India bank credit, May 2026: up 18%, and the mix is where the story is
Published: July 10, 2026
The RBI sectoral credit numbers are out for May 2026. I have loaded them into the interactive tool on tigzig, so you can sort them and download the full series. Here is what I make of them.
What this data is
Every month RBI publishes the Sectoral Deployment of Bank Credit. It takes the total credit given out by banks and splits it across about 20 sectors and sub sectors, from agriculture and industry to housing, vehicle loans, credit cards, gold loans and lending to NBFCs. It is the closest thing we have to a monthly view of where bank credit is actually going. On the tool it is updated monthly, from December 2022 onwards, which is 42 months, roughly three and a half years.
The latest number, at the end of May 2026, puts total bank credit at about 215 lakh crore rupees, which is the entire outstanding credit of Indian banks. What I am doing below is picking out a few of the pieces that make up that number and sharing what stands out.
The headline
Over the year to May 2026, total credit grew about 18%, and in rupee terms banks added roughly 32 lakh crore over the twelve months. That is a healthy number on its own, but the average hides a fair amount, because some segments grew much faster than that and some grew much slower, and the gap between them is the interesting part.
The fast lane
Loans against gold jewellery grew about 105% over the year, so the segment has more than doubled in size. It is still a small segment at about 5.1 lakh crore, and its share of total credit went up from 1.4% to 2.4% over the twelve months, which is a large increase for something this size.
Lending by banks to NBFCs is the other fast growing segment, up about 34%, with its share going up from 8.5% to 9.7%. A good part of the gold story is also inside this number, because the gold loan companies are counted within the NBFC group. So both banks and NBFCs are now lending more against gold at the same time, one directly through the gold loan line and the other by funding the NBFCs that do the gold lending. Independent data points the same way, with Equifax reporting that gold loan originations were up about 103% year on year in the quarter to March 2026, and gold loans leading retail credit growth.
The slow lane
Housing is the largest segment on the consumer side, at about 33.7 lakh crore, which is roughly 16% of all bank credit. It grew about 11% over the year, well under the 18% average, and it lost more share than any other segment. In the latest month, housing added 0.5% while total credit added 1.4%, so the largest single segment is growing at under a third of the pace of everything else.
Credit card outstanding is the other slow one, up about 1% over the year, so it is more or less flat.
This is really about capital rules
Gold growing fast while credit cards stay flat looks like a story about how households are borrowing. It is really a story about how much capital banks have to set aside.
Go back to November 2023. RBI raised the risk weights on unsecured lending, on personal loans and credit card dues for both banks and NBFCs, and also raised the weight on bank lending to NBFCs. A higher risk weight means the lender has to hold more capital against that loan, which makes it more expensive to grow. When RBI did this, it left out loans secured by gold, along with housing, education and vehicle loans. So one type of lending became costlier to grow on capital while gold lending stayed cheap on capital, and lenders put more of their money into the secured lending that had been left alone, with gold being the clearest example.
In February 2025, RBI partly reversed this, easing the weights on bank lending to NBFCs and on microfinance loans. That made it easier for banks to lend to NBFCs again, which is part of why the NBFC number is as high as it is.
The gold price angle
There is a second thing going on here, which is the gold price itself. For most of this period the gold price was going up sharply, and a higher gold price increases the value of the jewellery being pledged, which lets a lender give a larger loan against the same ornaments. That helped the whole gold loan segment right through the period covered by this data. Gold, in dollar terms, was up around 39% year on year by the end of May 2026, so the collateral was worth a lot more than it was a year earlier.
Since then the gold price has come off its highs. It peaked in early March 2026 near 5,370 dollars and has since fallen to around 4,130 dollars by early July, which is a drop of roughly 23% from the top. This is worth keeping an eye on, because gold loans work on a loan to value basis, and if the collateral keeps falling then a loan given near the peak can end up close to the value of the gold that backs it, and the lender can ask the borrower to put in more. It is not a problem for the whole segment yet, because gold is still higher than it was a year ago, so most of the loans are sitting on collateral that has gone up in value since it was pledged. It is really the loans given near the early 2026 top that are the tight ones, and if the gold price keeps falling, that group gets larger.
