The Loan You Pay Last: Are Gold Loans Becoming India's Bad Bank?
By Deepak Parkash (CFO, indiagold) · 2026-10-20 · 11 min read
Two lenders held their quarterly earnings calls on the same day this week. Piramal Finance went first, on the afternoon of October 19. Bajaj Finance followed in the evening. In fact, Piramal's CEO Jairam Sridharan signed off his call by wishing analysts a happy Dussehra and telling them to enjoy Bajaj's call.
On the face of it, both calls told a happy story. Profits were up. Loan books were growing. Bad loans in home loans, business loans and personal loans were under control. If you only read the headlines, you would think the Indian borrower is in great shape.
But tucked away in both calls was a pattern that is worth paying attention to. And that pattern had everything to do with gold loans.
The question this piece tries to answer is a simple one. Is financial stress in the Indian economy quietly moving into gold loans, because that is the one place where it does not show up for a while? Let me try and explain.
What Piramal saw
Piramal Finance is not a big gold lender. It had just 100 gold loan branches at the end of September. Its business is mostly home loans, loans against property and unsecured loans. That makes what it said about gold loans more interesting, not less. It is looking at gold loans from the outside, through the credit files of its own customers.
Here is what Sridharan told analysts on the Q2 FY27 call. Two years ago, depending on the product, 20-25% of Piramal's customers had an active gold loan with some other lender. That number is now above 35%, and it is above 35% in every product category. So more than one in three people who have a home loan or a personal loan from Piramal have also pledged their gold somewhere.
He had a second data point, and it is the more uncomfortable one. In 2025-26, around ₹4.7 lakh crore of gold loans across the industry were given to borrowers who were already more than 90 days late on some other loan. Depending on how you count it, that is anywhere between 15% and 23% of all gold loans disbursed over the past year. Sridharan himself called it a shockingly high number.
Let me explain what 90 days late means. Lenders track how many days a borrower has gone past the date a payment was due. This is called days past due, or DPD. A borrower who is 90 days or more past due is, for all practical purposes, a bad loan. The lender has stopped expecting to be paid on time.
So, roughly one in every five gold loans in the past year went to someone who had already stopped paying another lender. Why would someone in that situation take a gold loan? The most likely answer is to buy time.
To be fair, Sridharan also said that having a gold loan on file is not yet a sign of higher risk in Piramal's own book. But he called it an important emerging risk worth tracking. When a lender that is otherwise talking about record profits goes out of its way to flag something, it is worth listening.
What happened at Bajaj Finance
Bajaj Finance is a newer entrant in gold loans, but it is moving fast. Its gold loan book stood at a little over ₹24,000 crore at the end of September, about 4.1% of its balance sheet. It added around 320 gold loan branches in the quarter alone and plans to have about 2,800 by March 2027.
The interesting bit was in the asset quality numbers. Before getting there, a little jargon needs to be explained.
Lenders sort their loans into three buckets. Stage 1 loans are being repaid on time. Stage 2 loans are those where the borrower has started slipping, typically 30 to 90 days late on a payment. Stage 3 loans are 90 days or more late, and are bad loans. Stage 2 is the waiting room. Some borrowers walk back to Stage 1. Others walk on to Stage 3.
In the September quarter, Bajaj's Stage 2 loans in the gold business rose by about ₹1,100 crore. On a gold book of around ₹24,000 crore, that is roughly 4.5% of the book sitting in the waiting room. For a product that is backed by gold and is supposed to be among the safest in lending, that is not a small number.
Bajaj's management was candid about why this happened. When the RBI's new gold loan rules came in this April, Bajaj, like much of the industry, launched three-month gold loans, with the interest payable at maturity. Under the rules, the 75% loan-to-value cap includes the interest due over the loan's tenure, so a shorter tenure lets a lender hand out a bigger loan against the same gold. The catch was that the loan fell due every three months, and the customer had to come to the branch to repay it, or pay the interest and roll it over. As
CFO Sandeep Jain explained, many customers found this cumbersome and accidentally slipped into delinquency. Once in delinquency, the only way out was to repay the full principal and interest.
The company stopped offering short-tenor schemes in August. Its new gold loans run for 12 months, with the principal due at maturity and interest due only once every six months. A Bajaj gold loan receipt from October shows exactly this: one interest instalment at month six, a second at month twelve along with the principal. Customers are nudged to pay earlier through a rebate: pay the interest within 90 days and the rate drops from 24.25% to 20% a year. Rajeev Jain noted that much of the industry still offers three-month loans. And management said that of the roughly ₹1,100 crore that slipped into Stage 2, about ₹550 crore has already come back to Stage 1.
