Home Loan Masterclass · Part 9 · Home loans · 13 min read · August 2026

RBI's new loan interest-rate rules: what could actually change for your home loan

A borrower walks in with a question that sounds completely reasonable. Repo rate has come down. Your website is showing home loans cheaper than mine. I have paid every EMI on time for five years. Why am I still paying more? The uncomfortable answer is that a floating home-loan rate has never been just one number — and RBI now wants that machinery to become considerably harder to hide.

First, understand the number on your sanction letter

Suppose a floating home loan is priced at 8.00%. That figure is not normally pulled from a table as a standalone number — it is two pieces stacked on top of each other. With the policy repo rate at 5.25%, a repo-linked loan at 8.00% is carrying a spread of 2.75%. The benchmark is the moving part that reflects the wider interest-rate environment; the spread is the lender's layer on top, and it is where the borrower becomes specific. It can reflect credit risk, operating cost, the tenor of the loan and the lender's own commercial strategy.

For a commercial-bank floating home loan, the benchmark today is usually an external one — most commonly the policy repo rate, though the eligible list also includes specified Treasury Bill yields, the Secured Overnight Rupee Rate, and other benchmarks published by Financial Benchmarks India. What RBI's draft does is formally break the spread into identifiable components instead of leaving it as one mysterious number.

That distinction matters because borrowers watch the first number and ignore the second. From the desk, that is where most rate disputes begin. A customer reads that RBI changed the repo rate and expects the home loan to move by exactly the same amount on the same day. Sometimes it eventually does. Sometimes the reset date has not arrived. Sometimes the loan belongs to an older benchmark regime altogether. And sometimes the benchmark has moved perfectly well, but the spread makes the final rate look nothing like what a new borrower is being quoted. You cannot understand your rate until you separate the two.

What RBI is actually trying to clean up

Rules already exist for how commercial banks price benchmark-linked floating loans. The problem is that the architecture grew in pieces. Banks got MCLR rules; external-benchmark rules arrived later; NBFCs and housing finance companies operated under a different and much thinner set of requirements; fixed-rate lending had almost no detailed regulatory treatment at all. Even among banks, RBI says it observed divergent practices in areas such as how MCLR and its components are determined.

The draft is an attempt to put those pieces under one principles-based roof, with stated objectives of effective monetary-policy transmission, appropriate pricing of credit risk, and fair and non-discriminatory treatment of borrowers. It is proposed for commercial banks, regional rural banks, urban and rural co-operative banks, all-India financial institutions, NBFCs and housing finance companies — but several requirements differ by lender category, and some are not mandatory for the smallest ones at all.

Harmonised does not mean identical. A commercial bank and a housing finance company will not necessarily have to price your home loan the same way.

The first big change: your reset date stops being a footnote

Imagine RBI cuts an external benchmark today. Does your EMI change tomorrow? Not necessarily — a floating loan does not recalculate itself every morning. It has a reset cycle, and if your benchmark is reviewed every three months with the next reset two months away, the old rate can simply continue until that date arrives. This is why two borrowers linked to the same benchmark can temporarily show different effective rates and both be correct.

Under the draft, the benchmark used, the reset periodicity and the date of reset would have to be explicitly specified in the loan agreement. The periodicity chosen by the lender could not exceed three months, and once fixed for a loan it would remain unchanged for the entire tenor. Where the reset is monthly or longer, the benchmark moves on the first calendar day of the month in which the reset falls due; where it is more frequent than monthly, it moves on the date the agreement specifies.

One honest caveat: the three-month cap is not proposed as mandatory for every lender. Rural co-operative banks with deposits up to ₹1,000 crore, NBFCs in the base layer, and urban co-operative banks in Tiers 1 and 2 are carved out of that requirement in the draft. For the lenders most home-loan borrowers deal with, though, it would apply.

The second big change: the spread gets pulled apart

For an existing home-loan borrower, this is probably the most important part of the whole document. RBI defines the spread as the mark-up over the benchmark covering costs and risk premiums, explicitly excluding charges and fees — so they cannot be quietly folded into it. Under the draft, a lender's board-approved policy would have to set out how the spread is constructed and the range applicable to different loan categories, built from named components: credit risk premium, operating cost, term premium and business strategy premium. Components may be zero, with one exception — the credit risk premium must always be positive.

That framing finally gives a clean answer to the question borrowers ask constantly: why is my home loan 8.40% when the bank is advertising 7.90%? Because the benchmark can be identical while the spread is not.

Nothing in that arithmetic requires the benchmark to differ. The entire gap sits in the spread — which is why comparing only the "repo-linked" label on two loans tells you almost nothing. Ask for the actual benchmark and the actual spread, separately.

Will the new rule force banks to give old customers the new-customer rate?

No — and this is an important place not to overpromise. The draft does not say every borrower at a lender must receive the same spread, and it does not say that when a cheaper offer launches, every old borrower automatically gets it. What it does is put much tighter logic around when the components of an existing borrower's spread can move at all.

