DSCR from Banking Data: Why the Annual Number Misses What Matters

The debt service coverage ratio asks a simple question: does the business generate enough cash to meet its debt obligations, and with how much room to spare.

DSCR = Cash available for debt service ÷ Total debt service obligations

A ratio of 1.0 means the business covers its obligations exactly, with nothing for a bad month.

Indian bank credit policy usually sets two numbers rather than one: an average across the loan tenor and a floor for any single year. Indian Bank's loan policy sets the MSME term loan benchmark at an average DSCR of 1.50 with a minimum of 1.25, then delegates relaxation down a ladder of sanctioning authorities: 1.25 average and 1.00 minimum at the field general manager committee, and 1.00 on both measures at the corporate office committee. These are credit policy, not regulation. There is no RBI prescribed DSCR for MSME lending, and the project finance regime that took effect on 1 October 2025, now carried in the credit facilities directions RBI consolidated in November 2025, treats DSCR as a field to be captured in the project finance database rather than a threshold to be met.

DSCR calculation is not difficult as arithmetic. The difficulty is that the standard way of computing it, from annual financial statements, answers a question adjacent to the one you need answered.

Why the financials-derived DSCR fails for MSME lending

The conventional construction takes profit after tax, adds back depreciation and interest, and divides by interest plus principal repayments for the year. It is the right method for a rated corporate with audited quarterlies. For a ₹40 lakh MSME facility it has five problems.

It is one observation per year. By the time audited financials reach a lender they routinely describe a year that closed well over twelve months earlier, once the filing calendar and the appraisal cycle are both accounted for. You are assessing a business's current capacity using a photograph of a previous condition.

It is accrual, not cash. Profit recognises revenue when invoiced. A business with excellent profitability and uncollected receivables has strong accrual DSCR and no money. Debt service is paid in cash.

It is prepared by the borrower. Financial statements involve judgement at every line, and the variance in statement quality across the MSME segment is wide enough that two businesses with identical underlying economics can present very differently.

It averages away the working capital cycle. An annual figure cannot see that the business runs a ₹30 lakh deficit for four months and recovers it in two.

The depreciation add-back assumes no reinvestment. For an asset-dependent business, adding back depreciation treats maintenance capex as available for debt service when it is not.

This is not a fringe view. The Reserve Bank of India's expert committee on micro, small and medium enterprises, reporting in June 2019, recommended that banks move towards cash flow based lending on precisely this reasoning, observing that the traditional system rests on the financial statements and the collateral of the borrower.

Banking data has the opposite properties. It is monthly or better, it is cash rather than accrual, it is produced by the bank rather than the borrower, and it shows the working capital cycle directly rather than in summary.

Building the numerator

Cash available for debt service, derived from the account:

Start with gross credits, then subtract everything that is not a business receipt. Self-transfers between the borrower's own accounts. Loan disbursals. Related-party and promoter inflows. Refunds and reversals. Investment redemptions.

This step is not a refinement — it determines whether the ratio means anything. Computing DSCR on gross credits inflates the numerator by every rupee the borrower circulated, and a business that recycles money aggressively can manufacture a comfortable ratio out of nothing. The netting method, and why gross related-party movement is meaningless while net flow is informative, is set out in our post on circular and related-party flows.

Then subtract operating outflows. Supplier payments, wages, rent, utilities, statutory dues, and the running costs of the business — excluding anything that is itself debt service, which belongs in the denominator.

What remains is operating cash generation: the money the business actually produced and had available.

Building the denominator

Total debt service, and this is where the most common omission lives.

Term loan EMIs, principal and interest. Straightforward and visible in the account.

Facility interest. Overdraft and cash credit interest is debited monthly, because the Reserve Bank of India requires interest on all advances to be charged at monthly rests. It is unambiguously debt service. It appears in no bureau EMI list, it is not an instalment, and it is routinely left out of obligation calculations altogether. A borrower carrying an average drawn balance of ₹80 lakh on a cash credit facility is servicing several lakh rupees of interest every year that never once appears as an EMI.

The Ministry of MSME's own project profile template already treats it this way, placing interest on working capital in the debt service denominator alongside term loan interest and instalments. Bank covenant tables often split the two instead, running a term loan DSCR next to a separate interest coverage ratio for working capital. Either construction is defensible. What is not defensible is an obligation list assembled from the bureau, which captures neither.

