During his 2nd year at Ahmedabad University, Nandish Patel pursuing his BA (Hons) in Economics at the Amrut Mody School of Management, was introduced to Professor Manish Singh from the Indian Institute of Technology (IIT) Roorkee by the economics major advisor, Professor Pallavi Vyas, to explore the possibility of a summer internship. Nandish shares his learnings from the research experience it offered.
When I was introduced to Professor Manish Singh from IIT Roorkee by my major advisor, Professor Pallavi Vyas, I was keen on the prospect of putting into practice the skills and theoretical concepts we were being taught at Ahmedabad University. The opportunity of getting early-hand research experience was exhilarating.
In the summer break of 2022, Professor Singh contacted me and assigned me two tasks, both of which would be the building blocks of our first two projects together. The first task involved designing the framework of an input-output model for the data of new firm formation provided by the Ministry of Corporate Affairs (MCA). This task later evolved into a project of making a data portal for tracking firm formation in India. The second task involved web scraping nested information on insolvent firms from the website of the Insolvency and Bankruptcy Board of India (IBBI). This information was then to be used in the second project of creating insolvency prediction models for firms in India.
After I completed these tasks, I started working on the said projects during my summer break. The curriculum of Ahmedabad University helped me immensely throughout this internship. By my 2nd year itself, I had acquired not only programming skills in statistical computing and GIS (Geographic Information System) but also developed the intuition behind statistical modelling. Over the summer, I developed an interactive web portal for the new firm formation data and started working towards making insolvency prediction models.
Later, in my 5th semester, I was contacted by Professor Singh to work with him on a Portfolio Management Services (PMS) project for a young financial firm called 21 Artha. He introduced me to Mr Shantana Banala, whom I worked under. My role in the project was to write codes for extracting stock price data, making patches to auto-update this dataset, and, thereafter, using it to create interactive visual charts and other tools like Hedge Ratio Button. Through this project, I further acquired advanced programming skills and an understanding of the financial markets.
In the winter break, before my 6th semester started, I got an opportunity to go to IIT Roorkee’s campus for an offline internship experience. There, I continued working toward the second project on insolvency prediction. We used data from two sources – Credit Risk Agencies (CRA) and IBBI – to identify defaults and filter the balance sheet data of firms in India accordingly. Then, we created the required financial ratios and qualitative markers to be used as independent variables in making binary classification models, starting with Logistic Regression. Insolvency prediction models suffer from an imbalanced dataset problem, as the incidence of ‘Default’ is rare in the population of all firms. Despite this, we could correctly predict a significant portion of Defaults (75%), with less than 10% False Positives, against consistent cut-off values for different models. We developed an e-book for this project which is to be made available publicly so that any interested academician or researcher can use our dataset to run their own models and post the results. We also plan to add to the different models, apart from Logistic Regression.
Overall, this internship opened up multi-faceted learning opportunities for me. It provided me with a platform to utilise my existing skills and simultaneously pick up new ones. I am thankful to my university professors for their support and motivation and to Professor Manish Singh for including me in his projects and guiding me through this internship.
Written by Nandish Patel, BA (Hons) in Economics, Class of 2024