A ratio like Revenue ÷ Assets asks how many rupees of sales each rupee parked in the business generates per year. Turn any turnover upside down and multiply by 365. You get something even more vivid: how many days money sits in that asset before it comes back. Speed is DuPont's second lever, and it's the lever management can least fake. Accounting choices can dress up margins for a while. They can't change how fast inventory physically leaves a warehouse.
The five turnovers, with the conventions that trip people up:
| Ratio | Formula | Days version | Convention to defend |
|---|---|---|---|
| Total asset turnover | Revenue ÷ average total assets | — | Average of opening and closing assets; ending balances distort growers |
| Fixed asset turnover | Revenue ÷ average net PP&E | — | Net of depreciation; old, written-down plants flatter the ratio |
| Inventory turnover | COGS ÷ average inventory | DIO = 365 ÷ turnover | COGS, not revenue: inventory sits on the books at cost |
| Receivables turnover | Credit sales ÷ average receivables | DSO = 365 ÷ turnover | Revenue is the standard proxy when credit sales aren't split out |
| Payables turnover | Purchases ÷ average payables | DPO = 365 ÷ turnover | Purchases = COGS + ending inventory − beginning inventory; COGS is the common fallback proxy |
Advertisement
Is inventory turnover based on COGS or revenue?
The numerator is COGS because inventory is carried at cost, while revenue includes the markup. Divide revenue by inventory and you count the gross margin as if it were sitting on the shelf. Meet this week's synthetic drill company, Kaveri Consumer Products (all figures ₹ cr). Revenue is 7,300, COGS 5,110 and average inventory 840. Done correctly, the turn is 5,110 ÷ 840 = 6.08×, so DIO = 365 ÷ 6.08 = 60 days. With revenue in the numerator you'd get 7,300 ÷ 840 = 8.69×, which is 42 days. Same shelf, 18 phantom days of speed. That's the 30% gross margin laundered into fake velocity. The revenue-based version isn't illegal, and some data vendors publish it. You just need to know which one you're holding before you compare two companies.
DPO asks how long you take to pay suppliers, so the denominator should be what you bought. What you sold is the wrong base. What you bought is what you consumed (COGS) plus what you added to the shelf (the inventory build). For Kaveri, purchases = 5,110 + (880 − 800) = 5,190. With average payables of 640, DPO = 365 × 640 ÷ 5,190 = 45 days. Using the COGS shortcut gives 45.7 days here. That's close, because inventory barely moved. In a year of heavy stocking or destocking the two drift apart, so use one convention throughout when you compare companies.
Why use average balances and 365 days?
Flows such as revenue and COGS build up over the year. Stocks such as inventory and receivables are one-day photographs. Divide a year's flow by a year-end stock and the periods don't match, and for a company that doubled during the year that badly overstates the days. Use (opening + closing) ÷ 2 when you have two balance sheets. Use ending balances only when you must, and then use them for every company in the comparison. One more convention is 365 vs 360 days, since bank credit analysts often use 360. Either works. Mix them and your answers shift by 1.4% without anyone noticing.
To finish Kaveri, average receivables of 600 give DSO = 365 × 600 ÷ 7,300 = 30 days. Average total assets of 4,800 give asset turnover of 7,300 ÷ 4,800 = 1.52×, so each rupee of assets produces ₹1.52 of sales a year. Average net PP&E of 2,000 gives fixed asset turnover of 3.65×. One caution about a seductive ratio you'll meet in screeners, working-capital turnover (revenue ÷ average operating WC). It explodes toward infinity as WC approaches zero, and it flips sign when WC goes negative. For HUL or Costco it's meaningless. Stick with days.
How does seasonality distort the ratios?
Seasonality is the photograph problem squared, because every stock ratio inherits the balance-sheet date. Indian companies close on March 31, after the January to March quarter. Costco closes its fiscal year on the Sunday nearest August 31. Most US retailers close at the end of January, after holiday inventory has been sold down and before spring stocking, so their year-end inventory is the leanest photograph of the whole year. Compare same date with same date and prefer averages. For anything seasonal, read the quarterly balance sheets too.