A message we received from a founder in Lower Parel, sent at 11.40 pm:

“Our audited accounts show ₹1.8 crore profit last year. Best year ever. So why is my overdraft at the limit and why did I just delay salaries by four days? Where did the money go?”

He was not mismanaging anything. His accounts were clean, his auditor was happy, his tax was paid. He had simply been reading the wrong report. The profit and loss statement tells you whether your business model works. It says nothing about whether you will make payroll on the 1st. Those are different questions, and in a growing Indian business they have different answers more often than you would think.

This article explains where his ₹1.8 crore went, the five cash traps that are specific to running a business in India, and the one tool that would have shown him the problem in June instead of at midnight in March.

Applies to: Owner-managed businesses with ₹5–100 crore turnover · Family businesses in transition · Founders who make collection calls themselves

Profit is an opinion. Cash is a fact.

Profit is calculated on an accrual basis: you record the sale when you raise the invoice, not when the customer pays. That is correct accounting and it is also why profit and cash drift apart. Here is the bridge for the founder above (figures rounded):

Read the middle rows again. ₹3 crore of his profit was converted into receivables and inventory: sales he had made but not collected, and goods he had bought but not sold. The business was growing, and growth eats cash before it produces it. Add the machinery, the loan principal and the household, and a ₹1.8 crore profit became a ₹1.9 crore hole in the overdraft.

None of this appears on the P&L. All of it appears in the bank.

The five cash traps that are specific to India

Every business has a working-capital cycle. Indian businesses have five extra pressures layered on top, and most owners discover them one at a time, expensively.

1. You pay GST on invoices your customer hasn’t paid

GST is due on the 20th of the month following the invoice, whether or not the customer has paid you. Invoice ₹1 crore in September at 18% and ₹18 lakh leaves your account on 20 October. If your customers pay at 75 days, you have financed the government’s tax for well over a month on every sale you make. At ₹20 crore of annual turnover, that is something like ₹40–50 lakh of cash permanently tied up in GST timing alone.

The rule cuts the other way too: if you have not paid a supplier within 180 days of their invoice, the input credit you claimed has to be reversed with interest. Stretching vendors past six months is not a free source of funds.

2. Your customers deduct TDS, and you get it back next year

Corporate customers deduct tax at source before paying you: 1% or 2% on contracts, 10% on professional fees, a smaller slice under purchase TDS on large volumes. That money is yours. It sits with the income-tax department, visible in your 26AS, until you file your return and the refund is processed, typically 12 to 18 months after the deduction. A services firm with ₹15 crore of billings at 10% TDS is lending the government ₹1.5 crore a year, interest-free, while paying 11% on its overdraft.

3. The MSME 45-day rule ended the old way of managing payables

Under Section 43B(h) of the old Act (now Section 37(2)(g) of the Income-tax Act, 2025), if you have not paid a micro or small enterprise supplier within 15 days (no written agreement) or 45 days (written agreement, and 45 is the ceiling), the expense is not deductible in that year. It moves to the year you actually pay. Unlike other Section 43B items, paying before the return due date does not rescue it. On top of the disallowance, the MSMED Act charges compound interest at three times the RBI bank rate, and that interest is not deductible either.

What makes this bite harder now: since 1 April 2025, a “small” enterprise is one with up to ₹25 crore invested in plant and equipment and up to ₹100 crore of turnover, and “micro” goes up to ₹2.5 crore and ₹10 crore. That is most of the vendor base of most Mumbai businesses. The strategy your father used, of paying suppliers when the customers paid him, now creates a tax bill in March. Your payables are no longer a flexible cushion; they are a clock.

4. Advance tax and the March squeeze

Income tax is not paid once a year. Advance tax is due in four instalments: 15% by 15 June, 45% by 15 September, 75% by 15 December and 100% by 15 March. Now look at March in a typical business: the advance tax balance on the 15th, GST on the 20th, TDS on the 7th, PF and ESI on the 15th, year-end vendor clearing to protect the MSME deduction, annual bonuses, and the auditor asking why the overdraft is at its limit on 31 March. Businesses that are fine for eleven months routinely run out of cash in the twelfth, and it is entirely predictable.

