Technology

Gold Loan Management Software: RBI 2026 Rules in Code

|Posted by Hitul Mistry / 12 Nov 26

Gold Loan Management Software: Valuation, LTV Monitoring and Auction Workflows for NBFCs

Last reviewed: November 2026

Gold loan management software is the system that records, values, monitors and closes loans secured by gold and silver jewellery. In 2026, it is also the main way an NBFC proves it follows RBI's harmonised rules on lending against gold and silver collateral.

This guide is for CEOs, COOs and CROs at gold loan NBFCs and diversified lenders with a gold book. It turns each rule into a system control and shows the LTV arithmetic your software must get right.

Key Takeaways

  • RBI's gold and silver directions of June 2025 now sit in Chapter IV of the NBFC Credit Facilities Directions, 2025, with conduct rules in the Responsible Business Conduct Directions, 2025. The compliance deadline was 1 April 2026.
  • LTV caps are tiered at 85%, 80% and 75% by loan size, and bullet loans count interest due at maturity.
  • Valuation uses the lower of the 30-day average and previous-day closing price for that purity.
  • Release within seven working days is enforced by ₹5,000 per day compensation, and auctions need a 90% reserve price.

Which RBI Rules Must the Software Enforce?

Nearly every gold loan rule now maps to a field, a calculation or a workflow step. The table pairs each rule with the control that enforces it.

RuleSource (Paragraph)Software Control
Valuation at lower of 30-day average or previous-day closeCredit Facilities, para 40Daily price feed by purity from IBJA or an exchange
Stones and gems excludedCredit Facilities, para 42Net weight field after deductions
LTV tiers 85/80/75% by total loan per borrowerCredit Facilities, para 43Borrower-level exposure aggregation before sanction
Bullet LTV on amount due at maturityCredit Facilities, para 43LTV numerator includes scheduled interest
Credit assessment above ₹2.5 lakhCredit Facilities, para 33Mandatory income and repayment capacity step
Bullet consumption loans up to 12 monthsCredit Facilities, para 38Tenor cap in product configuration
Renewals and top-ups identifiableCredit Facilities, para 34Separate transaction codes and reports
Quantity caps per borrowerCredit Facilities, para 39Cumulative weight check across all loans
Borrower present at assayingResponsible Business Conduct, para 46Assay record with photo and acknowledgement
Pledge certificate contentsResponsible Business Conduct, para 50Auto-generated certificate with images and deductions
Release within seven working daysResponsible Business Conduct, paras 58, 69Release clock with ₹5,000 per day liability tracker
Auction reserve price 90%, then 85%Responsible Business Conduct, para 63Reserve price locked from valuation engine
Surplus refunded within seven working daysResponsible Business Conduct, para 66Surplus payable with due-date alerts
Unclaimed after two yearsResponsible Business Conduct, paras 71–72Unclaimed register and half-yearly report

Paragraph numbers reflect the RBI pages as updated in 2026. Confirm them before citing in policies, as amendments can renumber clauses.

How Should Valuation and LTV Work in the System?

Valuation must be automatic and reproducible. The system should store the price source, both candidate prices and the one applied for every loan, every day.

For bullet loans, the LTV numerator is the total amount repayable at maturity, not just principal. The tier is set by the borrower's total consumption loans, so the software must aggregate exposure across loans and branches.

Illustrative LTV Calculation

The figures below are illustrative, with assumed prices that are not market quotes. Interest is simple interest for a 12-month bullet loan.

A borrower pledges 22-carat ornaments with a net weight of 37 grams after deductions. The 30-day average price is ₹11,200 per gram and the previous day's close is ₹11,050, so the system applies ₹11,050.

Collateral value is 37 × ₹11,050, which equals ₹4,08,850. The borrower asks for ₹3,20,000 at 12% a year, so the amount due at maturity is ₹3,58,400.

The loan sits in the above-₹2.5-lakh to ₹5-lakh band, so the cap is 80%. LTV is ₹3,58,400 divided by ₹4,08,850, or about 87.7%, which breaches the cap.

At 80%, the maximum amount due at maturity is ₹3,27,080, so the maximum principal is ₹3,27,080 divided by 1.12, or about ₹2,92,036. Because the loan exceeds ₹2.5 lakh, a credit assessment is also required.

Now assume prices fall 8% after disbursal. Collateral value becomes ₹3,76,142 and LTV rises to about 87.0%, so the borrower needs to pay about ₹26,166 or add collateral to return to 80%.

The lesson for leadership: branch staff cannot do this arithmetic reliably at the counter. The engine must calculate eligibility, block over-limit sanctions and flag breaches the morning prices move.

Can your system prove the price, purity and LTV behind every gold loan you sanctioned this year?

Talk to Our Lending Technology Specialists

Visit Digiqt to see how we build secured lending journeys with valuation and LTV checks built in.

What Does the Branch-to-Vault Workflow Look Like?

Gold loans are physical products run through digital records. The workflow has five custody points, and each needs a timestamped record.

  1. Appraisal. Assay in front of the borrower, record purity, gross and net weight, deductions and photos.
  2. Sanction and certificate. Issue the pledge certificate in duplicate and capture the borrower's acknowledgement.
  3. Packeting and vault. Seal the packet, record the seal number and log vault entry under dual control.
  4. Audits. Record periodic and surprise checks, with the borrower's prior consent for checks in their absence.
  5. Release. Match the packet to the certificate, release on closure and close the release clock.