Other credit servicing. Bill discounting charges, processing and renewal fees on facilities, card repayments where the borrower revolves rather than clears.

Obligations absent from the bureau. Recently disbursed loans that have not yet reported, small-ticket and informal lenders that report inconsistently, and BNPL. The bureau establishes what is owed on record; the account establishes what is actually being debited.

The drawings question

Should promoter drawings be deducted before computing DSCR? It matters more than it sounds, and the honest answer is that it depends on the structure.

For a proprietorship or owner-managed firm, drawings are the promoter's household income. They are not discretionary — the family eats from them. Treating the full pre-drawings surplus as available for debt service overstates capacity, sometimes dramatically.

For a company with a salaried promoter, remuneration sits in operating costs already, and additional extractions may genuinely be discretionary.

The workable approach is to establish the sustainable drawing level from the account's own history — what the promoter has consistently taken over twelve months — deduct that as a fixed commitment, and treat extraction above that level as discretionary while noting it. A promoter extracting ₹7 lakh from a business generating ₹8 lakh of surplus has a DSCR problem regardless of what the pre-drawings ratio says.

Monthly, and why the minimum month beats the average

This is the argument that matters most, and it is where an annual DSCR is not merely stale but structurally wrong.

Debt service is a monthly obligation. The borrower must clear it every month, not on average across twelve. A business with an annual DSCR of 1.4 may be running at 1.75 for eight months and 0.7 for four — and in those four months it will bounce, or it will fund the gap with fresh borrowing, or the promoter will cover it. All three outcomes are credit events, and none of them is visible in the annual figure.

Build DSCR as a monthly series and read three things:

• Minimum-month DSCR. The worst month in the period. This is the number that predicts bounces.
• Count of months below 1.0. How often the business could not cover itself from operations.
• How those months were funded. Fresh borrowing, related-party inflows, or drawdown of balances. The funding source is more informative than the shortfall.

There is a regulatory version of this argument, and it is stricter than most credit policies. Under the asset classification directions RBI consolidated in November 2025, a cash credit or overdraft account is treated as out of order if the outstanding continuously exceeds the sanctioned limit or drawing power, whichever is lower, for 90 days, or if there are no credits for 90 continuous days, or if the credits over the previous 90 days are not enough to cover the interest debited in that period. That last limb is a monthly coverage test written directly into the classification rules. An account whose own inflows do not cover its own interest is not a borrower to keep an eye on. It is on the path to being a non-performing asset, and an annual DSCR will not see it coming.

Revolving facilities also skip the gentlest warning stage. Under the stressed assets directions there is no SMA-0 bucket for cash credit and overdraft accounts; they move straight into SMA-1 after 30 days of continuous excess over the limit or drawing power, whichever is lower. Worth noting that this vocabulary has a shelf life: the asset classification directions RBI issued in April 2026 replace the SMA buckets with expected credit loss staging from 1 April 2027, which makes a monthly view of coverage more central rather than less.

A borrower with an average DSCR of 1.15 and no month below 1.0 is a better risk than one averaging 1.45 with three months at 0.6. Conventional analysis ranks them the other way round.

What this looks like on a real series

A business on a ₹5,20,000 monthly debt service, read across one year:

The annual figure is 1.40, which clears most policy floors. The minimum month is 0.60 and four consecutive months sit below 1.0. Those four months were funded by something, and the analysis is not finished until you know what: fresh borrowing, promoter money, or a drawdown of the facility. Each answer changes the decision, and none of them is visible in the 1.40.

Seasonality is structure, not noise

A business with peak-month DSCR of 1.6 and trough-month DSCR of 0.7 is not a 1.15 DSCR borrower who needs watching. It is a seasonal business whose repayment structure is misaligned with its cash cycle.

Twelve months of banking establishes the actual trading calendar — which months generate, which consume, and how deep the trough runs. That is an input to structuring, not just to approval: repayment weighted to the strong months, or a facility sized to bridge the trough, converts a borrower who will bounce into one who will not.

It also means a shorter statement period is genuinely inadequate for business lending. Six months that happen to sit inside the peak produce a DSCR that describes a season rather than a business.

