CMA Data Rejections: 10 Mistakes Indian Banks Flag

A CMA file is usually not declined in a meeting. It dies quietly on a credit officer's desk, circled in red, returned with a one-line note. The reasons are almost always the same ten. This article lists them the way the officer reads them, so you can self-audit before you submit.

Why do banks return CMA data when the figures do not match ITR or GST?

Because a mismatch between your CMA's net sales and your filed returns is the single fastest credibility destroyer in the file. The officer starts here for a reason: RBI's Master Directions on Fraud Risk Management (effective 15 July 2024) require banks to run an early warning signal framework whose triggers include “material discrepancies in the annual report” and “significant inconsistencies within the annual report (between various sections)”. A CMA whose Form II sales do not foot to the ITR and GST filings lands on the same kind of list the bank's fraud-monitoring team watches.

CA firms report this as one of the more serious rejection categories, because it raises a question about the entire document, not one line in it. The fix is mechanical. Reconcile net sales across the CMA, ITR, GST returns, and MCA filings before submission, and attach a short reconciliation note for any genuine timing or classification difference. A disclosed difference is a footnote. An undisclosed one is a red flag.

What happens when DSCR falls below 1.25 in a repayment year?

The term-loan portion of your proposal gets restructured, reduced, or rejected. Most banks' credit policies require DSCR of at least 1.25 in every individual repayment year, not just on average, and officers judge the weakest year, not the best one. CA firms report this as the most common reason term-loan projections bounce back.

First, say the regulatory status plainly: no RBI circular fixes DSCR at 1.25. It is a bank-policy convention, full stop. Then fix the math the way preparers actually do: Net Cash Accruals = PAT + depreciation (adding back the non-cash charge most often brings a borderline year above the line), and if the number is genuinely low, the structural answer is a longer tenure or a moratorium, a point our DSCR guide works through in detail. Never fix a weak DSCR by nudging projected sales upward; that hands the officer rejection reason 5 for free.

Why is a balance sheet that does not balance a first-look rejection?

Because it tells the officer, within seconds, that the file was never self-checked. Total sources must equal total applications, and the two classic linkage breaks are operating-statement profit not carried into retained earnings, and depreciation in Form III not matching Form II. Practitioners describe this as the mechanical integrity test the officer runs first: if the arithmetic fails, nothing else in the file gets the benefit of the doubt.

The fix is a ten-minute reconciliation loop: profit from Form II into Form III reserves, depreciation across Form II and Form III, net worth carried consistently into Form VI. CA firms list this linkage check among the first things officers verify. Treat it as a gate: the file does not leave your desk until every statement foots and cross-foots.

Which MPBF method should a borrower use: Nayak or Tandon?

Whichever one your bank's credit policy says. Since April 1997, when RBI withdrew the MPBF prescription and gave banks operational freedom, the choice of method is a per-bank policy question, not a statute. Our MPBF method guide traces the full history.

The common convention, repeated by practitioners, is the Nayak Committee's simplified method (working capital assessed at 20% of estimated annual turnover) for limits up to ₹5 crore, and Tandon Method II (the borrower funds 25% of total current assets; the bank finances the balance of the gap) for larger limits. Use the wrong one and the MPBF computation in Form V collapses at review. Note the anchor: Nayak's lineage is real (the RBI implementation circular RPCD.PLNFS/BC.No.61/06.0262/2000-01, March 2001), but applying it to your file today is the bank's call. Ask the relationship manager which method their bank uses before you prepare, and recompute Form V afresh for each projected year instead of copying last year's figure.

Why do banks distrust projections built backwards from the limit?

Because the officer can see the construction. When projected sales jump to exactly the number that makes the desired limit work, with no capacity, order book, or market evidence behind it, the file reads as wishful arithmetic. A banker's checklist notes that unrealistic sales projections not supported by historical performance are among the first appraisal red flags, and a practitioner example widely quoted in CA circles makes the point concrete: projected debtors of ₹50 crore set against audited debtors of ₹1.5 crore is a red flag no cover letter can fix.

Build projections forward, not backward. Start from installed capacity, past growth, and confirmed orders; show the assumptions note (reason 10) that connects each figure to its source. Growth a little above trend with a stated reason is credible. A round-number leap to the eligibility figure is not.

