Leveraged Buyout: The Structure and the Return Arithmetic

A leveraged buyout is a purchase where most of the price is funded with borrowing placed on the company that was bought. Sthira Capital Partners, invented, buys Sankalp Industrial Systems Limited, invented, at 8.50 times earnings before interest, tax, depreciation and amortisation (EBITDA), being Rs 24,48,00,00,000, funded with Rs 13,00,00,00,000 of new borrowing and Rs 12,00,00,00,000 of its own equity out of Rs 27,20,00,00,000 of total sources.

Start with something plain enough to picture on a street. Four cousins want to buy a bus and run it on a fixed route. Between them they can raise Rs 6,00,000. The bus costs Rs 20,00,000. So they set up a small private company, the company borrows Rs 14,00,000 against the bus itself, and the company buys the bus. Every month the fares come in, the driver and the diesel are paid, and whatever is left goes to the lender until the loan is gone. If the route fails, the lender takes the bus. The lender does not take the cousins' houses. The cousins did not borrow anything. The company did.

The bus is the whole shape of a leveraged buyout, and a corporate buyout is the same arrangement with more zeros and better paperwork. Four decisions were actually made. The cousins decided what the bus was worth. The cousins decided how much of the price they were willing to fund with their own money. The cousins decided that the borrowing would sit on the bus rather than on themselves. And they accepted that from the first day, the fares belong to the lender before they belong to anybody else. Change any one of those four decisions and the transaction is a different one; keep all four and it is this one, whatever the size.

What makes a leveraged buyout different from any other purchase?

One structural feature, and it is not the amount of borrowing. Plenty of ordinary companies carry more borrowing than Sankalp Industrial Systems Limited will carry after this transaction. What makes a leveraged buyoutA purchase where most of the price is funded with borrowing placed on the company being bought. its own thing is where the borrowing is placed. The buyer does not borrow money and then hand it over. The company being bought borrows the money, on its own account, secured on its own assets, and the price is paid out of that.

The buyer in this worked example is Sthira Capital PartnersThe financial buyer in this invented example: the party that puts up the equity and controls the structure., invented, a financial buyer. Sthira Capital Partners is not a manufacturer looking for a factory. The firm has no other valve business to fold this one into and no cost savings to take out. Everything the firm can earn has to come from the company it is buying, from the borrowing it puts on that company, and from the price a later buyer pays for it. The constraint of having only the company itself to earn from is what makes a buyout so useful to learn on: there is nowhere for a vague benefit to hide.

Three things have to hold together for the arrangement to work at all, and it is worth naming them before any number appears. First, the borrowing sits on the target, so the target's own cash flows service it and the buyer's other holdings are not reached. Second, interest is deductible against profit, so a rupee of interest costs less than a rupee out of the shareholder's pocket. Third, the cash that used to fund the old shareholders' dividends now funds repayment instead. A structure like this is only available to a business whose cash flows are steady enough to carry a fixed obligation month after month, and every judgement about a buyout eventually reduces to whether that is true of the particular company being bought.

The 9.50 and 13.00 per cent below are rates these two lenders contracted for this structure, negotiated against this balance sheet, rather than a going rate for borrowing anywhere.

Where does the price come from, and what does 8.50 times produce?

The price starts with a multiple and a profit figure, and nothing more elaborate than that. The entry multipleThe multiple of earnings at which the business is bought, applied here to EBITDA. agreed here is 8.50 times, applied to Sankalp Industrial Systems Limited's Year 0 EBITDA of Rs 2,88,00,00,000. Multiply the two and the enterprise value is Rs 24,48,00,00,000. The single line of Rs 24,48,00,00,000 is the entry price of the whole business, before anybody has worked out who gets what.

Two things are worth noticing about that arithmetic. The first is that it is applied to the last completed year, not to a forecast. Year 0 is the year that has already happened and been reported, so both sides can argue about what it means but not about what it was. The second is that the multiple is the price this invented buyer and these invented sellers arrived at, rather than a figure derived or claimed to be right. How a price is negotiated, and how anybody forms a view about whether it is a sensible one, is covered separately.

For orientation, the same company traded at 7.78 times its Year 0 EBITDA before any of this began, giving a traded enterprise value of Rs 22,40,00,00,000. So 8.50 times is a higher multiple than the market was applying. The difference between those two figures is where the premium lives, and it is measured properly in a moment. An entry multiple is a price, not a valuation: it says what was paid, and it says nothing at all about the value of the business.

How does an enterprise value become a price per share?