Now, Bajaj's explanation is that this was about friction, not financial stress. That may well be true for many of these customers. But think about what the episode tells us. When a quarterly payment was made compulsory, a meaningful chunk of gold loan borrowers did not pay on time. When the compulsion was removed, the problem went away. As MD Rajeev Jain told analysts, 99% of Bajaj's gold loan customers are on bullet repayment.
In other words, the gold loan customer has a strong preference for not having to pay anything until the very end. Hold on to that thought.
The puzzle: a stressed economy with clean loan books
Look at the other loan books and things seem fine. At Bajaj Finance, loan losses as a share of average assets improved to 1.56% in the September quarter, from 1.88% a year earlier. At Piramal, retail loans that are 90 days or more overdue stand at 0.7%, and that number has stayed in a narrow 0.6-0.8% band for four years. Piramal said its unsecured loans are showing risk at multi-year lows. Home loans, business loans and personal loans, broadly, look okay.
Now look at the economy around these borrowers.
First, oil. India imports most of the crude oil it uses. Brent crude, the global benchmark, crossed $100 a barrel in March 2026, spent much of the year between $90 and $110, and went back above $100 on September 9. It was trading around $110 earlier this week. Expensive oil works its way into everything: petrol, diesel, transport, food. Kotak Securities expects retail inflation to rise towards 6% if crude stays above $100. The RBI has already moved from a neutral stance to raising interest rates, which means EMIs on floating-rate loans will go up.
Second, the IT sector, which for three decades has been the great creator of middle-class jobs. The top IT firms together ended 2025-26 with over 7,000 fewer employees than they started with, as TCS cut around 12,000 jobs. TCS plans to hire around 25,000 freshers this year, against 40,000-42,000 in earlier years. Salary hikes have been muted. Piramal itself flagged continuing stress in home loans to IT employees in South India.
Third, the monsoon. Bajaj's management said rainfall came in at 87% of the long-period average and that Maharashtra was the most affected state. The company set aside an extra ₹337 crore as a precaution against the macroeconomic environment, with CFO Sandeep Jain pointing to the weak monsoon and oil hovering at $100-110 a barrel.
So here is the puzzle. Household budgets are under pressure from expensive fuel, slower job growth and rising interest rates. That stress should show up as rising late payments across many kinds of loans. Instead, the stress is barely visible in home loans, car loans, business loans and personal loans. The one place it is showing up is gold loans.
Why?
Why stress hides in gold loans
The answer lies in how a gold loan is structured, as opposed to almost every other loan an Indian household takes.
A home loan, a car loan or a personal loan is repaid through an EMI, an equated monthly instalment. Every month, a fixed amount leaves the borrower's bank account. Part of it pays the interest, part of it pays back the principal, which is the money originally borrowed. The moment a borrower misses an EMI, the lender knows. The loan starts counting days past due from the very next day. Stress shows up in a month.
A gold loan is different. Most gold loans are bullet loans. In a bullet loan, the principal is repaid in one shot at the end of the loan's term, typically 12 months. Nothing towards the principal is paid in between.
And it goes further than that. Bajaj Finance's new gold loans, for instance, ask for the first interest payment only at month six. So for a large share of gold loan borrowers, nothing at all is due for the first six months. Not the principal, not the interest. The interest keeps piling up quietly in the background, but no payment is asked for.
Now put yourself in the shoes of a borrower who is feeling the pinch. His monthly expenses have gone up. His salary has not. He has an EMI on his home loan and another on his car loan. If he misses either, there is a phone call the next week, a late fee, a hit to his credit score, and eventually the possibility of losing the house or the car.
The rational thing for him to do is to keep paying the EMIs that bite the quickest, and find money from somewhere else to bridge the gap. The gold at home is the obvious somewhere else. He pledges it, gets a loan, and does not have to pay a rupee for six months.
Even at the end of six months, all he has to pay is the interest that has built up. The principal is still not due. He can then often renew the loan and start the clock again.
So the stress does not go away. It gets deferred, and it gets shifted from one product to another. The home loan book looks clean. The car loan book looks clean. And the gold loan book quietly absorbs the strain, which will only become visible when the interest or the principal finally falls due.
A household's juggling act
Let us make this concrete with an illustrative example. The numbers are made up, but the behaviour is not.
Meet Ravi, who works at a mid-sized IT services firm in Pune. He pays an EMI of ₹40,000 a month on his home loan and ₹12,000 on his car loan. His salary has not gone up in two years. His fuel bill, his children's school fees and the family's grocery bill have. Every month he is about ₹20,000 short.
Ravi has a few choices. He can skip an EMI, but that means a call from the bank, a late fee and a damaged credit score. He can take a personal loan, but that comes with yet another EMI starting next month, which only makes the monthly squeeze worse.