One correction
It is tempting to say that NBFC and microfinance both boomed, but only the NBFC part is true. Microfinance actually shrank, with the portfolio falling about 9% over the year to roughly 3.33 lakh crore by May 2026 after a stressed stretch, and it is only now starting to recover. So the risk weight cut on microfinance in February 2025 was there to help a struggling segment, not to reward a growing one, and the growth inside the NBFC group is coming from gold and other secured lending rather than from microfinance.
What I am not claiming
I am not saying that people are consciously giving up their credit card and taking a gold loan instead. I cannot show that from this data, and I would not try to. What I can say is that the flat credit card number and the fast growing gold number are two separate things that share a common cause, which is the capital rules and the gold price, and that part is on the record. The idea that households are switching from one to the other is a reasonable guess, and I would leave it at that.
If you want to look at it yourself
You can sort the data by growth, by share of the total, and by share of the new credit added, change the time window, and download the full series as a CSV to work through it yourself.
tigzig.com → Tremor → India Credit
A note on how the data is put together
One nuance worth knowing. RBI only publishes two pieces of its NBFC lending, housing finance companies and public financial institutions, and that is only about a third of the total. The rest, which is where the gold loan companies and the microfinance lenders sit, RBI does not break out. I compute it as the residual so the numbers reconcile, and that is how you can even see the gold through NBFC story.
There is a similar 5% residual in non-food credit that RBI never breaks out either, and the validation section shows it openly rather than hiding it.
The data itself is small, about 1,400 rows. The real work was pulling every number across time, reconciling them so the parts add up to the whole, and building the interface so you can cut it without setting up a pivot. You can download the whole reconciled set as CSV and check it yourself.
The tool lets you cut it in many ways ... I designed it as I cut into the data to pull out the insights, smoothening the interface as I went along. There is also a composition view, a trial balance kind of layout, if you want to see the whole hierarchy add up.
Sources
RBI, Sectoral Deployment of Bank Credit (monthly), read through the tigzig Tremor tool.
The risk weights described here are in force now, and they sit in the Reserve Bank of India (Commercial Banks, Prudential Norms on Capital Adequacy) Directions, 2025, reference RBI/DOR/2025-26/151, updated as on 1 July 2026. Consumer credit and personal loans carry a 125 per cent risk weight, excluding housing, education, vehicle and gold-secured loans; credit card receivables carry 150 per cent for banks; microfinance loans in the nature of consumer credit carry 100 per cent; and bank exposure to NBFCs is rating based, with no add-on. The NBFC side sits in the matching capital adequacy Directions for NBFCs, reference RBI/DOR/2025-26/345, where credit card receivables carry 125 per cent.
The changes themselves were made by three circulars, now superseded: the hike of 16 November 2023, RBI/2023-24/85, Regulatory measures towards consumer credit and bank credit to NBFCs, and the two rollbacks of 25 February 2025, RBI/2024-25/119 on microfinance and RBI/2024-25/120 on bank exposure to NBFCs. On 4 December 2025 RBI folded roughly 9,000 circulars into 238 consolidated Master Directions, so these three pages, and the old Basel III Master Circular, now carry a "Withdrawn" mark dated 4 December 2025. This is a consolidation, not a reversal, and every value moved into the new Directions unchanged. The reason is on the record in RBI's consolidation press release of 10 October 2025 and in its Financial Stability Report of June 2026.
One forward point, for accuracy: RBI has issued a revised capital framework, the Standardised Approach Directions, 2026, which changes these risk weights from 1 April 2027. That is after the period this data covers, so it does not affect anything here, but the current weights should not be read as permanent.
Gold loans: Equifax, India's Retail Credit Market AUM Surges 16% to Rs 162 Lakh Crore; Gold Loans Lead Lending Growth, 5 June 2026. Gold loan originations up 103% year on year for the quarter ending March 2026.
Microfinance: Equifax, India's microfinance sector stabilises further, delinquencies fall to 2.1%, portfolio contracts 9%, 30 June 2026. Portfolio outstanding 3,33,110 crore, down 9% year on year as of May 2026.
Gold price: COMEX gold futures, read through the tigzig Tremor tool. Peak near 5,370 dollars in early March 2026, around 4,130 dollars by early July 2026.
Working on something similar? How I work covers the rates, the availability and what I take on.