Debt Financing: What It Costs and What It Constrains

Sankalp Industrial Systems Limited pays a blended 8.00 per cent on Rs 6,00,00,00,000 of borrowing. The bill is Rs 48,00,00,000 of interest, and the rate falls to 6.00 per cent once the deduction saves it Rs 12,00,00,000 of tax. The constraints cost more than the rate says: a claim served before shareholders, dates nobody can move, and named assets already pledged.

Underneath that answer sits one idea, and the rest of this guide is its consequences. A lender is not buying a share of how the business turns out. A lender is buying certainty, and pays a lower price for it. The claim gets served first, the amount is fixed in advance, a date is attached, and on two of these three borrowings named assets stand behind it. Four certainties, and the rate is low because the lender holds all four. Each of those four features is a restriction on the borrower, and the low rate is the borrower's payment for accepting them. So what borrowing costs and what borrowing restricts are not two subjects sitting next to each other. Cost and restriction are the two faces of one price, and a reader who takes the first without the second has taken half a bargain and called it the whole.

What is a lender actually buying, and why is certainty cheaper than participation?

The balance sheet is the wrong place to start. Suppose a cousin needs Rs 50,000/- to fit out a tea stall and comes to a relative with two offers. In the first the relative hands over the money and she pays back Rs 54,000/- in a year, whatever happens to the stall. In the second the relative hands over the same money and takes a quarter of whatever the stall earns, for as long as it stands. The two offers would be priced very differently, and the first one lower. The first offer gives a number and a date. The second offer gives a share of an outcome nobody can see yet.

The lower price on the first offer is not generosity, and it is not a mistake: it is what a lender charges when somebody else has agreed to carry the uncertainty. Everything a lender does to a company is a version of that first offer written down at length. The amount is fixed. The date is fixed. The claim is served before anything reaches the people who put in the risk money. And if the borrower will hand over a claim on specific assets, the lender will drop the price again. The last part of the uncertainty has come off the lender as well.

Run the year on this company and the ordering shows up in rupees rather than in principle. Sankalp Industrial Systems Limited, an invented manufacturer, earned Rs 2,40,00,00,000 of operating profit in Year 0. Interest of Rs 48,00,00,000 came out first. Tax of Rs 48,00,00,000 came out next, computed on what was left. Rs 1,44,00,00,000 remained for the owners of the shares. Nobody chose that sequence on the day; the loan agreement chose it years earlier and the tax law chose the rest.

The picture settles no argument about whether the lenders took too much or the shareholders too little. The sequence is what it fixes, and the sequence is not up for discussion in the year it runs. A shareholder holds whatever is left over; a lender holds a number. The difference between holding a number and holding a remainder is the difference in what each one charges, and every figure below is a consequence of it.

What does the borrowing cost before tax, and how do three different rates become one?

Sankalp Industrial Systems Limited borrows on three separate lines, and the three carry different rates for reasons that will matter later. Rather than describe them in a paragraph, here they are with everything that makes each one different sitting in its own column.

Three rates, one company. A single figure comes out of pairing every rate with the amount standing at it, and that pairing is the whole trick. Line 1 carries Rs 3,00,00,00,000 at 7.80 per cent, giving Rs 23,40,00,000 of interest. Line 2 carries Rs 2,00,00,00,000 at 8.50 per cent, giving Rs 17,00,00,000. Line 3 carries Rs 1,00,00,00,000 at 7.60 per cent, giving Rs 7,60,00,000. The three bills add to Rs 48,00,00,000. As a percentage of the Rs 6,00,00,00,000 borrowed that is 8.00 to the decimal. The blended rate is not the average of the three rates; it is the total bill divided by the total borrowing, and on this company the two answers are 3.3 basis points apart.

Averaging 7.80, 8.50 and 7.60 without weighting gives 7.9667 per cent, and that would put the bill at Rs 47,80,00,000. The gap is small here because the amounts are not wildly unequal. The gap gets large the moment one line dwarfs the others, and a plain average is most tempting in exactly that case.

There is a second reason the three lines differ. Line 1 is charged quarterly, line 2 half-yearly and line 3 monthly, and money charged more often costs more than the same quoted rate charged once a year. On these three the effective annual rateWhat a quoted rate actually comes to across a full year once the charging happens more often than once a year, so the earlier charges themselves start attracting interest. works out at 8.0311, 8.6806 and 7.8704 per cent respectively, against quotes of 7.80, 8.50 and 7.60. The rest of the case is built on the quoted 8.00 per cent, so that is the blend used throughout. How a quoted rate becomes an effective one is covered separately.

Why is the after-tax cost lower, and by exactly how much?

