FOIR: How to Calculate It, and Why One Number Hides More Than It Shows

FOIR — the fixed obligation to income ratio — is the proportion of a borrower's monthly income already committed to servicing debt. It is the most widely used capacity measure in Indian lending and among the least carefully constructed.

The formula is trivial. The inputs are not, and almost every material error in a FOIR calculation happens before the division.

The formula

FOIR = Total monthly fixed obligations ÷ Total monthly income

A borrower earning ₹1,00,000 a month with EMIs of ₹45,000 has a FOIR of 45%. Adding a proposed EMI of ₹15,000 takes them to 60%.

There is no centrally prescribed FOIR ceiling in India. Thresholds are set by each institution's credit policy — commonly somewhere between 50% and 60% for salaried unsecured lending, higher for secured products and for borrowers at higher income levels, where the same ratio leaves far more absolute money to live on. That last point is worth stating explicitly: 60% of ₹40,000 and 60% of ₹4,00,000 are not comparable risks, and a policy that treats them identically is measuring the wrong thing.

Why one FOIR is not enough

A single ratio compresses three different questions into one answer. Calculating it in layers keeps them separate.

FOIR (1) — debt obligations only. EMIs, card repayments, and any other credit servicing. This is the comparable number, the one that lines up across lenders and roughly matches what bureau-based logic sees.

FOIR (2) — plus fixed commitments. Rent, insurance premiums, school fees, committed investments. Not debt, but not discretionary either. A borrower does not stop paying rent to service an EMI.

FOIR (3) — plus essentials. Utilities, groceries, fuel, transport. The floor below which consumption cannot compress.

The headline number is the same in every case. The gap between the layers is where the information is.

Consider two borrowers who both present a FOIR (1) of 45%.

The first runs 45 / 48 / 52. Almost no fixed commitments beyond debt, modest essential spend, and roughly half of income uncommitted. There is real absorption capacity here — an income shock or an unexpected expense can be accommodated.

The second runs 45 / 68 / 85. Identical debt load, but rent and fixed commitments consume another 23 points and essentials another 17. Fifteen percent of income is free. Any disruption at all forces a choice between the EMI and something else.

On the number your credit policy tests, these borrowers are the same. They are not the same.

The denominator is where FOIR is usually wrong

Income sounds like the easy input. It is the harder one.

The test is recurring and repeatable. Not "money that arrived."

For salaried borrowers, the credits that need judgment are these:

• Reimbursements are the most common error in retail underwriting. An expense reimbursement lands in the account as a credit, indistinguishable in size and regularity from income. Counted as income, it improves FOIR while changing the borrower's actual capacity not at all — the money was already spent.
• Bonus and variable pay should be normalised across the period, not counted at face value in the month they arrive. A FOIR calculated on a month containing an annual bonus is not a FOIR, it is a coincidence.
• Arrears and one-time settlements are not income.
• EPFO and gratuity credits are asset movements, not earnings.
• Rental income and interest count if they are demonstrably recurring, at the actual received amount rather than the declared one.

For self-employed and business borrowers there is no salary line to anchor to, and income has to be derived from net inflow after removing everything that is not business receipts: self-transfers between the borrower's own accounts, loan disbursals, related-party inflows, refunds, and circular flows between the promoter and the entity.

This is not a refinement. It is the whole calculation. If a loan disbursal is counted as business income, FOIR improves at precisely the moment leverage rose — the ratio moves in the opposite direction to the risk. We measured how often exactly this misclassification occurs across 200,001 transactions, and disbursals read as ordinary counterparty credits were among the most frequent errors we found.

The numerator misses what the bureau cannot see

Obligations are usually sourced from the bureau. The bureau is incomplete in four specific ways, all of them visible in banking.

Bureau reporting lag. A loan disbursed five weeks ago may not yet appear, while its EMI is already debiting the account. Banking sees the debit immediately.

Obligations that never report. BNPL repayments, informal and small-ticket lenders, chit fund contributions, and borrowings from family all service through the account without a bureau trail.

Card treatment. A borrower who clears the full statement each month carries no obligation; one who revolves carries a minimum-payment obligation and a growing balance. These produce identical bureau entries and completely different debits. Only the payment pattern in the account distinguishes them.

Guarantor and co-borrower exposure. Reported inconsistently, and frequently serviced from an account that is not the primary borrower's.

