Faculty Experience Sharing Interaction/Seminar: Department of Economics and Finance: 2-Jan-2026 @ 11 AM in 6165-I

[hod.ecofin]

Notice

Department of Economics and Finance
BITS Pilani, Pilani Campus 

Prof. Arya Kumar recently delivered a talk on "Reforming India's Financial Economy - Strengthening Financial Access to MSMEs for Vikshit Bharat" at the 108th International IEA conference, held from 21st to 23rd December 2025. The following details are for the Faculty Experience Sharing Interaction scheduled for January 2, 2026, by the Department of Economics and Finance, BITS Pilani, Pilani Campus.

Title of the Talk: Reforming India's Financial Economy - Strengthening Financial Access to MSMEs for Vikshit Bharat
Date: Jan 2, 2026 (Friday)
Time: 11 AM
Venue: 6165 I (HoD Chamber)
Speaker: Prof. Arya Kumar, Senior Professor

Abstract: The realisation of India's 'Viksit Bharat' vision by 2047 necessitates robust participation of Micro, Small, and Medium Enterprises (MSMEs) in manufacturing, export generation, and employment creation. Despite contributing approximately 30 percent to GDP and employing over 100 million individuals, MSMEs confront a substantial financing constraint, with an estimated credit gap of USD 530 billion and merely 14 percent securing access to formal credit channels. This structural impediment fundamentally undermines their capacity to catalyze industrial growth and advance self-reliance objectives under the "Make in India" initiative. Conventional credit assessment methodologies, calibrated primarily for large corporate entities with standardized financial reporting, inadequately capture the heterogeneous and dynamic characteristics of MSME financial health, consequently resulting in either undue risk aversion or elevated portfolio non-performing assets. This study develops and validates a temporally robust machine learning framework designed to enhance credit risk assessment for unlisted firms and MSMEs in India, thereby enabling financial institutions to expand credit

access while maintaining prudential risk management standards. Utilizing financial data from the Insolvency and Bankruptcy Code (IBC) ecosystem spanning 2015-2017 and comprising 306 firms (146 distressed entities liquidated in 2018 and 160 healthy unlisted companies), we construct and benchmark multiple predictive architectures including Logistic Regression, Decision Trees, Random Forest, Support Vector Machines, XGBoost, and Artificial Neural Networks. The study demonstrates that ensemble methods—particularly Random Forest and XGBoost—achieve superior predictive performance (AUC > 0.90, Gini > 0.80) while maintaining exceptionally low Type I error rates (3.8% for XGBoost), thereby supporting the development of MSME-inclusive credit risk monitoring infrastructure in emerging markets.
 
All are cordially invited to attend.

DRC Convener
Department of Economics and Finance, BITS Pilani, Pilani campus