Recapitalisation: Resetting the Balance Sheet

A recapitalisation changes the funding and leaves the business alone. Sankalp Industrial Systems Limited, invented, could borrow Rs 3,60,00,00,000 and buy back 4,00,00,000 shares at Rs 90.00, moving its debt share from 25.0 to 40.0 per cent. Interest cover falls from 5.00 to 2.78 times and earnings per share falls from Rs 6.9000 to Rs 6.8250. Not one rupee of operating profit moves.

Start with a household and a house. The awkward part of the idea lands faster on something a reader can stand in front of. A household buys a house for eighty lakh rupees. In the first version it pays fifty lakh from savings and borrows thirty lakh. Five years later, with the house worth what it is worth and the salary what it is, the household goes back to the bank and borrows another twenty lakh against the same house, then hands that twenty lakh straight back to itself, into its own savings account. Nothing about the house changed. Nobody repainted a wall. The tap that drips still drips.

The claim on the house changed. The bank now goes first for a larger monthly amount, written into a document, in good months and bad. The household goes second, and takes whatever is left after the bank has been paid. The household has more cash sitting in its own account and a bigger fixed obligation every month, and those two things arrived at exactly the same moment because they are the same transaction seen from two sides. A recapitalisation is that and nothing more: the funding is rearranged and the thing being funded is not touched at all.

What is a recapitalisation, and what changes when one happens?

A recapitalisationA change in the mix of debt and equity funding a business, with the business itself unchanged. is a deliberate change in how a company is funded, made without changing the business itself. Think of the balance sheet as two columns that happen to add to the same figure. The left-hand column is the business: the foundries, the machine tools, the inventory sitting on the shop floor, the receivables owed by customers, the order book. The right-hand column is who paid for all of it, and in what order they get paid back. A recapitalisation is an operation performed on the right-hand column only.

Leaving the business alone is what makes a recapitalisation the cleanest test in the whole of this subject. If moving the mix genuinely changes anything, a recapitalisation is where the change becomes visible. Nothing else moved to confuse the reading. There is no new product, no acquisition, no cost programme, no change in the market. Any difference in the numbers afterwards has exactly one possible cause.

A recapitalisation runs in two directions, and the two are exact mirrors of each other. A leveraged recapitalisationBorrowing and using the proceeds to buy shares back, which raises the debt share. borrows money and uses it to buy shares back. The debt share rises. A deleveraging recapitalisationIssuing shares and using the proceeds to repay borrowing, which lowers the debt share. issues shares and uses the proceeds to repay borrowing. The debt share falls. The arithmetic is identical in both directions with one sign reversed, and both directions are worked below.

Here is the company this guide uses throughout. Sankalp Industrial Systems Limited, invented, makes industrial valves, precision castings and the aftermarket parts and service that go with them. The company is listed, so its equity has an observed price. At the end of the last completed year it had 20,00,00,000 shares at Rs 90.00, giving a market capitalisation of Rs 18,00,00,00,000, and gross debt of Rs 6,00,00,00,000 blended across its three borrowings at exactly 8.00 per cent, giving an interest charge of Rs 48,00,00,000 for the year. Add those two and total capitalDebt plus equity, measured here at market values rather than at what the accounts say. at market is Rs 24,00,00,00,000, of which debt is exactly 25.0 per cent. Operating profit was Rs 2,40,00,00,000 and earnings before interest, tax, depreciation and amortisation (EBITDA) Rs 2,88,00,00,000, so interest cover was exactly 5.00 times. All of that is settled earlier in this subject and is restated here rather than rebuilt.

How does a company actually move its mix, step by step?

In five steps, and they run in one order. Choose the target mix. Work out the amount. Raise it. Return it. Then restate every ratio. All of them moved at the same moment. The middle three are an afternoon of arithmetic that any careful person can do. The first is genuinely hard and the last is the one that gets dropped.

Step 1, choosing the target mixThe debt share a company decides to move towards., is the hard part rather than the execution. Deciding to move to a particular debt share means forming a view about what the borrowing will cost at that level, what the shareholders will then require, and how much room the company wants against a bad year. Choosing the target is covered separately: that work builds the relationship between the debt share, the blended cost of funding and the value of the business, and arrives at a figure. The figure is taken here as given: a 40.0 per cent debt share, belonging to this one invented company and to the invented schedule of borrowing rates the company was given.