Interest comes off profit first, and only what survives that subtraction is taxed. A rupee of interest therefore shrinks the taxable amount by a rupee, and the bill falls by whatever share of a rupee the company hands over in tax. On Sankalp Industrial Systems Limited that share is 25.0 per cent, an assumed effective rate that every figure below carries through.

Do it in rupees before doing it in per cent. The rupee version is the one that stays in the head. The company writes cheques for Rs 48,00,00,000 of interest. A quarter of Rs 48,00,00,000 is Rs 12,00,00,000, so the tax bill comes in Rs 12,00,00,000 under what the same year would have produced with nothing borrowed. A second route reaches the same figure and is worth walking: with no borrowing, tax at 25.0 per cent on the full Rs 2,40,00,00,000 of operating profit would be Rs 60,00,00,000; with the borrowing, tax on Rs 1,92,00,00,000 of profit before tax is Rs 48,00,00,000; and Rs 60,00,00,000 less Rs 48,00,00,000 is Rs 12,00,00,000. The borrowing therefore costs the company Rs 36,00,00,000 for the year. Set that against the Rs 6,00,00,00,000 borrowed and the rate lands on exactly 6.00 per cent.

The per cent version is one multiplication. Take 8.00 per cent and keep three quarters of it. A quarter comes back through the tax bill. 8.00 times 0.75 is 6.00. Both routes land on the same number, and being able to check a rate against a rupee figure is what stops the after-tax cost of borrowing turning into a formula somebody applies without looking.

What happens to that tax saving in a year when there is no profit?

The condition most treatments leave out is exactly this one, and leaving it out changes the shape of the whole idea. The deduction is worth Rs 12,00,00,000 only because there is Rs 1,92,00,00,000 of profit before tax standing behind it for the interest to be taken off. Take the profit away and the deduction has nothing to work on.

A coupon makes this obvious in a way a tax rule never does. A shop offers 25 paise back on every rupee spent. The coupon is worth a great deal to a customer who was going to spend Rs 4,000/- anyway and nothing whatsoever on a purchase that cannot be made. The coupon did not change; the ability to use it did.

Push Sankalp Industrial Systems Limited down and watch the saving move. Halve operating profit to Rs 1,20,00,00,000 and nothing has gone wrong with the deduction yet: profit before tax is Rs 72,00,00,000, the interest is still fully absorbed, and the saving is still the whole Rs 12,00,00,000. The cushion has changed: operating profit is now 2.50 times the interest bill rather than 5.00 times. Push further, to operating profit of Rs 48,00,00,000, and profit before tax is nil. The interest still has to be paid in cash. The saving has gone to zero.

Read the size of that fall, the honest way to hold this. Operating profit has to drop by 80.0 per cent from Rs 2,40,00,00,000 before the saving disappears entirely, and it shrinks gradually on the way down rather than switching off at the end. So the deduction is robust on this company and conditional everywhere. The direction of travel is what survives: the deduction shrinks fastest in the very year the borrowing is hardest to carry. A benefit is not normally described that way round.

Does the cheap rate stay cheap as the company borrows more?

Here is where the sentence about borrowing being the cheaper source starts to come apart. The 8.00 per cent is not a property of borrowing. The rate is the price this borrower is charged at this level of borrowing, and the company keeps its own schedule of what it believes it would be charged at other levels.

The right-hand column is the quoted rate multiplied by 0.75 and then rounded to two decimals, so 7.75 per cent becomes 5.8125 and prints as 5.81. Now read the shape of the left-hand column rather than any single row of it. From 10 per cent to 40 per cent the quoted rate moves 1.25 percentage points. From 40 per cent to 60 per cent it moves 4.00. Three quarters of the entire climb sits above the 40 per cent row, and below about a 20 per cent share the schedule is so flat that it carries no useful information at all. It even dips: 7.75 per cent at a 10 per cent share is a quarter of a point below the 8.00 per cent at 25 per cent, so a reader who states the general rule that borrowing gets dearer as a company borrows more is contradicted by this company's own opening rows. Take the 10 per cent row as the starting point instead, and the climb from there on is unbroken. The climb above a 10 per cent share is the part of the schedule worth trusting.

Take the far end seriously for a moment. At a 60 per cent borrowing share the quoted rate is 13.00 per cent, and after the assumed deduction that is 9.75 per cent. Against the 14.00 per cent the shares cost at the company's current mix, the difference has fallen from 8.00 percentage points to 4.25. Two things have to be said in the same breath as that number. First, the whole outstanding balance reprices at renewal, not merely the extra borrowing, so the higher rate is charged on money already drawn. Second, the 14.00 per cent it is being measured against is the cost of equityThe yearly return shareholders look for in exchange for holding the leftover claim on a business, written as a percentage of what their stake is worth. at today's mix, and shareholders do not stand still while borrowing triples either. Holding that side fixed makes the narrowing look smaller than it is, and where the equity side of the schedule is worked out is settled separately.

