Net Debt and Leverage: What the Ratios Measure and When They Mislead

Net debt is what a business owes on borrowings less the cash it holds against them. Anjani Stationers, an invented stationery maker, closed at Rs 5,20,000. At 0.10 times earnings before interest, tax, depreciation and amortisation (EBITDA), the closing figure looks like almost nothing. Its average borrowing across the year was about Rs 37,00,000, or 0.69 times. Both are correct. One is a photograph taken on a chosen date and the other describes the year. The gap between them is the point.

The panel opens on Anjani Stationers exactly as its year two accounts report it: a term loan of Rs 4,20,000, a lease liability of Rs 2,00,000 current and Rs 4,00,000 non-current, the facility drawn at nil and cash of Rs 5,00,000. Those lines give gross borrowings of Rs 10,20,000 and net debt of Rs 5,20,000. EBITDA is assembled rather than read off anything: earnings before interest and tax (EBIT) of Rs 41,50,000 plus depreciation and amortisation of Rs 12,00,000 gives Rs 53,50,000. On those figures net debt to EBITDA is 0.10 times, gearing is 6.7 per cent and interest cover is 11.9 times. Leave the limit at 1.50 times and the strip at the foot of the panel reads 0.10 times on the reporting date, 0.60 times on the average borrowings across the year and 0.94 times at the maximum, all three deducting the same cash balance, with headroom of 1.40, 0.90 and 0.56 turns. The column beside them, taking no cash off at all, reads 0.19, 0.69 and 1.03 times on the same three measurements. Both columns are correctly computed and they disagree by a third. A net debt figure therefore has to travel with the basis it was built on.

The reason lies one level down. A balance sheet is not a film of the year. A balance sheet is a single frame, taken on one date that the calendar chose rather than the business, and every figure standing on it inherits that date. Borrowings are a balance at a moment. Cash is a balance at a moment. So any ratio built from them describes that moment and nothing else, however confidently it is quoted.

The income statement works the other way. Revenue, EBITDA, EBIT and finance cost are all totals for twelve months of trading, and no single day can flatter or spoil them. The difference between a balance and a total sounds like bookkeeping trivia until the two kinds of figure are put into the same ratio. At that point it decides whether the answer describes a business or describes a date.

Everything the arithmetic needs is already in the accounts: the kinds of borrowing in the borrowings note, EBITDA, EBIT and the finance cost in the statement of profit and loss, and the equity of Rs 1,42,00,000 on the balance sheet. The accounts do not give the inclusion list, the three ratios built from it, or the date each ratio was taken on.

What counts as debt here, and what does not?

Start with the inclusion list. Everything downstream depends on it. Net debtMoney owed on borrowing arrangements, less the cash held against it. A single figure meant to answer what would be left owing if the cash on hand were applied to the borrowings today. counts what a business owes under an arrangement that obliges it to hand over cash on dates already fixed, and nothing else. Term loans count. Overdrafts and working capital facilities count while they are drawn. Debentures and bonds count. Lease liabilities count. A lease liability is a promise to pay fixed amounts on fixed dates, and that is the same obligation wearing a different name.

Now the items that get wrongly swept in. Trade payables are money owed to suppliers for goods already delivered, and they carry no interest and no fixed repayment schedule beyond ordinary credit terms; they are how trade works, not how funding works. A contract liability is money a customer paid in advance, and it will be settled by delivering notebooks rather than by paying cash. Deferred tax is an accounting timing difference, not an amount anybody has agreed to pay on a date. Provisions are estimates of obligations whose amount or timing is uncertain. A borrowing is precisely the opposite: its amount and its dates are fixed.

The line between the two lists is drawn by whether there is a contractual obligation to deliver cash on fixed dates, and different analysts draw it in different places, so a net debt figure quoted without its inclusion list cannot be compared to anything. The lease liability is where the disagreement almost always sits. Including it says that a four-year commitment to pay rent on a warehouse is a borrowing in substance. Excluding it says that rent is an operating cost that happens to have been capitalised. Both positions are held by serious people. On Anjani Stationers' own figures the choice changes the sign of the answer, as the computation below shows.

How is net debt computed here, and why does one honest choice flip the sign?

The computation itself is two subtractions. Add up the borrowings, take off the cash, and stop. Anjani Stationers held cash of Rs 5,00,000 on 31 March, so with the lease liability included the arithmetic is Rs 4,20,000 plus Rs 6,00,000 less Rs 5,00,000, giving net debt of Rs 5,20,000. Net debt of Rs 5,20,000 is the figure Anjani Stationers' accounts carry forward.

Now run it again on the other inclusion list. Leave the lease liability out and the borrowings side is the term loan alone, Rs 4,20,000, against cash of Rs 5,00,000. The answer is minus Rs 80,000. The same business on the same date is either Rs 5,20,000 in net debt or Rs 80,000 in net cash, and the only thing that moved was one line on an inclusion list.