Two things travel with that figure everywhere it goes. The first is that the relationship it comes from is flat near its own bottom, so the precision of the point is far lower than the precision with which it can be printed. Quoting a debt share to one decimal place claims an accuracy the arithmetic does not have. The second is that the whole shape of that relationship rests on an invented schedule of borrowing rates belonging to one invented company. Change the schedule and the point moves. Anybody who reads that 40.0 per cent as a general answer for any company has misread a worked example.

How much would this company have to borrow to reach a 40.0 per cent debt share?

Step 2 turns a percentage into rupees, and it is the step where the commonest error in the whole build gets made. The target share is a share of total capital, not of equity. Total capital at market is Rs 24,00,00,00,000. Forty per cent of that is Rs 9,60,00,00,000 of debt. The company already carries Rs 6,00,00,00,000. So it borrows Rs 3,60,00,00,000, and not a rupee more.

The other side has to agree. If debt is to be Rs 9,60,00,00,000 out of Rs 24,00,00,00,000, equity at market must be Rs 14,40,00,00,000, down from Rs 18,00,00,00,000. Equity falls by Rs 3,60,00,00,000, the same figure. The two sides cross by the same amount and the total does not move at all. The matching pair is the arithmetic signature of a recapitalisation. If the two figures do not match, an error has been made somewhere, not something discovered.

The error worth naming is taking the target share of equity instead. Forty per cent of Rs 18,00,00,00,000 is Rs 7,20,00,00,000. Borrowing that much would take gross debt to Rs 13,20,00,00,000 and land the company at a 55.0 per cent debt share rather than a 40.0 per cent one. The slip is easy to make: the equity figure is the one sitting in front of the analyst at the start. The denominator of a debt share is always debt plus equity.

What does the lender do when the borrower gears up?

The lender reprices, and here is the part most first attempts get wrong: it reprices the whole balance and not only the new money. A lender is not selling three hundred and sixty crore of rupees at a price. A lender is taking a position in a company. The position it holds afterwards is riskier than the position it held before: the same operating profit now has to service half again as much borrowing. A lender prices the position it ends up with, not the increment it just added. Pricing the whole position is why a company that gears up often finds its existing facilities being renegotiated at the same moment as its new one, and why covenants are rewritten rather than merely extended.

On this invented company's invented schedule the pre-tax rate at a 40 per cent debt share is 9.00 per cent, against 8.00 per cent at a 25 per cent debt share. The 9.00 per cent runs on the whole Rs 9,60,00,00,000 and not on the new borrowing alone. Interest becomes Rs 86,40,00,000 for the year, against Rs 48,00,00,000 before.

A lot of hurried work uses the other convention without noticing what it leaves out. If the old Rs 6,00,00,00,000 stayed at 8.00 per cent and only the new Rs 3,60,00,00,000 were priced at 9.00 per cent, interest would be Rs 48,00,00,000 plus Rs 32,40,00,000, being Rs 80,40,00,000. The difference of Rs 6,00,00,000 a year is exactly the old balance repriced by one percentage point. RepricingA lender changing the rate on the whole outstanding balance rather than only on new money. the old balance is not a rounding. The repricing is a real annual cost that a step 2 calculation on its own never sees.

What happens to the share count, and at what price?

Step 4 returns the money to the holders of shares by buying shares back. A buybackA company purchasing its own shares, which reduces the number in issue. here is simply the mechanism by which the equity side is made smaller. Why a company might prefer a buyback to a dividend, and how a buyback works as a matter of policy, are covered separately in this subject.

The arithmetic is a division. Rs 3,60,00,00,000 returned, at the market price of Rs 90.00 a share, buys 4,00,00,000 shares. The count falls from 20,00,00,000 to 16,00,00,000, a reduction of exactly 20.0 per cent. Then the mix is checked. A step that does not check itself cannot be trusted: 16,00,00,000 shares at Rs 90.00 is Rs 14,40,00,00,000 of equity, plus Rs 9,60,00,00,000 of debt is Rs 24,00,00,00,000 again, and Rs 9,60,00,00,000 over Rs 24,00,00,00,000 is exactly 40.0 per cent. The mix arrived where step 1 said it would.