What does the lender get besides interest, and which assets did it take?

A gold loan makes the idea physical. A chain is handed over, money comes back, and the rate is lower than it would be on the same money with nothing handed over. The chain has not been sold. The chain has been placed where the lender can reach it if the repayments stop, and that reach is what bought the lower rate.

Two of the three lines at Sankalp Industrial Systems Limited work the same way. Line 1 is secured and line 3 is secured, so Rs 4,00,00,00,000 of the Rs 6,00,00,00,000 has named assets standing behind it, or 66.67 per cent of the borrowing. Line 2, the listed debentures, has nothing named against it. Its rate is the highest of the three at 8.50 per cent even though it is not the largest borrowing, and the missing security is one reason.

Read which assets line 3 took: the receivables and the inventory, the assets the business turns over in order to trade at all. A claim over a spare plot of land is one kind of position for a lender to hold. A claim over the money customers owe and the stock waiting to be sold is a different kind. The borrower cannot go on operating without both. Security of that kind is very common on a facility of this sort. Reading the line off the schedule shows what the lender would be reaching for and how quickly it could get there.

Which of the repayment dates can the borrower move, and which cannot?

None of them, is the short answer, and that is the point. Rs 3,00,00,00,000 comes back as a single payment at the end of Year 5. Rs 2,00,00,00,000 comes back at the end of Year 7, a date the company's own five year forecast never reaches, and that placement is deliberate. Rs 1,00,00,00,000 sits on a line that has to be renewed every twelve months. Three different clocks, none of them running on the borrower's convenience.

A date is a constraint that costs nothing at all until the day it arrives. Costing nothing until then is exactly why it gets left out of the rate. The single payment at the end of Year 5 is the sharpest of the three, because a repayment spread over five years lets a company pay it out of five years of cash while a single payment has to be met out of one year of cash or out of somebody's willingness to lend again. Handling a payment of that shape, and arranging the replacement, has a treatment of its own elsewhere.

What promises does this record disclose, and what follows from finding none?

A lender ordinarily wants more than interest and security. A lender wants promises: undertakings to hold some ratio on the right side of a line the lender picked, to leave the pledged assets unpledged elsewhere, to hand over figures on time, and to ask first before doing certain things. Promises are the third thing the lender is buying, and on many borrowings they bite harder than the rate ever does.

The record discloses none of them, for any of the three lines. Silence in the record is a fact about the record and not a fact about the company, and the gap between those two readings is the whole of good practice. A reader who writes down that Sankalp Industrial Systems Limited has no promises attached to its borrowing has invented something. A reader who writes down that no promise is disclosed anywhere in what is available has said exactly what can be said.

Where would somebody look for the real thing? The loan agreement itself for line 1, and the debenture trust deed for line 2, are the documents that carry the terms. Charges registered against a company's assets are filed with the Ministry of Corporate Affairs and the register is reached from mca.gov.in. The Securities and Exchange Board of India sets what a listed borrower has to put into its own disclosures, and publishes the applicable requirements at sebi.gov.in. The limits, the tests and the thresholds themselves are set in those documents and in the current requirements.

How much of the firm's cash actually reaches shareholders?

A rate on its own says nothing about what a shareholder sees, so the interest now goes back into the cash flow. The starting point is what the business generates before anybody is paid. Free cash flow to the firmCash a business throws off in a year before a rupee of it has gone to anybody who lent money or anybody holding shares. in Year 1 is Rs 98,00,00,000. Take off Rs 36,00,00,000 of interest net of tax, the Rs 48,00,00,000 bill with the Rs 12,00,00,000 saving already removed. Add back the Rs 25,00,00,000 the company draws in new borrowing during the year. Money borrowed in a year is money available in that year, and Rs 87,00,00,000 is left.

Three lines, and there is no fourth to add. The line that gets dropped is the borrowing line. Drop it and what reaches shareholders looks like the firm figure less interest. The two are not the same, and the omission is a real error rather than a rounding.

Run the same three lines across all five forecast years and a pattern comes out that no single year shows.

The firm line steps up by Rs 18,00,00,000 a year and the shareholder line steps up by Rs 16,50,00,000, so the two never converge. Over five years the business generates Rs 6,70,00,00,000 and shareholders see Rs 6,00,00,00,000. The Rs 70,00,00,000 difference is exactly the Rs 1,95,00,00,000 of after-tax interest that went out less the Rs 1,25,00,00,000 of fresh borrowing that came in, and it checks both ways. The gap is exactly what the lender takes net of what the lender adds, and the adding is the half people forget.