How do officers spot padded inventory and receivables?

By holding-days arithmetic against your own past trend. If sales grew 10% but inventory doubled and receivables aged sharply, the numbers are telling a story the narrative is not. This check has regulatory weight: RBI's July 2024 fraud-direction EWS list names “significant movements in inventory, disproportionately differing vis-a-vis change in turnover” and “significant movements in receivables, disproportionately differing vis-a-vis change in turnover and/or increase in ageing of the receivables”. The officer's Form IV ageing review at appraisal is the same check, applied before sanction instead of after fraud.

A banker's checklist adds the practitioner texture: excessive inventory beyond normal industry requirements and slow recovery of receivables are standard flags. Keep holding days and collection periods in line with your audited past, and explain any genuine deviation (a seasonal build-up, a large confirmed order) in the assumptions note. Silence reads as padding.

What is the difference between provisional, projected, and audited years in CMA?

The confusion between these three columns causes more return letters than any single ratio. Audited years are final, signed financials. Provisional (sometimes called estimated) is the current year's figures as they stand mid-year: part actual, part estimate. Projected years are forward-looking and wholly assumed. The common layout practitioners describe is two audited years, one provisional/estimate, and two projected years, but the exact pattern is bank-specific.

The classic error, confessed by a practitioner on a CAclubindia forum thread, is deducting term-loan principal from the wrong base date: the mid-year outstanding balance instead of the opening balance of 1 April. When audited accounts later diverge from the estimate you submitted, the answer is to proactively send a reconciled revision, not to wait for the officer's review meeting to surface it. An officer who finds the error first treats it as a credibility issue. An error you flag yourself is a correction.

Why does Form I need fund-based and non-fund-based limits?

Because the bank's exposure to you is not just the cash credit. Form I exists to state existing and proposed fund-based limits (cash credit, overdraft) alongside non-fund-based limits (bank guarantees, letters of credit), and understating the non-fund side hides real risk. A banker's checklist treats unsatisfactory conduct of existing accounts as an appraisal input, and practitioners note that non-fund limits quietly inflated over the years surface at renewal and stall the file.

There is a regulatory adjacent point worth knowing: the same July 2024 RBI fraud directions list “frequent invocation of BGs and devolvement of LCs” as an early warning trigger. Declare every BG and LC limit honestly in Form I, and if utilisation has been heavy, explain why in the cover note. The bank will find the exposure anyway; finding it in your own disclosure costs you nothing.

How do banks catch ratio window dressing in CMA data?

By cross-checking the ratios against the statements they are computed from. The industry rules of thumb are familiar: current ratio around 1.33, total outside liabilities to tangible net worth in the comfort zone, sales at four or five times the loan amount. Practitioners repeat these Tandon-era yardsticks in forum after forum. The problem starts when the ratios are beautified while the underlying figures disagree: a polished current ratio with creditors ageing badly, or a healthy TOL/TNW with promoter funds parked in non-business assets.

Say the status honestly, as this blog always does: 1.33 and its cousins are industry conventions rooted in the 1970s-80s committee era, not live RBI requirements. But the bank's credit policy may still encode them, so ignoring them is not an option either. The self-audit is to compute every ratio from the filed statements and check that the story the ratios tell matches the story the statements tell. When they diverge, the officer believes the statements.

What should a CMA assumptions note actually contain?

The reasoning behind every projected number. Most banks expect assumptions and estimates to be stated separately, CA firms report, and the absence of the note is itself a return trigger: projections without stated assumptions are, to the officer, numbers without a source. The note should cover, in plain lines, the basis for sales growth, margin movement, holding periods for inventory and receivables, the capital expenditure plan, and the repayment schedule assumptions.

A second, duller failure belongs here too: submitting an outdated format not aligned with the bank's current requirement. Form layouts drift between banks and over time, and the Roman-numeral mapping itself varies (one bank's Form VI-A is another bank's Form VI-B). Confirm the current format with the branch before preparing, and anchor your review on the field-level meaning of each statement, which the mapping below lays out, not on the Roman numeral.

Which CMA “rules” are RBI mandates and which are bank conventions?