Two assumptions in that paragraph are doing a great deal of work and both are worth naming. The first is that the price stays at Rs 90.00 while four crore shares are bought. In reality a company buying a fifth of its own shares would move the price it was paying, and every figure downstream would move with it. The second is that total capital stays at Rs 24,00,00,00,000 throughout. Holding both still is what makes the illustration readable, and it is not what would happen.

Which ratios move, and do they all move the same way?

Step 5 is where the company finds out what it has done, and it is the step that gets dropped because by then the transaction feels finished. The transaction is not finished. Every ratio the company is measured on has just moved, at the same instant, and they did not all move in the same direction.

Interest coverOperating profit divided by the interest charge. The ratio says how many times over a year's operating profit would pay a year's interest. was Rs 2,40,00,00,000 over Rs 48,00,00,000, exactly 5.00 times. Interest cover becomes Rs 2,40,00,00,000 over Rs 86,40,00,000, being 2.78 times. Gross debt to EBITDA was Rs 6,00,00,00,000 over Rs 2,88,00,00,000, being 2.08 times, and becomes 3.33 times. Cash is untouched at Rs 1,20,00,00,000, so net debt goes from Rs 4,80,00,00,000 to Rs 8,40,00,00,000, and net debt to EBITDA from 1.67 to 2.92 times. Three measures of borrowing rose sharply, one measure of room for a bad year fell by nearly half, and the reported earnings figure barely moved at all. Reporting a recapitalisation on any single ratio therefore reports a fifth of it.

A ratio is easy to nod at, so interest cover deserves a second sentence. Restated as a distance it reads differently. At 5.00 times, operating profit could fall 80.0 per cent before the interest bill consumed all of it. At 2.78 times it could fall 64.0 per cent. The transaction took a large amount of room for a bad year and exchanged it for something else. Whether the something else is worth having is a separate question.

What does the whole move do to earnings per share?

As one column, from operating profit to the figure a shareholder reads: operating profit is Rs 2,40,00,00,000 in both cases. A recapitalisation cannot reach it. Interest goes from Rs 48,00,00,000 to Rs 86,40,00,000. Profit before tax therefore goes from Rs 1,92,00,00,000 to Rs 1,53,60,00,000. Tax at the company's own assumed effective rate of 25.0 per cent goes from Rs 48,00,00,000 to Rs 38,40,00,000. Profit after tax goes from Rs 1,44,00,00,000 to Rs 1,15,20,00,000. Out of that comes Rs 6,00,00,000 belonging to the minority in Sankalp Coatings Private Limited, an invented subsidiary the group consolidates in full but holds only three quarters of. The figure is held flat throughout: the subsidiary is untouched by the transaction.

So profit attributable to owners falls from Rs 1,38,00,00,000 to Rs 1,09,20,00,000, a fall of Rs 28,80,00,000. Divide each by its own share count and earnings per share goes from Rs 1,38,00,00,000 over 20,00,00,000 shares, being exactly Rs 6.9000, to Rs 1,09,20,00,000 over 16,00,00,000 shares, being Rs 6.8250. The fall is 1.09 per cent.

One check inside that column is worth doing once, and it confirms that the whole build hangs together. Interest rose by Rs 38,40,00,000. A quarter of that is deducted against tax at the assumed rate, so the after-tax cost of the extra interest is three quarters of it, being Rs 28,80,00,000. Rs 28,80,00,000 is exactly the fall in profit attributable to owners, to the rupee, and it has to be: the only thing that changed between the two columns is the interest line and the tax on it. If the two figures do not agree, something else has crept into the model that should not be there.

The deduction against the extra interest is real and worth stating plainly. Interest of Rs 86,40,00,000 at the assumed 25.0 per cent saves Rs 21,60,00,000 of tax, against Rs 12,00,00,000 before: a gain of Rs 9,60,00,000 a year. The saving is arithmetic on an assumed rate belonging to an invented company. Whether interest is deductible at all, whether there is a limit on how much of it is, and at what rate profit is taxed are matters of law.

Why must the market weights and the value case never be netted?

Two different totals are in play here, and they measure different things on different bases. Keeping them apart is the whole of what follows.