The practical rule: bureau establishes what is owed, banking establishes what is actually being paid, and where the two disagree the account is the more reliable witness.

When FOIR exceeds 100% and the borrower is still paying

The most instructive case in FOIR analysis is the one that appears impossible.

Take a borrower whose analysis produces this:

Read literally, this borrower spends a third more than they earn and should have defaulted months ago. Instead the account is closing each month with savings of ₹81,230 — a positive 23% savings ratio alongside a negative disposable income.

Both figures cannot be right. Something is entering the account that the income calculation is not capturing, and there are only three candidates:

1. Undeclared or unclassified income. Business receipts, rental, professional fees or family support arriving in a form the categorisation logic did not recognise as income. If so, the borrower's true capacity is materially better than the ratio suggests, and the file has been misjudged downward.
2. The obligations are being serviced from borrowings. New credit funding old credit, which is the most dangerous pattern in retail lending and the one FOIR was designed to catch.
3. Money is being routed through the account on someone else's behalf. Neither income nor obligation, and the account is not describing this borrower's finances at all.

Each demands a different decision. What matters is the general principle underneath, which applies well below 100%:

When FOIR says a borrower cannot service their obligations and the account shows them servicing their obligations, the inputs are wrong, not the borrower. FOIR is at its most useful not as a threshold test but as a consistency check between what you calculated and what the account actually did. A large contradiction is a signal to re-examine the classification of every significant credit before making a decision — not a signal to decline.

FOIR has to be a series, not a snapshot

A FOIR computed on one month is a reading taken at whatever point in the bonus, variable pay and seasonal cycle you happened to sample.

Build it monthly across the full statement period and read it month on month. Annual averages are worse than useless here: a year in which three new obligations appeared in the final quarter can average to a perfectly comfortable figure, and year-on-year comparison puts twelve months between observations, hiding the entire trajectory that matters.

Three patterns in the monthly series:

• Rising FOIR on flat income. New obligations appearing. Check whether they are on the bureau — if not, the borrower has been borrowing in places you cannot see.
• Falling FOIR. Distinguish the cause. Obligations closing out is genuine improvement. Income rising on a single large credit may be a one-off dressed as a raise.
• FOIR volatile on stable obligations. The income side is lumpy, which means the average understates the strain in weak months. Test against the worst month, not the mean.

What FOIR does not tell you

Capacity is not the same as conduct, and FOIR measures only the first.

A borrower at 40% FOIR who reaches zero balance before every EMI date, and clears it with a same-day transfer from elsewhere, is running tighter than a borrower at 55% with a stable end-of-day buffer. The ratio ranks them the wrong way round.

FOIR should be read alongside end-of-day balance behaviour on obligation dates, bounce history separated into technical and non-technical causes, the savings ratio, and the gap between FOIR (1) and FOIR (3). Together those describe whether a borrower can pay and whether they reliably do. FOIR alone answers only the first, and only if the inputs were built correctly.

Frequently asked questions

What is the full form of FOIR? Fixed Obligation to Income Ratio — the share of a borrower's monthly income already committed to fixed obligations, principally debt servicing.

How is FOIR calculated? Divide total monthly fixed obligations by total monthly income. The difficulty lies in the inputs: income should include only recurring, repeatable credits with reimbursements and one-time receipts excluded and bonuses normalised, while obligations should include debits the bureau does not capture, such as BNPL, informal lenders and recently disbursed loans.

What is a good FOIR for a loan? There is no centrally prescribed limit in India; each lender sets its own. Between 50% and 60% is common for salaried unsecured lending, with higher tolerance for secured products and for higher income levels, where the same percentage leaves more absolute income available.

What is the difference between FOIR and DTI? They measure the same thing — obligations against income — with FOIR the term used in Indian lending and debt-to-income ratio the international equivalent. Institutions differ on whether the numerator is restricted to debt or extended to all fixed commitments, which is why calculating the layers separately is worth doing.

How is FOIR calculated for self-employed borrowers? There is no salary credit to anchor to, so income is derived from net business inflow after removing self-transfers, loan disbursals, related-party credits and circular flows. Accuracy depends almost entirely on whether those non-revenue credits were correctly identified — misclassifying a disbursal as income improves FOIR at the exact moment leverage increased.

---

Fiscus computes FOIR in three layers across salaried and self-employed profiles, with income and obligations derived from categorised banking data rather than declared figures. Book a parallel evaluation on cases your team has already assessed.