Bank or NBFC for a Gold Loan? Borrower Choice in Coimbatore and Palakkad after RBI's 2025 Gold-Lending Directions

In Tamil Nadu and Kerala, a gold loan is the most ordinary kind of credit: school fees in June, a crop input bill before the monsoon, a hospital deposit at midnight. Households pledge jewellery with a public-sector bank, a cooperative bank or one of the large gold-loan NBFCs, and many switch between them. In June 2025 the Reserve Bank of India issued the Reserve Bank of India (Lending Against Gold and Silver Collateral) Directions, 2025, which apply to all regulated lenders from 1 April 2026: tiered loan-to-value (LTV) caps of 85% up to ₹2.5 lakh, 80% from ₹2.5–5 lakh and 75% above, standardised valuation and purity certificates, regional-language documentation, and compensation for delays in returning pledged gold.

This dissertation studies how borrowers choose between banks and NBFCs for gold loans, and whether the new directions change the basis of that choice. It is carried out with a fictional urban cooperative bank, Noyyal Valley Co-operative Urban Bank, which wants to win gold-loan business back from NBFCs in Coimbatore district and across the border in Palakkad.

About 300 borrowers are surveyed on speed, documentation, LTV, interest, hidden charges, branch convenience, trust and auction experience. Eight branch managers are interviewed. SPSS is used for chi-square, ANOVA and binary logistic regression, and an Excel model compares the effective annual cost of a typical ₹1 lakh loan across lenders. The report follows the Anna University MBA Regulation 2021 Project Work scheme with three reviews.

Abstract

This study examines factors influencing borrowers' choice between banks and non-banking financial companies for gold loans in Coimbatore (Tamil Nadu) and Palakkad (Kerala), in the context of the RBI Lending Against Gold and Silver Collateral Directions, 2025. A survey of about 300 gold-loan borrowers and interviews with eight branch managers are analysed using chi-square tests, ANOVA and binary logistic regression. An effective-cost comparison quantifies the price of speed. The study recommends how a cooperative bank can compete for gold-loan customers under the harmonised rules.

Introduction

Gold loans are secured, quick and familiar, which makes them the first choice for emergency and short-term credit in South India. Banks generally advertise lower interest rates, while NBFCs emphasise speed, longer hours and higher LTV. Supervisory concerns about irregular practices led RBI to issue a circular in September 2024 asking lenders to review their gold-loan policies, followed by draft directions in April 2025 and final directions in June 2025 that bring banks, NBFCs and cooperative banks under one framework.

Literature gap

Indian studies on gold loans are mostly descriptive surveys of NBFC customer satisfaction, or financial analyses of listed gold-loan companies. Few compare bank and NBFC borrowers in the same area, and none we found examines borrower choice after the 2025 directions narrowed differences in LTV and procedure. The border region between Coimbatore and Palakkad allows a comparison of two states with very high gold-loan penetration.

Proposed study

• Identify factors that drive the choice of lender type.
• Compare satisfaction and auction experience across bank and NBFC borrowers.
• Measure awareness of the new LTV caps and borrower protections.
• Estimate the effective annual cost of a typical loan under each lender's rate card.
• Recommend a competitive gold-loan proposition for the host bank.

Feasibility

• Operational: the host provides access to its own borrowers and introduces the student at a gold-traders' association for NBFC borrowers; 16 weeks of fieldwork fit the Regulation 2021 schedule.
• Technical: SPSS and Excel are used in the programme.
• Ethical: no loan account numbers, Aadhaar, PAN or jewellery details recorded; consent in the respondent's language.

The host cooperative bank has lower advertised gold-loan rates than nearby NBFCs, yet its gold-loan book has stagnated while NBFC branches in the same streets are crowded. Managers believe borrowers prefer NBFCs for speed and higher LTV, but they do not know how much weight borrowers place on each factor, whether borrowers compare total cost or only the headline rate, or whether the 2025 directions, which cap LTV for every lender and standardise valuation, have changed these preferences. The bank also does not know whether borrowers are aware of new protections such as purity certificates and compensation for late return of gold. Without this evidence, its marketing repeats "lowest interest rate" and fails. The problem for this study is to identify the determinants of lender choice for gold loans in a high-penetration region, quantify the cost of convenience, and assess borrower awareness of the new regulatory protections.

The study covers individual borrowers who took or renewed a gold loan in the last 12 months from a public-sector bank, private bank, cooperative bank or NBFC in Coimbatore district and Palakkad district. It excludes agricultural gold loans under special schemes and business borrowers with loans above ₹10 lakh. Cost comparisons use published rate cards and charges, not individual loan records. The study describes regulation for context and awareness measurement, not as legal advice.

Research design

Descriptive and analytical design with a borrower survey, branch-manager interviews and document analysis of the RBI directions, the September 2024 RBI circular on gold loans and lenders' published rate cards.

Sampling

• Population: gold-loan borrowers in the two districts (very large; treated as infinite).
• Sample size: Cochran (p = 0.5, e = 0.057, 95%) gives about 296; target 300, split Coimbatore 180 and Palakkad 120.
• Method: quota sampling by lender type (bank 150, NBFC 150) and district, with respondents intercepted outside branches in six market areas at different times of day, plus the host's own borrowers for the bank quota. Quotas are filled across age and gender bands.
• Interviews: 8 branch managers (4 banks, 4 NBFCs) using a semi-structured guide on process time, LTV practice, auction policy and preparation for the directions.

Instrument

Pilot n = 30; Cronbach's alpha ≥ 0.70 for Sections C and D.

Hypotheses and tests

Cost model

For a ₹1 lakh loan held for 6 and 12 months, compute the effective annual cost (interest + processing fee + valuation charge + any renewal fee) from each lender's published rate card, and the maximum loan available on 20 g of 22-carat jewellery under the new 85% LTV cap.

Timeline (20 weeks)