India's Household Gold Is an Asset Class Waiting for Infrastructure

Estimates of Indian household and institutional gold have long clustered around 23,000–25,000 tonnes — a figure a 2025 parliamentary answer priced at roughly US$1.4 trillion. More recent World Gold Council work pushes the estimate higher still, with an August 2026 report discussing around 31,000 tonnes of largely idle household gold. The precise number is genuinely contested. The order of magnitude is not.

And India is already demonstrating appetite for monetising some of it: by May 2026, retail gold loans outstanding stood at roughly ₹5.1 lakh crore at banks and ₹3.3 lakh crore at NBFCs — about ₹8.4 lakh crore combined, per the World Gold Council. That is a large, fast-growing market. It is also a rounding error against the underlying stock.

Today’s gold loan is structurally simple: a household pledges jewellery, a lender takes physical custody, disburses a loan, and returns the jewellery on repayment. It works, and it has scaled into a multi-lakh crore market precisely because it is simple. But it also treats gold the way a pawnbroker does — as something to be temporarily surrendered against a loan, not as a verified asset that lives on a household’s balance sheet.

A different architecture is possible. A household authenticates its gold, deposits it into a trusted repository, and receives a digital asset record — unique ID, weight, purity, fine-gold equivalent, KYC, custody location, valuation history, encumbrance status. The gold itself goes into a secure, audited vault. What the household holds is a claim against a specific, identifiable, insured physical asset — not a pawn ticket.

That distinction matters because once an asset is verifiable and traceable, it can support more than one loan structure. It becomes a platform other financial products can be built on top of — savings, revolving credit lines, term-based financing — rather than a single-purpose pledge that has to be unwound and re-pledged every time.

India already tried something adjacent to this. RBI’s Gold Monetisation Scheme still permits Short-Term Bank Deposits, but the government discontinued the medium- and long-term government components effective March 2025. That withdrawal is instructive, not incidental.

The problem was never that Indian households refuse to monetise gold. It was that GMS asked them to hand it over in a way that felt indistinguishable from selling it — melted, refined, and gone, in exchange for an interest rate that didn’t compensate for giving up a specific, sentimental, inheritable object. The scheme treated gold as a commodity to be mobilised. Households treat it as family wealth to be preserved.

The most immediately workable version of this idea doesn’t require households to hand over new gold at all. It refinances the gold loans that already exist.

Today, a family with ₹10 lakh of jewellery pledged for a ₹7 lakh loan is usually paying a short-tenure, frequently renewed facility at a relatively high rate. A refinancing institution — potentially government-backed — could pay off that ₹7 lakh outstanding balance, move the gold into a certified repository, and convert it into a term-based, reducing- balance facility running three to five years. The household repays on a predictable schedule instead of renewing every few months. The gold stays locked for the committed term and is released once the loan is settled.

The leverage in this structure is that the refinancer never has to buy the gold — only the much smaller outstanding liability against it. And because gold has historically appreciated over multi-year holding periods, the collateral buffer tends to widen over the life of the loan even as the balance amortises down.

Today, gold pledged to one lender is functionally stuck there. Moving it to a cheaper lender means physically retrieving it, closing the loan, and re-pledging elsewhere — friction most households never bother to overcome, even when a better rate exists.

A digital registry breaks that friction the same way dematerialisation did for securities and land titling is slowly doing for property. If a gold asset has a unique ID, a certified fine-gold equivalent, and a recorded encumbrance, a new lender doesn’t need to take possession of the metal to refinance the loan — it only needs to verify the registry entry and update the encumbrance. The gold never has to move for the credit relationship to change.

That’s not a literal payments-rail analogy, but the conceptual shift is the same one UPI made for money movement: a trusted digital representation standing in for a physical or account-based reality, so the underlying asset doesn’t need to change hands every time its economic relationship does.

This is where the idea gets legally serious, and where it has to. A cooperative or credit society cannot tell a household “give us your gold and we’ll issue something equivalent to a bank deposit” without risking exactly the kind of regulatory boundary-crossing that has caused problems elsewhere in Indian finance. A repository cannot casually become a lender either. RBI already sets out detailed requirements around assaying, valuation, loan-to-value limits, ownership verification and secure storage for gold lending — and any infrastructure layered on top has to work within that, not around it.

The workable structure keeps each function in its own regulated lane:

Go back to where this started. India runs a recurring cycle: households earn, buy gold, the gold sits idle, India imports more gold to meet demand, and capital flows out to pay for it — even as the gold itself appreciates and households get nominally wealthier without the asset ever becoming economically productive. Reducing reliance on gold imports was the explicit rationale behind GMS when it launched. The objective was right. The instrument didn’t fit how Indians actually hold gold.

A refinancing-led, registry-backed version of this idea reopens that objective through a different door. As existing gold loans convert into longer-duration, amortising structures, and as households repay, the freed-up collateral capacity doesn’t have to flow straight back into more consumer credit. A portion of it could fund a second, separate credit channel aimed at import-substituting domestic manufacturing — machine tools, components, batteries, industrial equipment — priced on the strength of that manufacturer’s economics, not political direction.

That second channel only works if it stays disciplined: gold provides the collateral base and funding cost advantage; competitive, transparent underwriting decides which manufacturers actually get financed. Mix the two up and you’ve recreated directed lending under a new name.

Consider a household holding ₹8 lakh of family jewellery against ₹1 lakh in bank savings. Today, the formal financial system reads that household through its bank balance and credit history — both thin. An asset-verified registry reads the same household through ₹8 lakh of authenticated collateral, plus whatever income and repayment history it can add on top. That’s a materially different financial identity, and it’s the identity that matters for the people this would help most: farmers, small business owners, women-led households, and self-employed and informal- sector workers who own real wealth but don’t currently show up as creditworthy on paper.