python scripts, and the files are stored on Cloud. PowerBI then calls and fetches the data from these 4 files through API calls. The whole process of data ingestion is completely automated.