The sign flip is not a rounding argument. One version of Anjani Stationers is a borrower and the other holds more cash than borrowings, and both are defensible from the same audited figures. The flip is why a credit memo states its inclusion list before it states its number, and why a net debt figure lifted from a screener or a summary table is close to useless until its inclusion list is known.

How are the three leverage ratios built, and what does each one compare?

Three ratios do most of the work, and the useful way to hold them apart is not by what they are called but by what kind of figure sits on each side of the line. A stock measureA figure that is a balance at one instant, such as cash, borrowings or equity. Change the instant and the figure changes, whatever the business did over the period. is a balance at one instant. A flow measureA figure that accumulates over a period, such as revenue, EBITDA or finance cost. A flow measure sums up the whole period, so no single date inside the period can move it. accumulates across the whole period. Every ratio here is one of three combinations.

Net debt to EBITDA divides a stock by a flow. Anjani Stationers' Rs 5,20,000 of net debt over EBITDA of Rs 53,50,000 gives 0.10 times. The conventional reading is a rough number of years of trading earnings needed to clear the borrowings, so the ratio is quoted as turns rather than as a percentage. GearingThe share of a business funded by borrowing rather than by the shareholders. Written here as borrowings divided by borrowings plus equity, and stated that way every time because the phrase is also used for other formulas. divides a stock by a stock: gross borrowingsThe total owed on borrowing arrangements before any cash is deducted. Also called gross debt, and the figure a balance sheet actually shows. of Rs 10,20,000 over borrowings plus equity of Rs 1,52,20,000, giving 6.7 per cent. Interest coverHow many times the year's operating profit covers the year's interest bill. Both figures are totals for the whole period, so the measure has no reporting date inside it at all. divides a flow by a flow: EBIT of Rs 41,50,000 over the finance cost of Rs 3,50,000, giving 11.9 times.

Count the stock figures in each ratio. That count is exactly how exposed the measure is to the date it was computed on: two stocks means fully exposed, one stock means half exposed, and no stocks means immune. Gearing carries two, so both sides of it move with the date. Net debt to EBITDA carries one, so its numerator moves while its denominator cannot. Interest cover carries none, and that is not a small technical detail. It is the reason interest cover is the one ratio the calendar cannot touch.

Why does the same business read 0.10 times and 0.69 times in the same year?

Begin with a household. The shape is easier to feel there. A household borrows Rs 2,00,000 for a wedding in November and clears it by the middle of March out of the year's savings. Ask that household on 31 March whether it carries any debt and the honest answer is none at all. Ask whether it carried any debt during the year and the honest answer is yes, for four months, quite a lot of it. Both answers are true, and only one of them would appear on anything written on 31 March.

Anjani Stationers runs the same shape at business scale. Anjani Stationers supplies school notebooks, so it buys paper and prints months before the session opens and then waits about 128 days for its customers to pay. The gap between paying for paper and being paid for notebooks is funded by a cash credit facilityBorrowing a business dips into and clears again as trading requires, within a ceiling the bank has sanctioned, instead of drawing one fixed sum for one fixed term. from its bank, drawn heavily through the production and despatch months and repaid as the collections arrive. On the twelve month-end balances it averaged Rs 26,40,000 drawn and peaked at Rs 45,00,000. On 31 March it was nil.

So the borrowings figure the balance sheet publishes is Rs 10,20,000, and it is entirely correct. But add the average facility drawing of Rs 26,40,000 to the average term loan of Rs 4,10,000 and the average lease liability of Rs 6,50,000 and average borrowings across the year come to about Rs 37,00,000. Against EBITDA of Rs 53,50,000 that is 0.69 times, where the year-end net debt reading was 0.10 times. Seven times apart, on the same business, in the same year, with the same audited accounts, purely because one figure was measured on a single date and the other describes twelve months. Nobody did anything wrong: the facility rose and fell with the printing season, and the accounts report the balance that genuinely stood on 31 March.

The worked instance, both bases side by side

Here is every division on screen. The year-end column is taken straight from the published balance sheet. The across-the-year column takes the average facility drawing of Rs 26,40,000, the term loan averaged between its opening Rs 4,00,000 and its published closing Rs 4,20,000, and the lease liability averaged between its opening Rs 7,00,000 and its published closing Rs 6,00,000. Nothing in the second column restates or replaces anything published.

Read the third ratio row before anything else. It answers the obvious objection. Average cash by date is not something a set of accounts publishes, so the 0.10 against 0.69 comparison puts a net figure beside a gross one. So the table also runs gross against gross: 0.19 times at the year end against 0.69 times across the year. The gap narrows from about seven times to about three and a half, and it does not go away. The cash deduction was never what caused it. The measurement date caused it, and no amount of care about the inclusion list will fix a problem that lives in the calendar.

Which measures survive the measurement date, and which do not?