The credit risk premium could be revised only when the borrower's credit profile undergoes a change, in accordance with the lender's policy and the loan agreement, and only after a comprehensive review of that credit risk. Every other component of the spread generally could not be revised before three years on a floating-rate loan — with that three-year clock counted from the date of first disbursement or the date of the last revision of the spread, whichever is later.

There is one genuinely borrower-friendly exception written into it. A lender may reduce those other components before the three years are up for customer retention, on justifiable grounds and in a non-discriminatory manner under its policy. Read that again, because it does not create a right to demand the lowest rate you saw in an advertisement. What it does is put inside the framework something borrowers and bankers already know happens across the desk: a good existing borrower says another lender is offering 0.40% less, and the existing lender has a commercial reason to keep them. Under the proposal, that repricing must sit inside a defensible policy rather than being an arbitrary favour.

Can the lender increase your spread because it feels like it?

The proposed answer moves much closer to no. The credit-risk component is meant to track your actual credit risk, so if a lender says your risk premium has gone up, there should be a credit reason behind it and a documented review supporting it — deterioration in the credit profile, or another risk factor captured by the approved methodology. What the draft tries to make harder is an unexplained spread adjustment with no transparent connection to the risk being priced. The other components get their own three-year restriction.

This is one of those rules whose value is invisible on the day a loan is sanctioned. It becomes important four years later, when nobody remembers the sales conversation and the only things left are the contract and the system.

MCLR gets rewired too — which matters if your loan is older

Plenty of home loans are still running on MCLR rather than an external benchmark, and the draft touches those as well. The marginal cost of funds would be computed as a moving average over the trailing three months, with each month's figure being an annualised weighted average interest cost on the volume of fresh deposits and borrowings — system-generated and independently verifiable, rather than assembled by hand. Lenders publishing an internal benchmark would publish it on the first calendar day of each month, and that published figure applies to loans linked to it and sanctioned during that month.

The point of that change is consistency. Two banks facing the same funding environment should not be able to produce very different internal benchmarks because they measured the same thing differently.

Fixed-rate loans come inside a clearer framework too

Most of the discussion around the draft has focused on floating rates, because that is where borrowers feel benchmark movements. But the proposed Directions cover fixed-rate loans explicitly: a fixed-rate loan would also have to be determined with reference to an internal or external benchmark plus a risk-based spread. For hybrid products — fixed for a period, floating afterwards, or the reverse — the relevant fixed or floating provisions apply during the respective periods.

That matters because "fixed" describes what happens to the customer's rate during an agreed period. It should not mean the lender has no transparent basis for how the rate was arrived at in the first place.

Are all home loans going to become repo-linked?

No, and this is the easiest headline to get wrong. For commercial banks, the draft continues the external-benchmark requirement for floating-rate personal loans and floating-rate MSME loans, and an ordinary individual home loan sits inside the personal-loan category used for this regulatory purpose. Commercial banks may also offer external-benchmark loans to other categories of borrowers if they choose.

But RBI is not proposing to force every NBFC, housing finance company, regional rural bank or co-operative bank into the same model — those categories retain discretion over whether to offer external-benchmark-linked floating loans at all. So if somebody tells you that from April 2027 every housing finance company must link every home loan directly to the repo rate, that is not what the draft says. Different lender types can still use different permitted benchmark structures. The improvement is meant to be in how clearly that structure is defined, applied and disclosed.

A small line with a large effect: no loan priced below the benchmark

For both fixed and floating loans, the draft says a lender shall not price a loan below the applicable benchmark for that loan. At first reading that sounds obvious. It is far more consequential for lenders that historically used a high internal reference rate and then advertised large "discounts" beneath it. The proposed structure pushes the industry towards an equation a borrower can actually follow — benchmark plus a positive spread equals the lending rate — rather than a very high reference rate minus a mysterious discount. In the first version you can see what moved because the market changed, and what belongs to your lender or your own risk profile.

How the interest itself would be calculated

The draft also standardises several mechanical items borrowers almost never see discussed. Interest would generally be charged at monthly rests, computed on a daily reducing balance, using an Actual/Actual day-count convention — meaning the system works from the real outstanding balance and the real number of days, rather than an assumption that every month and year contains the same number of them.

For a normal EMI borrower this stays invisible day to day. Where it surfaces is around part-payments, disbursement dates, broken periods and account closure. If you make a substantial part-payment today, daily reducing balance is exactly what you want: the reduced principal starts affecting the interest calculation from the relevant date rather than waiting for artificial monthly arithmetic. Our guide to the 2026 prepayment rules covers the other half of that decision.

What happens to a home loan you already have?

This may be the most important question on the page, because the proposed Directions are not designed only for loans sanctioned after April 2027. The draft provides for all existing loans linked to an internal or external benchmark to be brought into the new framework by 1 April 2029 through a one-time mapping exercise, with three protections written into that transition: your consent is required, the revised rate after transition cannot exceed the rate applicable immediately before it, and the lender cannot levy any charge for the migration.

