CMA Data: What a Bank Actually Reads, and the Four Ratios That Decide It

Every business that has applied for a working capital limit has been asked for "CMA data", usually at short notice, often with a blank template attached. Treated as a form-filling exercise it produces a file the credit officer quietly distrusts. Treated properly it is the single most influential document in the application.

Understanding why means understanding what the officer does with it.

What CMA data is

Credit Monitoring Arrangement data is a standardised presentation of a borrower's financial position — past, present and projected — in a format that lets a lender compute a credit decision consistently across borrowers.

Its structure descends from the Tandon and Chore Committee recommendations on bank lending to industry, and although the Reserve Bank long ago freed banks to set their own norms, the format and the methods have stuck because they work and because every credit officer was trained on them.

It conventionally comes in six statements:

The usual period covered is two audited years, the current year on a provisional or estimated basis, and one or two projected years. Term loan proposals extend the projections across the repayment tenor.

What the officer computes from it

This is the part worth internalising, because these are the numbers that determine the answer. A credit officer will compute them before reading your narrative.

Current ratio

Current assets divided by current liabilities. The conventional benchmark for working capital lending is 1.33, which follows arithmetically from the second method of lending: if the borrower funds 25% of current assets from long-term sources, the ratio lands at about 1.33.

Below it, you are asking the bank to fund a larger share of your working capital than the standard assumes. That is not fatal, but it has to be explained — and explained by you, in the file, rather than discovered by the officer.

TOL/TNW

Total outside liabilities divided by tangible net worth: how leveraged the business is. Tangible net worth is paid-up capital plus reserves, less intangibles, accumulated losses and, in most banks' methodology, loans to related parties and investments in group entities.

That deduction is where files come apart. A balance sheet showing a healthy net worth can produce a thin tangible net worth once advances to sister concerns and director loans are stripped out. Compute it the bank's way before you submit, and know your number.

DSCR

Debt service coverage ratio, for term loans: cash available for debt service — profit after tax plus depreciation plus interest on term debt — divided by the debt service obligation of interest plus principal repayment.

Lenders generally want an average DSCR of around 1.5 to 2 across the tenor, and a minimum in any single year of not less than about 1.25. The average matters less than the worst year, so a projection that averages well but dips below 1 in year two will be questioned on year two.

MPBF

Maximum permissible bank finance — the output of Form V, and in practice your limit.

Under the second method of lending, which is the usual one:

1. Compute working capital gap = current assets less current liabilities other than bank borrowings.
2. Compute 25% of current assets — the margin the borrower is expected to bring.
3. MPBF is the working capital gap less that margin.

Under the first method, which applies to smaller limits, the margin is 25% of the working capital gap rather than of current assets, which produces a larger permissible figure.

The practical consequence: MPBF is driven by your current assets, and current assets are driven by your inventory and receivable holding periods. A business that collects in 45 days gets a materially smaller limit than an identical business that collects in 90 — which is the right answer, and also the lever you actually control.

Where files actually fail

In the files we are asked to rescue, the reasons are consistent and they are almost never that the business was unviable.

The projections do not tie to the audited figures. Sales growing 40% a year when the last three audited years grew 6%, with no new contract, no new capacity and no explanation. Every subsequent number in the file inherits the doubt.

Inconsistency between statements. Form II profit that does not reconcile to the Form III reserves movement. Depreciation in the operating statement that does not match the fixed asset schedule. Each one is a small arithmetic error and each one costs credibility.

Holding periods that improve for free. Debtor days falling from 90 to 55 in the projected year, with no change in customer mix or collection policy. This is the most common single plug, and officers look for it specifically.

Contingent liabilities omitted. Guarantees given for group companies, disputed tax demands, pending litigation. These surface in the audited accounts and in the credit bureau report, and omitting them from the file is worse than disclosing them.

Related-party exposure not disclosed. Loans to and from directors and group entities, which the bank will find in the notes and which hit tangible net worth.

No reconciliation to GST and income tax filings. The bank's analyst will compare declared turnover in the CMA file against GST returns and the income tax return. A difference is often perfectly legitimate — different definitions, exempt supplies, timing — but it needs a note explaining it, not silence.

Preparing it so it stands up

1. Start from audited figures, not from management accounts. Use the signed financial statements, and keep the audit report with the file.
2. Build projections from drivers, not from a growth percentage. Volume times price, with capacity and a named source of demand behind the volume.
3. Hold the ratios honest. Keep the operating cycle flat unless something specific changes it, and if something does, say what.
4. Compute the bank's ratios yourself — current ratio, TOL/TNW, DSCR, MPBF — before submission, and write a line on each.
5. Reconcile to filings. Turnover to GST, profit to the income tax return, with a note on the differences.
6. Disclose everything. Contingent liabilities, related-party balances, existing limits with other lenders, any restructuring history.
7. Write a one-page covering note stating the ask, the purpose, the security offered and the three numbers you want the officer to take away. It will be read.

The point

A bank is not deciding whether your business is good. It is deciding whether it can be repaid, from cash flows it can see, with security it can realise, by a borrower whose numbers it can trust. CMA data is where trust is either established or lost, and it is lost through inconsistency far more often than through weakness.