This map is the article's backbone. Read it before arguing with your officer about any number below: almost everything people call a “CMA rule” is the bank's policy, which means the bank can flex it, and RBI cannot be cited to force it.

What does the officer circle in each CMA form, field by field?

Forget the Roman numerals for a moment. Banks number their forms differently; the fields are what the officer reads. Here is the field-level map behind the ten rejection reasons.

The scale of what this file unlocks is worth remembering while you audit it. Non-food bank credit stood at ₹194.5 lakh crore as of 28 November 2025, with ₹42.2 lakh crore to industry, per the RBI Monthly Bulletin of January 2026. Bank credit to MSMEs alone was about ₹28 lakh crore in February 2025, roughly 15% of non-food credit. A clean CMA is the key to a share of that pool; a flagged one is the lock on it. For the full field-by-field preparation method, see our CMA data guide, and for how the officer's review feeds the bank's internal memo, our CAM guide.

How do the rejection reasons map to RBI's fraud-direction warning signals?

This is the connection no competitor has drawn. RBI's Master Directions on Fraud Risk Management (effective 15 July 2024) require every bank to run a board-approved early warning signal framework, and the trigger list reads like an officer's CMA review notes. The appraisal checks that return your file are the pre-sanction version of signals the bank must monitor after sanction. A clean CMA therefore does double duty: it gets the limit sanctioned, and it keeps the account off the bank's EWS watchlist later. The ELP-hosted copy of the RBI circular carries the full trigger list; the appraisal-stage parallels are below. For the EWS framework itself, see our early warning systems guide.

Did CMA die once, or twice?

Twice, in a sense no competitor mentions. The first date is firm: April 1997, when RBI withdrew the MPBF prescription and handed banks operational freedom over working-capital assessment. The second date is convention, not statute: bank materials commonly place the end of the CMA reporting return itself around December 1997, and the date should be read as practitioner-reported rather than RBI-confirmed. But the appraisal formats never died inside the banks: RBI's own working-group material shows a CMA-style Form A return for limits of ₹10 crore and above still being received by RBI's industrial credit department around 2003. So the reporting to the regulator ended while the appraisal format lived on, bank by bank, as policy convention. That is why “required by RBI” is stale and “irrelevant” is wrong: the CMA format is neither mandated nor dead. It is the bank's own house format, and the bank decides its terms.

The context matters because scrutiny is rising, not falling. The MSME gross NPA ratio fell from 4.5% in March 2024 to 3.6% in March 2025, and the SMA-2 bucket sits at just 0.8%, per RBI Financial Stability Report data. Banks are lending more carefully and watching earlier. A CMA that survives the ten checks below is built for exactly that environment.

The 10-row pre-submit self-audit

Run this table like the officer runs the file: left column is the failure, middle column is what they check, right column is what you verify before the file leaves your desk.

If you want this table as a working file, our free CMA self-audit workbook walks the same ten checks with space to record each figure and its source. Work through it line by line before your CA finalises the file; the hour it takes is cheaper than the week a returned file costs.

Sources

• RBI Master Directions on Fraud Risk Management (July 15, 2024), via Economic Laws Practice: board-approved fraud risk management policy; mandatory early warning signal and red-flagging framework (clause 8.3); EWS triggers including material discrepancies in the annual report, significant inconsistencies within the annual report, significant movements in inventory and receivables disproportionate to turnover, frequent invocation of BGs and devolvement of LCs, and increase in fixed assets without corresponding long-term sources.
• RBI speech: industrial credit in India: on the April 1997 withdrawal of MPBF prescription and greater operational freedom for banks in working-capital assessment.

Regulatory claims were checked against the primary RBI sources above. Threshold figures (1.25 DSCR, 1.33 current ratio, MPBF methods) are described as bank-policy convention, not RBI law; the binding source is the individual sanction letter.

A rejection is rarely about one ratio. It is about where the file broke in the credit appraisal process, step by step: documents, financial preparation, proposal and rating each have their own failure points, and the CMA sits in the middle of them.

Frequently Asked Questions

Last updated: September 2026. Regulatory claims were checked against the RBI Master Directions on Fraud Risk Management (July 15, 2024); bank-policy thresholds are stated as convention, not RBI law.