Profitable yet out of cash: the cash conversion cycle for SME owners

A dry-goods wholesaler's income statement shows a profit after tax of 1,280 million dong for the year. The bank balance at year end is 620 million, 880 million lower than a year ago. The owner asks the bookkeeper: where did the profit go? The company is my construction for illustration; every number about it below is invented.

Why can a profitable company see its cash fall?

Because profit and cash follow two different recording rules. Under accrual accounting, revenue is booked when goods are delivered and costs when they are incurred, whatever the date cash moves. The gap between profit and cash lives in receivables, stock and payables.

Invented example. The wholesaler supplies supermarkets on 60-day contract terms. Sales this year are 18,000 million dong, up 50 per cent from 12,000 million last year, with a 20 per cent gross margin in both years. This year: cash operating costs 1,800, depreciation 120, interest 80, and tax at 20 per cent of the 1,600 pre-tax profit, which is 320, leaving 1,280 profit after tax. Tax and interest are paid in cash during the year.

This is where the three statements link up. A cash-flow statement prepared by the indirect method starts from profit after tax, adds back depreciation, adjusts for the change in working capital, and arrives at the closing cash on the balance sheet:

Retained earnings rise by 1,280 less the 200 dividend, which is 1,080, and closing cash is 1,500 minus 880, which is 620. These are standard accounting identities rather than research findings, so I cite nothing for them. Operations absorbed 780 million, yet the company still paid a 200 million dividend and borrowed another 400 million.

How is the cash conversion cycle calculated?

Add receivable days to inventory days, then subtract payable days. Deloof describes the cash conversion cycle as the time lag between paying for raw materials and collecting the sales of finished goods; the longer it is, the more is invested in working capital (Deloof, 2003, p. 3). He computes receivables times 365 over sales, inventories times 365 over cost of sales, and payables times 365 over purchases (Deloof, 2003, p. 4). The working capital cycle framework uses the same three components.

My convention is close: year-end balances, 365 days, receivables over sales, inventory and payables over cost of sales. Over purchases (15,480, which is cost of sales plus the 1,080 stock build-up) payable days would be 42 rather than 46.

Supermarkets pay after 71 days against 60 in the contract, and goods sit in the warehouse 61 days instead of 50. Operating working capital (receivables plus inventory minus payables) rises from 1,920 to 4,100, an increase of 2,180. Had the days stayed at last year's levels, 50 per cent more sales would have needed only 2,880, an increase of 960; the remaining 1,220 is the price of 27 extra days. Profit after tax plus depreciation is 1,400, so the working capital needed for growth alone could have paid for itself (before fixed assets and dividends): the longer cycle turns cash flow negative.

In my judgement this convention treats sales as even across the months, so a fast-growing firm will see receivable days inflated.

How much cash does each day of reduction free?

One day of receivables is worth 18,000 over 365, about 49.3 million dong; one day of inventory or payables is worth 14,400 over 365, about 39.5 million. Multiply the days you want to cut by that figure to get the cash released, before counting what the move costs.

The three rows add up to about 1,503 million, enough to cover the 880 million lost. That is arithmetic, not a forecast. Deloof gives the other side: a large inventory reduces the risk of a stock-out, and a generous trade-credit policy may lift sales because customers can check quality before paying (Deloof, 2003, p. 2). Supermarkets asked to pay sooner may demand a price cut, and a stock-out may cost shelf space (my inference). Neither cost is in the table. The method suits a business that sells on credit or carries a lot of stock; it helps little where customers pay at the till.

Is stretching supplier payments the cheapest lever?

Not necessarily. It frees the most cash, but the cost lies in forgone early-payment discounts and in the supplier relationship. Deloof cites Svensson's 1996 survey, in which 75 per cent of Belgian firms offered a discount for prompt payment, averaging 3 per cent (Deloof, 2003, p. 2). The post knows that survey only through Deloof's account of it.

Take assumed terms of "2/10, net 45": a 2 per cent discount for paying within 10 days, otherwise the full amount in 45. Skipping the discount is borrowing for 35 days at a price of 2/98, about 2.04 per cent. Annualised as simple interest, 2.04 per cent × 365/35 is about 21.3 per cent. If your working-capital loan costs far less, paying early for the discount beats holding the cash. The company in the example already pays after 46 days against terms of 45, so stretching to 60 is paying late, not using credit it has been granted.

The Belgian data do not show that paying late pays. Median payable days in the two lowest gross-operating-income groups are 62.5 and 65.8, against 47.1 to 54.3 in the other groups (Deloof, 2003, p. 10). Deloof says the figures fit the hypothesis that less profitable firms wait longer to pay their bills (Deloof, 2003, p. 10): late payment is a consequence of a shortage of cash, not proof that it helps.

What did Deloof find in 1,009 Belgian firms, and how far can it be trusted?

He found that longer receivable and inventory days went with lower profit, but the link weakens when profit is measured after depreciation. The sample is 1,009 large non-financial Belgian firms, 5,045 firm-year observations over 1992 to 1996 (Deloof, 2003, p. 4). The mean cash conversion cycle is 44.5 days, made up of 54.6 receivable days, 46.6 inventory days and 56.8 payable days (Deloof, 2003, p. 5).

In the fixed-effects regression, each extra receivable day goes with gross operating income lower by 0.048 percentage points (measured on total assets minus financial assets; the mean is 12.2 per cent) (Deloof, 2003, pp. 4-5, 7). Inventory days also carry a negative coefficient (p = 0.015) (Deloof, 2003, p. 7). The combined cycle, however, has a coefficient not different from zero (p = 0.668); Deloof explains that receivables, inventory and payables all go with lower profit, and payables are subtracted in the cycle (Deloof, 2003, p. 8).

With net operating income (after depreciation), the inventory coefficient is not significant (Deloof, 2003, p. 9), and the cycle's coefficient in the fixed-effects regression is actually positive, 0.10 × 10⁻³ with p = 0.006 (Deloof, 2003, p. 15). The manuscript's text does not discuss that positive sign.

The first limit is causation. Deloof writes that it cannot be ruled out that profitability affects working capital rather than the reverse, for instance falling sales lowering profit and raising stock (Deloof, 2003, p. 9). The second is scope: one country, five years, large firms, and only those with statements for every year from 1991 to 1996 (Deloof, 2003, p. 3), a condition that in my inference excludes firms that disappeared midway. The sample has no small businesses and no Vietnam.

Going deeper: what the post rests on, and what it does not

The post uses the full text of Deloof's (2003) author manuscript in the University of Antwerp repository, 16 pages including tables and figure (Deloof, 2003, p. 1). Page numbers here count the PDF file, with the cover as page 1, so they are one higher than the numbers printed on the manuscript and are not the printed pages of the Journal of Business Finance & Accounting. The site library lists pages 573–587; Crossref and OpenAlex give 573–588; the printed version has not been checked.

The manuscript has one inconsistency: page 5 gives the median cycle as 59.6 days, but Table I on page 12 shows 59.56 in the standard-deviation column and 34.59 as the median. The post uses only the mean.

The library's working capital cycle framework credits Richards and Laughlin (1980) as its origin. No legitimate full text was available, only the opening lines via OpenAlex, so the definition here comes from Deloof. The three-statements framework credits the textbook by Elliott and Elliott (2019); the post uses no part of it and treats the links as standard accounting identities. The wholesaler and every number in the tables are invented by me and computed by script.