CMA Data Mistakes That Get Bank Loans Rejected — and How Credit Officers Actually Read the File
- A credit officer reads CMA data in a fixed order — current ratio, TOL/TNW, DSCR, holding periods, then the assumptions — and stops at the first number that fails the bank's benchmark.
- Most rejections are internal inconsistencies, not weak businesses: sales that grow 40% while debtors stay flat, or a profit margin the audited accounts never showed.
- Every figure in the CMA must tie to a document the bank already has — GSTR-3B turnover, the ITR, the audited balance sheet — or the whole file loses credibility.
Banks seldom reject a working-capital or term-loan proposal by saying "your business is weak". They say the file is "returned for clarification", "needs revision", or simply goes quiet. Behind that language is usually a CMA (Credit Monitoring Arrangement) data set that failed one of a short list of tests the credit officer applies in the first fifteen minutes. This guide lists those tests, the mistakes that fail them, and the fix for each — written from the officer's side of the desk.
How the file is actually read
| Order | What the officer checks | Benchmark most banks apply (MSME) | If it fails |
|---|---|---|---|
| 1 | Current ratio in the projected year | 1.33 or better (1.17–1.25 accepted for some schemes) | File returned: working capital under-funded or inventory inflated |
| 2 | TOL / TNW (total outside liabilities to tangible net worth) | Below 3:1 (4:1 tolerated for traders) | Promoter contribution demanded or limit cut |
| 3 | DSCR (term loans) | Average 1.5, minimum 1.2 in any year | Tenure stretched or loan amount reduced |
| 4 | Holding periods — stock, debtors, creditors | Consistent with the last audited year and the industry | MPBF recomputed on lower holdings; limit falls |
| 5 | Sales growth versus capacity, debtors and creditors | Growth explained by capacity, orders or market | Projections treated as optimistic; discounted 20–30% |
| 6 | Tie-up with GSTR-3B, ITR and audited accounts | Exact match for past years | Credibility of the whole file questioned |
The mistakes, in the order they are caught
1. Projected current assets that do not fund themselves
Sales projected to double, inventory and debtors projected to stay the same. The officer knows a doubling of sales needs more stock and more credit to customers; a flat current-asset line means the proposer has not modelled the working-capital cycle. Fix: build current assets from holding periods — stock in days of cost of sales, debtors in days of sales, creditors in days of purchases — and let them move with sales.
2. A margin the audited accounts never showed
Last year's net margin was 3%; the projection assumes 9% from year one. Unless there is a documented reason (new machine, dropped loss-making line), the officer takes the historical margin and reworks DSCR — which then fails. Fix: project margins from the last two audited years, improving gradually with a stated reason per year.
3. Promoter contribution missing from the balance sheet
The proposal says 25% margin; the projected balance sheet shows no new capital or unsecured loan from promoters. Fix: show the margin money as capital or subordinated unsecured loan in the first projected year, with a source-of-funds note.
4. Figures that do not tie to GST and income-tax filings
Turnover in the CMA for FY 2025-26 differs from GSTR-3B by ₹60 lakh; the ITR shows another number. Officers now download GSTR-3B summaries themselves. Fix: reconcile and footnote every difference — exempt sales, exports, branch transfers, year-end cut-off.
5. No assumptions sheet
Seven statements and no page saying where the numbers came from. The officer cannot defend the proposal to the sanctioning committee. Fix: a one-page assumptions sheet: capacity, price, growth, holding days, interest rate, depreciation method, tax rate, repayment schedule.
6. Term-loan repayment that starts before the asset earns
Machine delivered in month six, instalments from month one. DSCR in year one collapses. Fix: model the moratorium the bank actually offers and put interest during construction into project cost.
7. Treating the limit as the need
Asking for ₹1 crore because that is what the business "wants", when MPBF on the projected cycle supports ₹55 lakh. Fix: compute MPBF (Tandon second method) from the projections and ask for that figure, or justify the gap with order books.
8. Inconsistent depreciation, tax and interest
Depreciation at Income-tax rates in the P&L but Companies Act rates in the balance sheet; tax at 25% for a proprietorship; interest computed on the opening balance only. Each is small; together they signal the file was assembled, not prepared.
A self-check before submission
- Past years tie exactly to audited accounts, ITR and GSTR-3B (with reconciliations).
- Current ratio ≥ 1.33 and TOL/TNW ≤ 3 in every projected year.
- DSCR ≥ 1.5 average and ≥ 1.2 in each year of the term loan.
- Holding periods within ±15% of the last audited year unless explained.
- Promoter contribution visible in the balance sheet with a source.
- Loan asked = MPBF or term-loan requirement computed in the file.
- Assumptions sheet, repayment schedule and sensitivity (sales −10%) attached.
The DPR and CMA Studio builds the seven statements from holding periods and assumptions rather than from typed-in totals, computes MPBF and the five ratios, and flags any year that breaches a benchmark before you print — which is the same check the officer will run.
Common mistakes
- Copying last year's CMA and changing the dates. Officers compare with the previous sanction note.
- Round numbers everywhere. Sales of exactly ₹5,00,00,000 in every projected year signal a guess.
- Negative net working capital in the projected year while asking for a cash-credit limit.
- Ignoring GST on the working-capital cycle. Input credit stuck in GSTR-2B mismatches is real cash blocked; model it in current assets.
- No sensitivity analysis. A file that survives a 10% sales drop with DSCR above 1.2 is sanctioned faster.
- Submitting without the auditor's or CA's certification where the bank's format asks for it.
Do it yourself in minutes — free to try, no login needed.
Open DPR and CMA Studio →Frequently asked questions
What is the most common reason banks reject CMA data?
Internal inconsistency — projections that do not follow the working-capital cycle (sales rising while stock and debtors stay flat), margins that depart from the audited history without a reason, or past-year figures that do not match GSTR-3B and the income-tax return. The business may be sound; the file fails its own arithmetic.
What current ratio do banks expect in CMA data?
A minimum of 1.33 in the projected year for most working-capital assessments, with 1.17 to 1.25 accepted under some MSME schemes. Below that, the bank concludes the business is under-capitalised and asks for promoter contribution or reduces the limit.
How is the working-capital limit decided from CMA data?
Most banks use the Tandon second method of lending: maximum permissible bank finance equals 75% of projected current assets less other current liabilities, with the balance 25% funded from long-term sources. Holding periods for stock, debtors and creditors drive the current-asset figure, so unrealistic holding days inflate or deflate the limit.
Should CMA data match GST returns exactly?
For completed years, yes, or every difference should be reconciled in a footnote — exports, exempt supplies, branch transfers, cut-off differences. Credit officers download GSTR-3B summaries and compare; an unexplained gap undermines the whole file.
General information for FY 2026-27, not professional advice for your specific case. Rules change — verify against the latest notification or ask a KyaTax expert.
Related: All free tools · More guides · Virtual CFO