That does not mean the rate can never rise afterwards. If yours is a floating loan and the benchmark later rises, ordinary floating-rate mechanics still operate. The protection is against the transition itself being used as an excuse to leave you worse off.

So if your existing floating rate immediately before migration is 8.10%, the mapping should not produce a letter saying your new framework rate is 8.35%, and it should not produce a framework conversion charge of ₹4,999. Afterwards, if the applicable benchmark moves, your rate moves with it under the new contract.

A related protection covers a situation many borrowers meet by accident. Where lenders merge or a portfolio is acquired, the transferee has to carry out its own one-time mapping without placing the borrower in a disadvantageous position — the revised rate cannot exceed what applied with the previous lender immediately before the merger. If your loan has ever been moved between institutions and the rate quietly went up, that is the paragraph to know about.

What if your benchmark disappears altogether?

Long home loans outlive products, policies and sometimes entire benchmark systems. India has already moved through BPLR, Base Rate, MCLR and external-benchmark regimes, so a twenty- or thirty-year loan faces a very real chance that the reference rate it started with will one day be replaced. The draft addresses this directly: if the benchmark used for a floating loan is discontinued, the lender must shift to another benchmark without putting the borrower at a disadvantage in terms of the applicable rate, and the loan agreement may carry a fallback mechanism explaining in advance what happens.

That clause deserves more attention than it usually gets. A benchmark transition should be an administrative necessity, not an opportunity to quietly reprice an old customer.

So does a repo cut have to reach me within three months?

If the final Directions retain the proposal substantially as drafted and your loan falls under the reset provision, the benchmark reset periodicity cannot exceed three months. But be precise about what that means: it is not a promise that your EMI falls three months after every policy decision. Your loan's benchmark has to move first, your contractual reset has to occur, and your final rate still contains your spread. Depending on the lender and your repayment arrangement, a rate change may adjust the EMI, the tenor, or a combination of the two under the applicable rules and loan terms.

Which is why the useful question at a branch is not "RBI cut the rate, why didn't you cut mine?" Ask these five instead:

• What benchmark is my loan linked to?
• What was its previous value, and its current value?
• What is my spread?
• What is my reset frequency?
• What is my next reset date?

Five answers. The entire loan rate should become explainable from them.

What the draft does not do

A regulation gets easier to understand once you strip away what it never promised. It does not fix one home-loan rate for India — lenders still price risk and still compete, so a stronger credit profile and a lower-risk property can still earn a different price. It does not abolish the spread; it structures what sits inside it and when the parts can move. It does not guarantee an old customer every new-customer offer, though it does leave room for retention-based reductions. It does not force every NBFC and housing finance company onto the repo rate. And most importantly, it does not take effect today: comments were invited until 11 September 2026, the document is still a draft, and RBI has said final Directions will be issued separately for each category of regulated entity after the feedback is examined.

Treat anyone describing every proposal on this page as an already-enforceable borrower right with some caution.

New borrower in 2027? Four lines to red-pen before signing

If this framework becomes final, four items on a floating home-loan sanction letter deserve a pen. The benchmark — not the word "floating," the actual name: repo, a Treasury Bill yield, SORR, or an internal benchmark. The spread — if the benchmark is 5.25% and your rate is 8.10%, your spread is 2.85%, and that number deserves as much attention as the 8.10%. The reset frequency, because a cheaper benchmark is less useful if your loan takes longer to pick up the change. And the reset date, because a quarter beginning in January and one beginning in March react to the same RBI decision at very different times. Do not leave the branch knowing your loan is "quarterly reset" without knowing when the quarter turns.

The point of transparency is not that the answer exists somewhere in the document. The point is that you can work it out yourself.

Existing borrower? Do this instead

Do not wait until 2029 to understand your own loan. Open your latest statement and your sanction letter and find five numbers now: current outstanding, current interest rate, benchmark, spread, next reset date. If the statement does not make them clear, ask for the break-up in writing. Then look at what the same lender is currently offering a comparable new borrower. A small difference is normal. A large and persistent one deserves a calculation.

The real change is not the EMI. It is the explanation.

It is tempting to read every RBI interest-rate headline as will my EMI come down. That is not the interesting part of this proposal. RBI is not offering anyone a cheaper home loan. It is trying to make a much more basic question harder to evade: why am I being charged this rate?

If your rate is 8.20%, the lender should be able to show you the benchmark and the spread. The agreement should tell you when the benchmark resets. If the credit risk premium changes, there should be a credit reason behind it. And if your old loan is migrated into a new regulatory structure, the transition itself should not leave you worse off or generate another fee.

That is less exciting than an overnight one-percent cut. Across a twenty-year loan, it is probably more useful — because the biggest disadvantage a borrower carries into a lending desk is rarely arithmetic. It is information. The cleaner the rate becomes, the less of that information stays on only one side of the table.

Written at the MoneyClarityTech desk — by a working retail-credit professional in Indian banking who reads loan files, credit reports and bank statements every working day. Patterns from hundreds of real cases; every identifying detail removed. More about MoneyClarityTech →