By walking the bridge between the two, in reverse. An enterprise value is what the operating business is worth to everybody who has a claim on it, lenders included. A price per share is what one particular claimant, the ordinary shareholder, gets. The route between them is a fixed set of additions and deductions, and every line has a reason.

Start at the entry enterprise value of Rs 24,48,00,00,000. Take off the existing gross borrowing of Rs 6,00,00,00,000. The lenders have a prior claim and the buyer is going to have to settle it. Take off the minority interest of Rs 60,00,00,000. The Rs 60,00,00,000 is the quarter of Sankalp Coatings Private Limited, invented, that the group does not hold, and that quarter of the coatings unit's cash flow was inside the forecast but belongs to somebody else. Add back the cash of Rs 1,20,00,00,000. Cash sits on the balance sheet and is not inside EBITDA. Add the non-operating assets of Rs 1,00,00,00,000, being a surplus land parcel at Rs 45,00,00,000 and the 26.0 per cent holding in Aruna Tooling Private Limited, invented, at Rs 55,00,00,000. Neither produces a rupee of the EBITDA that was just multiplied.

The figure that comes out the other end is Rs 20,08,00,00,000, and that is the price for all the equity. Divide by the 20,00,00,000 shares in issue and the price per share is Rs 100.40. Notice that the price per share is not the multiple times anything: it is what survives after four separate claims have been settled against the value of the whole business. There is no shortcut from 8.50 times straight to a share price. Anybody who tries one gets a different answer every time.

What is the premium, and why must two of them always be printed?

Because the same money produces two different percentages depending on the base it is divided by, and quoting only one of them misleads. Sankalp Industrial Systems Limited's unaffected share price, meaning the price before anything about a transaction reached the market, was Rs 90.00. The buyout price is Rs 100.40. A gap of Rs 10.40 a share is a premium of 11.56 per cent on the share price.

Now measure the same transaction on the other base. The traded enterprise value was Rs 22,40,00,00,000 and the entry enterprise value is Rs 24,48,00,00,000. The gap between those two is a premium of 9.29 per cent. Both are correct. Both describe the identical transaction. And in rupees they are identical too: Rs 2,08,00,00,000 either way. Rs 10.40 more per share across 20,00,00,000 shares is Rs 2,08,00,00,000, and Rs 24,48,00,00,000 less Rs 22,40,00,00,000 is also Rs 2,08,00,00,000.

So why do the percentages differ? Because the extra money is being spread over a bigger base the second time. The borrowing and the minority interest transfer at their carrying amounts and take no premium at all: a lender owed Rs 6,00,00,00,000 gets Rs 6,00,00,00,000 and not a rupee more for the inconvenience of a change of control. The premium is paid entirely to the shareholders, but it is measured against a base that includes people who were not paid a premium. The enterprise value premium is therefore always the smaller of the two.

There is a tidy consequence hiding in that, and it is worth carrying away. For this company the relationship between the two premiums is fixed. Whatever price is offered, the enterprise value premium comes out at exactly 80.36 per cent of the share price premium. The ratio collapses to market capitalisation over traded enterprise value, being Rs 18,00,00,00,000 over Rs 22,40,00,00,000. Check it here: 9.2857 divided by 11.5556 is 0.803571. The ratio would hold at Rs 115.00 a share and at Rs 150.00 a share as well. But it is a property of this balance sheet, not a rule of thumb. A company with more borrowing would have a lower ratio and a company with none at all would have a ratio of one, so a reader comparing two premiums quoted on different bases needs each company's own ratio.

Where does Rs 27,20,00,00,000 come from, and where does it go?

Every rupee that has to be found is a use. Every rupee that is found is a source. A transaction that is a rupee short does not complete, so the two lists have to come to the same figure, not approximately but exactly. The mechanics of building such a statement are settled separately; what matters here is what is on each list and what each item is doing.

The five uses, numbered, are these. One, the purchase of the equity at Rs 20,08,00,00,000, the figure the bridge produced. Two, the repayment of the existing gross borrowing at Rs 6,00,00,00,000. The old lenders' agreements will not survive a change of control, and they are settled at their carrying amounts. Three, the purchase of the minority interest at Rs 60,00,00,000. The buyer wants the whole of Sankalp Coatings Private Limited and not three quarters of it. Four, financing feesAmounts paid to the lenders arranging the borrowing, usually a percentage of what they arrange. of Rs 32,00,00,000, paid to the lenders arranging the new borrowing. Five, advisory and other transaction fees of Rs 20,00,00,000. The five uses come to Rs 27,20,00,00,000.