Or he can pledge his family's gold. Say he borrows ₹3 lakh at 18% a year on a 12-month bullet loan where interest is also deferred for six months. Here is what his monthly outgo looks like.
| Period | Home loan EMI (₹) | Car loan EMI (₹) | Gold loan payment (₹) |
|---|---|---|---|
| Months 1-5 | 40,000 | 12,000 | 0 |
| Month 6 | 40,000 | 12,000 | about 27,000 (six months' interest) |
| Months 7-11 | 40,000 | 12,000 | 0 (or interest, if paid monthly) |
| Month 12 | 40,000 | 12,000 | 3,00,000 principal + remaining interest, or renew |
For five months, the gold loan asks for nothing. Ravi uses the ₹3 lakh to keep paying both EMIs on time. His home loan lender sees a perfect borrower. So does his car loan lender. Both report him as a clean customer to the credit bureau.
At month six, he has to find about ₹27,000 in interest. That is a meaningful sum, but it is a fraction of what he owes elsewhere, and it buys him another six months. At the end of the year, if things have not improved, he renews the loan or takes a fresh one, perhaps from another lender.
Now multiply Ravi by a few lakh households. The stress in the economy is real, but on the books of lenders it has been moved into the one product where it does not show up for six to twelve months.
Bajaj's U-turn: right for Bajaj, wrong for the segment
This brings us back to Bajaj Finance's decision to drop its short-tenor schemes and return to 12-month bullet loans. On the face of it, the move makes perfect sense. Look closer, and it is a good decision for Bajaj that may not be a good one for gold loans as a whole.
Let us start with why it makes sense for Bajaj. In India, a lender has to classify bad loans borrower by borrower, not loan by loan. If any one loan of a borrower turns into a non-performing asset, or NPA, which is a loan that has gone unpaid for 90 days or more, then every other loan that borrower has with the same lender is also treated as an NPA. This holds even if he is paying those other loans on time. And to be upgraded back to a good borrower, he has to clear all arrears across all his loans with that lender. The RBI's revised norms, which kick in from April 1, 2027, keep this borrower-level rule at their core.
Now think about what this means for a diversified lender like Bajaj. It has a cross-sell franchise of 82 million customers. Many of its gold loan customers also have a consumer durable loan, a personal loan or a credit card from it. If such a customer misses a payment on his gold loan and stays behind for 90 days, it is not just a small gold loan that turns bad. Every loan he has with Bajaj turns bad with it, even the ones he is paying diligently. A slip in a ₹2 lakh gold loan can drag a much larger relationship into the NPA column. Rajeev Jain said as much on the call: for a lender that is not a monoline, a customer who slips into delinquency on a gold loan cannot then be lent to for any other product either.
A bullet loan solves this problem neatly. When nothing is due for six months, nothing can be overdue for six months. Under Bajaj's new structure, the earliest a gold loan can turn into an NPA is about nine months after it is given. For a lender whose real business is the rest of the balance sheet, that is a very attractive feature.
But here is the catch. Bullet repayment is exactly the structure that allows subordination of gold loans: the borrower keeps paying the loans that bite every month and parks his stress in the loan that asks for nothing. Bajaj's own experience shows this. When its loans fell due every three months, about ₹1,100 crore slipped into Stage 2, and within weeks about half of it was back on track. That is not a failure. That is an early-warning system doing its job. It flagged which customers were struggling while there was still time to act. Going back to bullet loans does not remove the stress. It removes the alarm. And there was a middle path. A three-month tenure is the harshest form of regular repayment: miss the date and the entire loan is in default, repayable in full. A 12-month loan with compulsory monthly or quarterly interest keeps the habit and the early warning, while a missed instalment can be caught up without repaying everything.
There is a longer-term cost as well. A regular repayment schedule, monthly or quarterly, builds a habit. The customer budgets for it, the way he budgets for his EMI and his electricity bill. The lender, in turn, gets a running read on the borrower's health every month, rather than a single verdict at the end of the year. Over time, this is what turns gold loans from an emergency product into a disciplined form of credit that the rest of the system can trust.
To be clear, this comes at a price. A monthly or quarterly scheme will look worse in the short term. More payments fall due, so more of them are missed, and Stage 2 numbers go up, just as they did at Bajaj this quarter. A bullet book, by contrast, looks spotless until the day it doesn't. But the early slippages in a regular-repayment book are mostly customers learning a new habit, and many of them come back on track, as half of Bajaj's did within weeks. Over a few cycles, the customers who stay are the ones who pay, and the book gets steadily cleaner. It is short-term pain for long-term gain. The bullet offers the opposite: short-term comfort, with the bill arriving later and all at once.