The first total is the one the transaction is sized on. Total capital at market is Rs 24,00,00,00,000, being the observed market capitalisation of Rs 18,00,00,00,000 plus gross debt of Rs 6,00,00,00,000. Both figures are market weightsWeights computed on the observed market value of equity and on debt, rather than on the values carried in the accounts.. The company is listed and its equity has a quoted price. A figure somebody actually paid is better evidence about today's equity value than any estimate, so the quoted price is the one used. Forty per cent of that observed total is what step 2 borrows against.

The second total is a modelled figure, and the move is judged against it. The modelled figure comes from taking the same unchanged stream of operating cash flows and discounting it at a blended rate that moves as the mix moves. No price enters it anywhere. The modelled figure is what a model says the operating business is worth on a stated set of assumptions, and moving the assumptions moves it. The discounting arithmetic is covered separately.

Neither figure is solved back into the other, and a reader who subtracts one from the other has stopped doing arithmetic. The market weights size the transaction. A modelled value tests the case for making it. The two answer different questions and they are allowed to disagree; the gap between a modelled value and a market price is itself covered separately. Mixing them is how a recapitalisation analysis quietly stops being arithmetic and starts being a story.

What does leverage take from what is left for owners?

The whole case about leverage can be made without a valuation model at all, and that is the cleanest way to make it. The funding side cannot reach operating profit, so it is held at Rs 2,40,00,00,000, unchanged at every debt share. The share left after the lender has been paid is what changes.

At the current 25.0 per cent debt share, debt is Rs 6,00,00,00,000 at 8.00 per cent, interest is Rs 48,00,00,000, and 80.00 per cent of operating profit is left for owners. Interest cover is exactly 5.00 times. At a 40.0 per cent debt share the whole balance reprices to 9.00 per cent, debt is Rs 9,60,00,00,000, interest is Rs 86,40,00,000, and 64.00 per cent is left. Cover is 2.78 times. At a 60.0 per cent debt share the schedule charges 13.00 per cent, debt is Rs 14,40,00,00,000, interest is Rs 1,87,20,00,000, and 22.00 per cent is left. Cover is 1.28 times.

The share left for owners falls 80.00 to 64.00 to 22.00 per cent and interest cover falls 5.00 to 2.78 to 1.28 times, on an operating profit that never moved by a single rupee. The entire cost of leverage is stated there without one cell of any valuation table. The same two forces are why the relationship the target comes from eventually turns downwards. The deduction against interest pulls the blended cost of funding down. The rising cost of the borrowing and the rising return the shareholders require push it back up, and somewhere between those two forces the blended figure stops falling. Where exactly is worked out separately, the relationship is flat near its bottom so the point is far less precise than it can be printed, and the whole shape depends on the invented rate schedule above.

Do value and earnings per share peak in the same place?

Value and earnings per share do not peak in the same place, and that is the finding that makes the whole exercise worth doing. Run the same five-step arithmetic at every point on the invented rate schedule, holding operating profit at Rs 2,40,00,00,000, holding total capital at Rs 24,00,00,00,000, moving equity in and out at Rs 90.00 a share, charging the schedule rate on the whole balance, taxing at the assumed 25.0 per cent and deducting the Rs 6,00,00,000 attributable to the minority. The construction returns exactly Rs 6.9000 at the company's actual 25 per cent debt share. Rs 6.9000 is the reported figure to the paisa, and that agreement is the check that the construction is the right one rather than an assertion.

Read across the ladder and it climbs, tops out and then collapses. With no borrowing at all it is Rs 6.5250. At a 10 per cent debt share Rs 6.6688, at 20 per cent Rs 6.8231, at 25 per cent Rs 6.9000, at 30 per cent Rs 6.9348, at 35 per cent Rs 6.9127, at 40 per cent Rs 6.8250, at 45 per cent Rs 6.4790, at 50 per cent Rs 5.7938 and at 60 per cent Rs 3.1500.

Earnings per share peaks at a 30 per cent debt share, lower than the debt share the value arithmetic points at, so the move with the better modelled case lowers the reported number. At 40 per cent, the target of this entire worked move, earnings per share is Rs 6.8250, below the Rs 6.9000 the company reports today. A board watching earnings per share and a board watching value would recapitalise to two different places, and neither of them would be making an arithmetic error. Both statements are arithmetic. Neither is a verdict.