Which raises the obvious question about the borrowing line. The borrowing line only adds while it keeps growing. Freeze it in Year 5 and that year's shareholder figure gives up the whole Rs 25,00,00,000, landing at Rs 1,28,00,00,000 instead of Rs 1,53,00,00,000, with nothing whatever having happened to the business.

Why is permanent capital sitting on a line that is renewed every year?

Look at where the Rs 25,00,00,000 of new borrowing comes from each year. The draw is on line 3, the working capital facility, and that facility has to be renewed every twelve months. Draw it five times and the facility stands at Rs 2,25,00,00,000 at the end of Year 5, having started at Rs 1,00,00,00,000.

Now look at what the money buys. Net new invested capitalMoney put into fixed assets and into working capital in a year over and above whatever wore out or was used up in the same year. runs at Rs 1,00,00,00,000 a year here, and the Rs 25,00,00,000 borrowed is a quarter of that. Net new invested capital here is machinery and stock, and machinery and stock stay in a business for years. Funding something that stays for years on a line somebody else renews every twelve months is a mismatch, and naming it is the whole of what this record supports.

Household version, and it is uncomfortably close. Suppose a refrigerator that will last a decade gets bought on a credit line the bank reviews every year. Nothing has gone wrong. The refrigerator works, the payments are made, the line is renewed. The mismatch is not a problem anyone can point at on any given day; it is a dependence on a decision that somebody else gets to make repeatedly. In a company the same shape has a second edge. A renewal is also an opportunity to reprice, and repricing reaches the whole outstanding balance rather than only the newest slice of it.

The story has no ending in this record. The record does not say whether the facility was ever renewed, refused or repriced, and no outcome hides in it for a careful reader to find. Naming the mismatch is the finding. Adding a comfortable ending or a frightening one puts a fact into circulation that nobody ever established.

How does a lender, an analyst and a shareholder each use these same numbers?

Three readers open the same schedule of figures and take three different things off it, and watching that happen is the fastest way to see why both halves of this guide matter.

A lender reads the cushion rather than the rate. Operating profit of Rs 2,40,00,00,000 against an interest bill of Rs 48,00,00,000 is 5.00 times over, and a lender's first move is to ask what that becomes under pressure. Halve operating profit and the cover falls to 2.50 times, still comfortable. Take it down to Rs 48,00,00,000 and it is 1.00 time, meaning every rupee the business earns before interest is spoken for. The lender is not pricing today's number; the lender is pricing how far today's number can fall before it stops being a number.

An analyst reads the 6.00 per cent and immediately asks what borrowing share it belongs to. That is the question the schedule answers and a single figure never can. An analyst building anything on a 6.00 per cent after-tax cost has to check that the amount of borrowing assumed in the model is the amount the 6.00 per cent was quoted at. Carrying a rate from one level of borrowing into a model that assumes another is a mistake that produces a perfectly tidy answer.

A shareholder reads the difference column. Rs 11,00,00,000 in Year 1 and Rs 17,00,00,000 in Year 5, growing every year, is the lender's take net of the lender's contribution, and it grows because the balance grows. A shareholder also notices that the incoming half of that netting depends on the facility continuing to grow. The dependence is on somebody else's annual decision rather than on the business.

And a household reads the same thing without any of the vocabulary. A fixed monthly payment does not care what kind of month the household had. A fixed payment indifferent to the month is the entire idea, and every figure in this guide is a version of it.

What does borrowing stop a company from doing, when nothing measures it?

The last constraint has no number attached and is probably the largest. A company with Rs 48,00,00,000 of contracted interest and Rs 3,00,00,00,000 falling due four years out makes different decisions from a company with neither. Such a company is slower to take on a project that pays back late, thinks harder about a downturn, and is more conscious of how a lender will read anything it announces. None of that shows up in any ratio, and all of it is real. The operating cycleMoney leaves to buy materials long before a customer settles the bill for what was made from them, and the wait in between is what this names. gets managed a little tighter, the marginal project gets deferred, and the effect is a change in behaviour that no line item records.

There is a serious argument that this is worth having rather than merely a cost. Jensen and Meckling put the case in 1976, in a Journal of Financial Economics paper called Theory of the Firm, that a manager with a fixed claim to serve has less room to spend cash badly, and that the discipline of having to produce the payment is itself worth something to the people who put up the risk money. Read that way, the same Rs 48,00,00,000 that one description calls a constraint is what another description calls a useful restraint.

Both readings are held by people who have thought about it carefully, and both are describing the identical contractual fact. Neither of the two readings rules out the other. Which one applies depends on the company, on what the cash would otherwise have been spent on, and on judgements that no figure in this record settles.

How the price of borrowing responds to the amount borrowed, and how borrowing more or less changes what the whole business is worth, are both covered separately.

Sankalp Industrial Systems Limited is invented.
Educational material. Not advice on any investment, tax, budget or market position.