Bajaj does nudge customers to pay sooner, with a rate rebate for paying interest within 90 days. But a rebate is not the same as a structure. The borrowers most likely to take it up are the ones who need it least. The borrower who is quietly pushing his gold loan to the back of the queue will, quite rationally, pick the bullet.
This is where specialist gold lenders differ. For a lender whose main product is the gold loan, a missed gold loan payment does not contaminate a home loan or a credit card on the same books. Such a lender can afford regular repayment structures, and the segment is better off when it uses them. When the biggest new entrants push customers the other way, they teach a generation of borrowers that the gold loan is the loan you pay last. That is a lesson the whole segment will pay for when the bullets come due.
Deferred is not the same as gone
There are a few reasons to take this seriously.
The first is the sheer pace of growth. As per RBI data, NBFC loans against gold jewellery grew 69.1% year on year to about ₹3.65 lakh crore in August 2026. Some of this growth is because gold prices went up and each gram fetches a bigger loan. Some of it is because lenders are opening gold branches at a rapid clip. But some of it, as Piramal's data suggests, is borrowers who are already struggling elsewhere.
The second is that bullet loans make the books look cleaner than they are, for a while. A loan where nothing is due cannot be late. So a bullet gold loan book can show very few overdue loans right up until the point where payments fall due. The risk is not removed. It is bunched up at renewal time.
The third is the price of gold itself. When gold prices were rising, a struggling borrower had an easy way out. The same gold was worth more, so he could renew the loan at a higher amount, pay off the accumulated interest from the extra money, and carry on. Rising gold prices were the safety valve. With gold flat or falling, that valve is closing, and the interest has to come out of the borrower's pocket.
None of this means gold loans are about to go bust. Gold is still among the best collateral a lender can have. It is liquid, easy to value and, unlike a house, can be sold quickly if a borrower defaults. Lenders typically lend up to 75% of the gold's value, which leaves a cushion. Most gold loan borrowers repay.
But collateral protects the lender, not the borrower. If enough households end up unable to pay, the outcome is gold being auctioned, families losing jewellery that often carries deep emotional value, and borrowers losing access to the cheapest form of credit they have. That is a real cost, even if the lender recovers its money.
So what should one watch? Here are a few things:
- Stage 2 in gold loan books across lenders over the next two or three quarters, especially at renewal time.
- Overlap with other loans. Piramal's 35% number, and the share of gold loans going to borrowers who are already 90+ days late elsewhere.
- Repayment structure. Lenders whose customers pay interest every month see stress early, because a missed payment shows up in weeks rather than months. Lenders with mostly bullet books will see it later, and all at once.
- Auctions. A rise in gold auctions is the final, visible end of the stress that started elsewhere.
To conclude
So, are gold loans becoming India's bad bank? Not in the sense that they are about to blow up. But they may well be becoming a bad bank that nobody set up: the place where the economy's stress goes to hide, before anyone has recognised it as bad.
When money is tight, people do not stop paying everything at once. They choose. They protect the loans that hurt the most if missed, and they lean on the loan that asks for nothing until later. In India, that loan is the gold loan. It has always been the household's emergency fund. What has changed is that more people seem to be reaching for it, and some of them are already behind on other loans when they do.
But let me be clear about one thing. None of this makes gold loans a bad asset for lenders. Gold remains the safest asset class in Indian lending. It is the most liquid collateral there is, easy to value and quick to sell, and the loan-to-value cap leaves a sufficient margin of safety even after the interest piles up. If a borrower stops paying, an auction will, in most cases, recover the principal along with the interest. The lender may not take an actual credit hit at all.
And that is precisely why the signal matters. When the lender is protected, the stress does not show up as losses on its books. It shows up at the auction counter. Any increase in gold loan auctions will tell us what the clean numbers elsewhere do not: that households are deferring their stress, shifting it from one loan to another, and finally paying for it with their family's jewellery.
So, gold loans will likely remain the safest loans on any lender's books. The question to ask is not whether lenders will get their money back. It is what the auctions tell us about the borrower on the other side. Watch them closely. That is where the economy's deferred stress will finally show up.
Sources
- Piramal Finance, Q2 FY27 earnings call transcript, October 19, 2026
- Bajaj Finance, Q2 FY27 earnings call transcript, October 19, 2026
- Reserve Bank of India, Sectoral Deployment of Credit by NBFC - August 2026
- SLM, RBI norms on income recognition and asset classification (borrower-wise NPA classification and upgradation)
- La Excellence, RBI NPA norms effective April 1, 2027
- Equentis, Brent crude crosses $100 on September 9, 2026
- Asianet Newsable, India's CAD and inflation as Brent nears $100
- NewKerala / ANI, Crude above $100 could push inflation towards 6%: Kotak Securities
- IANS, Top IT firms' headcount falls over 7,000 in FY26