Digiqt's collateral valuation AI agent shows how valuation exceptions, such as unusual purity readings by branch, can be surfaced for review.

How Should Auctions Be Managed?

Auctions are the most scrutinised step in gold lending. The software should treat each auction as a case with a checklist that cannot be skipped.

Before an auction, the system needs proof of adequate notice to the borrower and, where the borrower cannot be traced, a public notice at least one month earlier. It should also hold the two newspaper advertisements, one regional and one national.

During and after the auction, the system should lock the reserve price from the valuation engine and block related-party bidders. It should then post proceeds, compute any surplus and pay it within seven working days.

Venue rules belong in the workflow too. The first auction must be held physically in the same district as the lending branch, and only after a failed first auction can the NBFC move to an adjoining district or an online auction.

Reminder and notice journeys can be automated. Our guide to voice agents in digital lending covers how multilingual reminders can reach borrowers before an account reaches auction, and the early-stage collections AI agent shows how outreach can be prioritised.

What Should Leadership See Every Morning?

A CEO or CRO dashboard should answer one question: where is the gold book exposed today? The measures below come straight from the controls above.

MeasureWhy It Matters
Loans above their LTV tier, by branchShows breach exposure after price moves
Bullet loans maturing in the next 30 daysDrives renewal and recovery planning
Releases pending beyond five working daysEarly warning before compensation applies
Auctions scheduled, with notice statusFlags cases missing evidence
Surplus refunds outstandingTracks the seven-working-day duty
Unclaimed packets older than two yearsFeeds tracing drives and board reporting

When you evaluate software, ask vendors to demonstrate these views on your own sample data. A system that cannot aggregate borrower exposure across branches, or store the daily price used for each loan, will struggle with the 2025 rules however good its counter screens look.

Which Gold Loan System Mistakes Create Regulatory Risk?

These gaps are common in older gold loan systems built before the 2025 directions.

  • Branch-level LTV checks. Tiers depend on total borrower exposure, which a single branch cannot see.
  • Principal-only LTV on bullet loans. Ignoring interest due at maturity overstates headroom.
  • No release clock. Without one, ₹5,000-per-day liabilities build up unseen.
  • Manual auction files. Missing notices or reserve price evidence are hard to defend in a complaint.

What Is Changing Next?

The 1 April 2026 compliance date has passed, but loans sanctioned before a lender adopted the directions remain under earlier rules until they run off. Your system must tag each loan by regime so reports and checks apply the right rules.

Watch for clarifications and amendments on RBI's website, as gold lending has been updated more than once since June 2025. If your current core cannot support these controls, a purpose-built digital lending platform may cost less than ongoing patches.

Want auctions, releases and LTV breaches tracked automatically across every branch?

Talk to Our Lending Technology Specialists

Visit Digiqt to see how we build lending platforms and AI agents for secured lenders.

Sources

Read our latest blogs and research

Featured Resources

Technology

Loan Against Mutual Funds Platform: LAMF Launch Guide

Loan against mutual funds platform guide for NBFCs: RTA lien marking, RBI LTV rules, NAV monitoring, margin calls and launching a fully digital LAMF product.

Read more
Technology

Loan Against Property Underwriting: Faster LAP Checks

Loan against property underwriting for NBFC credit teams: run legal, technical and income checks in parallel, automate safely and keep sanction in-house.

Read more
Technology

Repossession Management Software for Vehicle Finance

Repossession management software for vehicle financiers: RBI recovery rules from January 2027, yard and auction cost control, and a worked recovery model.

Read more

About Us

We are a technology services company focused on enabling businesses to scale through AI-driven transformation. At the intersection of innovation, automation, and design, we help our clients rethink how technology can create real business value.

From AI-powered product development to intelligent automation and custom GenAI solutions, we bring deep technical expertise and a problem-solving mindset to every project. Whether you're a startup or an enterprise, we act as your technology partner, building scalable, future-ready solutions tailored to your industry.

Driven by curiosity and built on trust, we believe in turning complexity into clarity and ideas into impact.

Our key clients

Companies we are associated with

Life99
Edelweiss
Aura
Kotak Securities
Coverfox
Phyllo
Quantify Capital
ArtistOnGo
Unimon Energy

Our Offices

Ahmedabad

B-714, K P Epitome, near Dav International School, Makarba, Ahmedabad, Gujarat 380051

+91 99747 29554

Mumbai

C-20, G Block, WeWork, Enam Sambhav, Bandra-Kurla Complex, Mumbai, Maharashtra 400051

+91 99747 29554

Stockholm

Bäverbäcksgränd 10 12462 Bandhagen, Stockholm, Sweden.

+46 72789 9039

Malaysia

Level 23-1, Premier Suite One Mont Kiara, No 1, Jalan Kiara, Mont Kiara, 50480 Kuala Lumpur

Lewes

16192 Coastal Highway, Lewes, Delaware 19958, USA

software developers ahmedabad
ISO 9001:2015 Certified

Call us

Career: +91 90165 81674

Sales: +91 99747 29554

Email us

Career: hr@digiqt.com

Sales: hitul@digiqt.com

© Digiqt 2026, All Rights Reserved