Two honest qualifications travel with that finding. The first is that the peak is barely a peak. Rs 6.9348 at a 30 per cent debt share beats the Rs 6.9000 at 25 per cent by three and a half paise and the Rs 6.9127 at 35 per cent by about two. On a share priced at Rs 90.00 that is a rounding nobody would act on. The figure below therefore shades the top of the ladder rather than marking a point on it. The second is that the value side is flat near its own bottom too, so the gap between the two peaks is far less precise than two figures quoted to four decimal places suggest, and the whole shape of both rests on one invented rate schedule belonging to one invented company.

What happens if the same five steps run backwards?

Everything reverses and nothing about the method changes. The mechanism rather than the direction is what does the work. Take a company moving the other way: one that wants more room for a bad year, or one whose lenders have asked for it, or one that has just been through a difficult year and would rather owe less.

Run the five steps to a 10 per cent debt share. Step 1, the target is 10 per cent. Step 2, ten per cent of Rs 24,00,00,00,000 is Rs 2,40,00,00,000 of debt, so Rs 3,60,00,00,000 of borrowing is repaid. Step 3 becomes raising equity rather than debt: issuing Rs 3,60,00,00,000 of new shares at Rs 90.00 means 4,00,00,000 new shares, taking the count to 24,00,00,000. Step 4 returns the money to the lender instead of the shareholder. Step 5 restates everything: on the schedule the rate at a 10 per cent debt share is 7.75 per cent, interest falls to Rs 18,60,00,000, interest cover rises to 12.90 times, profit attributable to owners rises to Rs 1,60,05,00,000, and earnings per share falls to Rs 6.6688.

Notice that profit rose and earnings per share fell. The mirror of the move in the other direction is the clearest possible demonstration that earnings per share is a division rather than a measure of how well the business did. More profit spread over more shares can be less per share. The move is therefore dilutiveLowering earnings per share. even though it made the company safer against a bad year by every leverage measure there is. Lower risk of a bad year, lower reported number. Higher reported number, less room. Both directions cost something.

What does a recapitalisation not change, and why does that matter most?

The shortest list, and the most important one. Revenue of Rs 12,00,00,00,000. EBITDA of Rs 2,88,00,00,000. Earnings before interest and tax (EBIT) of Rs 2,40,00,00,000. Capital expenditure of Rs 1,34,80,00,000 in Year 1 and the Rs 18,00,00,000 movement in working capital that goes with it. Free cash flow to the firm of Rs 98,00,00,000 in Year 1 and Rs 1,16,00,00,000, Rs 1,34,00,00,000, Rs 1,52,00,00,000 and Rs 1,70,00,00,000 in the four years after it. The three divisions, the customers, the order book and the aftermarket service business.

Not one of those figures moves. Not one of them knows how the assets were paid for. A valve does not cost more to machine because the machine was bought with borrowed money. A customer does not order differently because the share count fell. The cash the business generates before anybody with a claim gets paid is a property of the business, and a recapitalisation happens entirely on the other side of the balance sheet.

Holding that list in mind is what stops the commonest misreading of all: crediting a recapitalisation with an operating improvement it could not possibly have produced. When a company announces a change in its mix and the reported earnings figure moves, the list above is the first thing to check, and it will show that the movement came from the funding every time.

So what travels out of this worked example, and what does not?

The method travels: five steps in that order, a target share taken of total capital and never of equity, a rate applied to the whole balance rather than only to new money, a share count that falls or rises in exact proportion to the money moved at whatever price it actually moves at, and a restatement of every ratio at the end rather than one of them. The method works at any scale, on any company, in either direction, and on a household with a housing loan.

Not one figure from this worked example travels with it. The 40.0 per cent, the Rs 3,60,00,00,000, the 4,00,00,000 shares, the 2.78 times, the Rs 6.8250, the peak at a 30 per cent debt share: all of them are arithmetic on one invented company's locked inputs and one invented schedule of borrowing rates. A different schedule gives a different ladder, a different peak and a different answer to every question asked here. Carrying any of those numbers away as though it were a fact about companies would be the most expensive possible mistake, precisely because the arithmetic behind it is sound and therefore looks transferable.