The five sources, numbered, are these. One, a senior term loanBorrowing that ranks first, is secured, and is repaid before anything junior to it. of Rs 9,00,00,00,000 at 9.50 per cent a year. Two, subordinated notesBorrowing that ranks behind the senior lender and accepts a higher rate for standing there. of Rs 4,00,00,00,000 at 13.00 per cent a year. Three, the cash already sitting on the company's balance sheet, Rs 1,20,00,00,000. Four, the non-operating assets sold at their carrying value, Rs 1,00,00,00,000. Five, sponsor equity of Rs 12,00,00,00,000. The five sources also come to Rs 27,20,00,00,000.

Readers skate past sources three and four, and both repay a lot of attention. Rs 2,20,00,00,000 of the price is funded by the company that is being bought, out of its own cash and its own surplus assets. The buyer added those two lines to the bridge on the way in, valuing them, and is now taking them straight back out to help pay for the purchase. None of that is a trick and none of it is double counting: the cash and the land were worth what they were worth, the buyer paid for them, and the buyer is entitled to use them. But it does mean the sponsor's cheque is smaller than a reader who only looked at the price would expect.

A small identity falls out of the two lists, and it allows a figure to be checked without rebuilding anything. The uses add Rs 2,72,00,00,000 to the enterprise value, being the cash, the non-operating assets and the fees. The four sources other than the sponsor supply Rs 15,20,00,00,000. So sponsor equity always equals the entry enterprise value less a flat Rs 12,48,00,00,000. At this price: Rs 24,48,00,00,000 less Rs 12,48,00,00,000 is Rs 12,00,00,00,000, exactly the locked figure. None of the other four sources moves with the price, so the same wedge holds at a different entry multiple.

What does each of the two lenders demand in return?

Two lenders put in Rs 13,00,00,00,000 between them, against the same business and the same cash flows. The two lenders are lending into the same risk. The difference between them is where each stands if things go wrong, and how long each is prepared to wait. Both differences carry a price, and it is worth being precise about how much.

Tranche 1 is the senior term loan. Rs 9,00,00,00,000 at 9.50 per cent a year. The senior loan ranks first, it is secured on the business, and it amortises by a 100 per cent cash sweepA term requiring spare cash to be applied to repaying borrowing rather than kept in the business.: every rupee of cash the business generates after interest, tax, capital expenditure and working capital goes straight to repaying it, and nothing is retained. The company does not get to decide what to do with its own spare cash for as long as this loan is outstanding.

Tranche 2 is the subordinated notes. Rs 4,00,00,00,000 at 13.00 per cent a year, bulletRepayable in a single instalment at maturity, with no repayment along the way., with no amortisation and no sweep at all. The subordinated lender receives its coupon and nothing else until the very end, and stands behind tranche 1 in any loss. If the business is sold for less than Rs 9,00,00,00,000, tranche 2 receives nothing.

Both lenders face the same business, so the 350 basis points between 9.50 and 13.00 per cent is not buying anything about the business: it is buying position in a queue and it is buying patience. The 350 basis points are worth sitting with, and they are the cleanest illustration in corporate finance of what risk actually costs. Nothing about Sankalp Industrial Systems Limited's valves, customers or margins is different for the two lenders. Only the order in which they are paid.

Blend the two and the borrowing costs 10.58 per cent before tax. Check it: Rs 9,00,00,00,000 at 9.50 per cent is Rs 85,50,00,000, Rs 4,00,00,00,000 at 13.00 per cent is Rs 52,00,00,000, so the interest is Rs 1,37,50,00,000, and Rs 1,37,50,00,000 over Rs 13,00,00,00,000 is 10.58 per cent. The company was previously borrowing at a blended 8.00 per cent. Sankalp Industrial Systems Limited is now paying more per rupee, and borrowing more than twice as many rupees.

How much borrowing is that, against what the business earns?

Entry leverageBorrowing at completion divided by that year's earnings before interest, tax, depreciation and amortisation. is the standard way of stating the size of it: Rs 13,00,00,00,000 of borrowing over Rs 2,88,00,00,000 of Year 0 EBITDA, being 4.51 times. Before the transaction the same company carried net borrowing of Rs 4,80,00,00,000, being 1.67 times. So the size of the obligation, measured this way, has been multiplied by roughly two and seven tenths.

The same funding, set out as rows.

The share column adds to 100.01 per cent rather than 100.00. Each share is rounded to two decimals for display and five roundings do not have to cancel. The rupee column is the check and it totals Rs 27,20,00,00,000 on both sides exactly. No figure has been adjusted to make the printed percentages add up.