Now put the three ratios on the same twelve months and watch them. Net debt to EBITDA runs from 0.10 times on 31 March to 0.94 times in May. Gearing runs from 6.7 per cent to 28.0 per cent. Interest cover reads 11.9 times in April, 11.9 times in May and 11.9 times on 31 March, and it would read 11.9 times if the accounts were made up on any other day of the year.

The reason is not that interest cover is a better designed ratio. The reason is that both of its inputs are flows. EBIT of Rs 41,50,000 is what twelve months of trading produced. The finance cost of Rs 3,50,000 is what twelve months of borrowing cost, and critically it includes every rupee of interest on the seasonal facility during the months it was drawn, whether or not the facility was still drawn on the last day. There is no date inside the ratio for a date to move.

A measure built from two flows cannot be moved by the reporting date, and interest cover is the only one of the three built that way. Quote it alongside net debt to EBITDA rather than instead of it. Net debt to EBITDA asks how big the obligation was at a moment; interest cover asks whether the year's trading covered the year's cost of borrowing. A reader who has only the first has a photograph, and a reader who has both has the photograph and a description of the year it was taken in.

What does the finance cost line say that the balance sheet did not?

A cross-check worth carrying into every set of accounts takes about twenty seconds: the finance cost divided by the borrowings on the balance sheet. On Anjani Stationers that is Rs 3,50,000 over Rs 10,20,000, an implied rate above 34 per cent. Nobody is paying that on a secured term loan and a warehouse lease. The arithmetic is not wrong; the denominator is simply not the balance the interest was charged on.

The same division on average borrowings of about Rs 37,00,000 gives about 9.5 per cent, an ordinary blended figure for a business of this shape. An implied rate that looks absurd is not usually evidence of an absurd rate, it is evidence that the balance sheet figure is not the balance the business actually carried through the year. The finance cost line is a flow, so it quietly reports the whole year even when the borrowings line reports only one day of it, and holding the two against each other is how a reader spots the gap without needing the facility papers at all.

Anjani Stationers' own split confirms it. Of the Rs 3,50,000 charged, Rs 2,64,000 arose on the seasonal facility. The term loan accounted for Rs 41,000 and the warehouse lease for Rs 45,000. Look at the first line: three quarters of the year's interest cost was paid on an instrument that appears on the balance sheet at nil. The rates are Anjani Stationers' own contracted rates, and a different borrower would carry different ones.

One more reconciliation shows the same point from the cash flow statement. Financing activities were an outflow of Rs 4,30,000 in year two. Take out Rs 3,50,000 of interest paid and Rs 1,00,000 of lease principal and what is left is plus Rs 20,000 of net new borrowing. The Rs 20,000 is exactly the movement in the term loan from Rs 4,00,000 to Rs 4,20,000. The facility drew and repaid many times inside that year and contributed nothing at all to the closing figure. No line of the published statements therefore shows how large it got.

What would a lender ask for that no balance sheet carries?

A credit officer reviewing a working capital limit does not begin with the balance sheet. She begins with the bank's own record of the account. The people who use these numbers for a living settled the date problem long ago by asking for three figures no published statement contains.

The first figure is the maximum amount drawn at any point in the year, the size of the exposure the lender actually carried. The second is the average drawn, the ordinary running position rather than either extreme. The third is the facility limitThe ceiling a lender has agreed a business may draw up to on a facility. The drawn amount sits below the ceiling. The unused part is available but not borrowed. and how much of it was used. A business running at ninety per cent of its limit through the season has far less room than one running at forty per cent.

An analyst outside the bank cannot demand any of those three, but can usually get close, by reading the borrowings note for the facility limits, by dividing the finance cost by the borrowings to test whether the implied rate makes sense, and by asking the business directly at a results call. An investor reading published accounts alone should at minimum stop treating a single year-end borrowings figure as a description of the year, and should hold interest cover beside it precisely because interest cover already contains the months the balance sheet left out.

Where does each of these numbers sit in a set of accounts?

The calculator above carries a location under every field it asks for. Three of them are the ones readers most often go looking for in the wrong statement, so they bear repeating.

EBIT is not a caption in most statutory formats: it is read as profit before tax plus finance cost, or taken from an operating profit line where one is presented. EBITDA appears nowhere at all and is built by adding the depreciation and amortisation line back to EBIT. And the cash figure the balance sheet gives has to agree with the closing balance of the cash flow statement. Agreement there is the quickest check that the right line has been picked up.

The maximum drawn during the year, the average drawn and the facility limit appear in the borrowings note if the business chooses to give them, in the sanction papers available to the lender, and nowhere else. Guarantees given on behalf of another business appear in the contingent liabilities note and on no line of the balance sheet. The finance cost split by instrument, where a business gives it, is in the finance cost note.

Anjani Stationers Private Limited, Chitra Binding Works Private Limited and Vaidehi Rao are invented.
Educational material. Not advice on any investment, tax, budget or market position.