When DSCR and the account disagree

The same consistency logic that applies to FOIR applies here, and it is worth stating separately because the failure mode is different.

If your computed DSCR is below 1.0 and the account has cleared every obligation for twelve months, one of three things is true. Revenue is entering the account in a form your categorisation did not recognise as business receipts. Obligations are being serviced from borrowings or promoter funds rather than operations. Or money is transiting the account on someone else's behalf.

Each has a different implication and the first is the most common — which means a low DSCR on a clean account is a prompt to re-examine the classification of significant credits before it is a prompt to decline.

The reverse case deserves more attention than it gets. A comfortable DSCR on an account showing rising utilisation, lengthening EMI delay days and increasing related-party inflows is a ratio computed on inputs that have not caught up with the business. Those signals move first; ratios move later. Our post on early warning indicators sets out the sequence.

What DSCR does not tell you

It is a flow measure and it has the limits of one.

It says nothing about the security backing the exposure or what that security is worth, which matters more as the collateral-free perimeter widens: RBI raised the collateral-free lending ceiling for micro and small enterprises to ₹20 lakh for loans sanctioned or renewed from 1 April 2026, with banks able to take that to ₹25 lakh where the unit has a good track record. It carries no contingent liabilities — guarantees given to group entities, disputed statutory dues, pending litigation. It says nothing about the quality of the receivables behind the credits, including whether a dominant customer is itself in difficulty. And it cannot distinguish a business generating ₹50 lakh from four customers from one generating the same from forty, which is a materially different risk at the identical ratio.

DSCR answers whether the cash covers the obligations. Concentration, security, contingent exposure and receivable quality are separate questions, and a strong DSCR on a single-customer business is not the comfort it appears to be.

Frequently asked questions

What is a good DSCR for a business loan? Indian bank policy typically specifies two numbers: an average across the loan tenor and a floor for any single year. Indian Bank's loan policy sets the benchmark for MSME term loans at an average of 1.50 with a minimum of 1.25. Successively higher sanctioning authorities can accept 1.25 average and 1.00 minimum, and at the highest level 1.00 on both. These are credit policy rather than regulatory requirement. The distribution across months matters as much as the level: a 1.50 average with three months below 1.0 is weaker than a 1.25 with none.

How do you calculate DSCR from bank statements? Take gross credits, remove self-transfers, loan disbursals, related-party inflows and refunds to arrive at genuine business receipts, subtract operating outflows to get cash generation, and divide by total debt service including term loan EMIs, facility interest and obligations absent from the bureau.

Should working capital interest be included in DSCR? Yes. Overdraft and cash credit interest is debt service, is debited monthly because RBI requires interest on advances to be charged at monthly rests, and appears in no EMI schedule or bureau obligation list. The Ministry of MSME's own project profile template places interest on working capital in the debt service denominator. Bank covenant tables sometimes split it out into a separate interest coverage ratio instead. Omitting it from both is what overstates the ratio.

Is there an RBI prescribed DSCR for MSME loans? No. The RBI directions on lending to the MSME sector prescribe no DSCR, and the project finance regime effective 1 October 2025 treats DSCR as a reporting field in the project finance database rather than a threshold. Every number a borrower is tested against comes from the lender's own credit policy.

What is the difference between DSCR and interest coverage ratio? DSCR measures cash against the full debt service obligation, interest plus principal. Interest coverage measures it against interest alone. Indian bank covenant tables frequently run both, applying DSCR to the term loan and interest coverage to the working capital facility, which is one reason facility interest falls out of obligation calculations that look only at the DSCR line.

Why is monthly DSCR better than annual? Debt service is a monthly obligation, so an annual figure averages across months the borrower could not cover. A business meeting its obligations comfortably for eight months and failing for four shows an acceptable annual ratio and will bounce four times a year.

Should promoter drawings be deducted when calculating DSCR? For proprietorships and owner-managed firms, drawings are household income rather than discretionary extraction and should be deducted at the level the account history shows to be sustainable. Extraction above that level can be treated as discretionary but should be noted.

---

Fiscus derives business cash generation and debt service directly from categorised banking data across all accounts in a case, reported month on month rather than as a period total. Book a parallel evaluation on business cases your team has already assessed.