5. Growth itself

This is the trap that caught the founder above. Take a business that adds ₹1 crore of monthly sales. At 75-day collections, that is ₹2.5 crore of new receivables. If cost of goods is 70% of sales and you hold 45 days of stock, that is roughly ₹1 crore more inventory. Suppliers paid at 30 days give back about ₹0.7 crore. Net: about ₹2.8 crore of additional working capital for every extra ₹1 crore of monthly sales, all of it needed before the profit on those sales arrives. Growth is a cash-consuming activity until the cycle turns, and the faster you grow, the deeper the hole before it does.

Why the annual budget doesn’t save you

Most businesses at this size have a budget. It was built in March, it is monthly, it is in P&L format, and nobody has opened the file since May. We call it a dead budget: a document, not a decision tool. It cannot tell you that the second week of September is a problem, because it does not know that your largest customer pays on the 25th, that advance tax falls on the 15th, and that the MSME clock on your packaging supplier runs out on the 18th.

What replaces it is a rolling 13-week cash forecast: a simple, week-by-week picture of cash in and cash out for the next quarter, updated every week. One week drops off the front, one is added at the back, and the picture is always thirteen weeks deep.

Why weekly, and why thirteen? Because a week is the unit in which cash problems actually happen (the 7th, the 15th, the 20th), and thirteen weeks is the horizon in which you can still do something about them: pull a collection forward, renegotiate a delivery, defer a capex, or walk into the bank with a plan rather than a request.

How to build one, in a spreadsheet, this afternoon

The rows. Start with the opening bank balance plus undrawn overdraft. Then receipts, listed by customer, placed in the week they will actually pay based on their history, not their promise. Then payments, in order of how non-negotiable they are:

1. Statutory dues on their dates: TDS (7th), PF and ESI (15th), GST (20th), advance tax instalments
2. Payroll
3. Loan EMIs and interest
4. Rent and utilities
5. MSME suppliers, sequenced by when their 15/45-day clock expires
6. Other suppliers
7. Capex
8. Owner drawings

Then the closing balance, and a line showing how far above or below your minimum buffer you are.

The columns. Thirteen weeks. Here is what the first four look like for a ₹30 crore business, in lakhs, to show the shape:

Nothing in that table is a surprise once it is written down. Week 3 is tight because advance tax, GST and PF all land in the same seven days while the big collections come in week 1 and week 4. The owner now has three weeks’ notice, and the options are obvious: ask the two large customers due in week 4 for a partial payment in week 3, shift the discretionary vendor payment from week 3 to week 4, or arrange a short-term enhancement with the bank in week 1 while nobody is panicking. Without the table, the same owner discovers the problem on the morning of 15 September, and the only option left is to delay the advance tax and pay interest on it.

The weekly ritual. Thirty minutes on Monday morning. Replace last week’s forecast with actuals. Note what slipped and why (the customer who paid late twice is now a 90-day customer in your forecast, not a 60-day one). Add the new week thirteen. Look at the buffer line. That is the whole discipline, and in our experience it takes about three weeks before an owner stops trusting the old way of doing things.

Three rules that make it work

• Forecast collections on history, not hope. Your debtor ageing report tells you the truth about each customer.
• Never fund capex from the overdraft. The machine is a five-year asset; the OD is a one-year facility. Term it out.
• Set a buffer (four to six weeks of fixed costs is a reasonable start) and treat any week below it as a decision, not a surprise.

What changes after a quarter

Owners who run this for three months report the same things. The overdraft comes down, because receivables get chased in the week they are due rather than the week the cash runs out. The March squeeze disappears, because advance tax was provided for in week 1 of the quarter, not week 11. Hiring and capex decisions get made against a number instead of a feeling. And the conversation with the bank changes completely: a 13-week forecast, reconciled to actuals for a few months, is exactly what a credit manager wants to see before enhancing a limit, and almost nobody at this size brings one.

When you need a CFO (and why probably not a full-time one)

The forecast is the tool. The harder question is who owns it. Signals that the answer is no longer “the accountant, in his spare time”:

• Turnover has crossed ₹10–15 crore and the overdraft has been above 80% utilisation for more than two consecutive months
• The owner personally makes collection calls or approves every payment
• There are three or more bank accounts and nobody reconciles them daily
• The bank has asked for CMA data or projections and the request sat for a fortnight
• A second product line, a second location, or a promoter loan is being funded from working capital
• The March numbers were a surprise, again

A full-time CFO in Mumbai costs ₹50 lakh to well over ₹1 crore a year, and a ₹30 crore business does not