Now the part that matters more than the leverage figure, and that the leverage figure hides. A multiple of EBITDA states how big the obligation is; it states nothing whatever about whether the business can carry it. EBITDA sits above interest, above tax and above every rupee of capital expenditure, so a company can be at four times EBITDA and comfortable, or at four times EBITDA and in serious trouble, depending entirely on what sits underneath.

So look underneath. Before the transaction, Sankalp Industrial Systems Limited earned earnings before interest and tax (EBIT) of Rs 2,40,00,00,000 in Year 0 and paid interest of Rs 48,00,00,000. Cover was exactly 5.00 times. Put the new structure on the same unchanged Year 0 profit and the interest becomes Rs 1,37,50,00,000, so cover falls to 1.75 times. Nothing about the trading changed. The valves are the same valves and the customers are the same customers. The change is in how much of the profit is already committed before anybody in the business gets to make a decision with it.

Why does the borrowing sit at the target rather than at the buyer?

Come back to the four cousins and the bus. The answer is exactly the same at both scales. If the cousins had borrowed Rs 14,00,000 personally and then bought the bus with cash, the lender's claim would be against them. Bad route, empty bus, the lender comes for the house. Because the company borrowed instead, and the security is the bus, the lender's claim stops at the bus.

Sthira Capital Partners borrows nothing. Sankalp Industrial Systems Limited borrows Rs 13,00,00,00,000, secured on its own assets, and services it out of its own cash flows. In Year 1 the business is forecast to produce EBITDA of Rs 3,16,80,00,000 and must find Rs 1,37,50,00,000 of interest out of it before a rupee reaches anybody else. If it cannot, the lenders enforce against the business. The lenders do not reach past it.

Two consequences follow and they point in opposite directions. The placement is a genuine decision rather than a free lunch. The buyer's money at riskWhat the buyer can lose, which here is the equity cheque and not the price of the business. is the Rs 12,00,00,00,000 it put in, and not a rupee more, however badly the transaction goes. The first consequence is the attractive one. The second is that a business which was previously carrying Rs 48,00,00,000 of interest a year is now carrying Rs 1,37,50,00,000, and it did not ask for that and did not get anything for it. Every rupee of that extra obligation is a rupee of margin for error that the business no longer has.

Who carries what is worth being blunt about. The sponsor carries the risk of losing its cheque. The lenders carry the risk of the business failing to pay. The employees, suppliers and customers of the business carry the consequences of a company that now has less room to absorb a bad year, and they were not asked. None of that makes the structure wrong. The structure has distributional consequences, and that is a different and more useful thing to understand than a structure that is simply good or bad.

What did the Rs 52,00,00,000 of fees actually buy?

Nothing that appears in any valuation of the business, and that is not a criticism of the people who were paid. The fee line is a statement about where the money went. Use four is a financing fee of Rs 32,00,00,000, paid to the lenders for arranging Rs 13,00,00,00,000 of borrowing. Use five is Rs 20,00,00,000 of advisory and other transaction fees. Together Rs 52,00,00,000, funded out of sources like everything else.

Both amounts simply leave. The money is spent on arranging and on advising. Both are real work that had to be done, and neither leaves behind an asset that anybody would pay for. The valves are the same valves the day after as the day before. The customers are the same. The factory is the same. A valuation of the business the morning after completion would not show the fees anywhere. There is nothing there to value.

The fees are the only part of the Rs 27,20,00,00,000 of uses that buys nothing, and that makes them the sharpest line on the whole statement. Every other use bought something: use one bought the equity, use two settled a real claim, use three bought out a real minority holding. Uses four and five bought the transaction itself.

The same thing shows up in an ordinary household. Somebody buys a flat for Rs 60,00,000 and pays Rs 2,00,000 in stamp duty, registration and brokerage. Their outlay is Rs 62,00,000. Their flat is worth Rs 60,00,000. The buyers are not poorer in any moral sense and nobody cheated them, but a sale the next morning at the same price would leave them Rs 2,00,000 down. The buyout version of that fact follows below.

What is the sponsor's stake worth on the morning after completion?

Work it from the two figures already on the table and it takes one line. The business is worth Rs 24,48,00,00,000 on the entry assumptions. The company carries Rs 13,00,00,00,000 of borrowing. So the equity sitting underneath that borrowing is worth Rs 11,48,00,00,000. Rs 11,48,00,00,000 is what the sponsor's stake is worth on the day it completes.

The sponsor put in Rs 12,00,00,00,000.

The Rs 52,00,00,000 difference is exactly the fees: money that left and bought nothing, and it is the first thing the return has to earn back. Not approximately the fees. Exactly them, to the rupee, and the reason is arithmetic rather than coincidence. The sponsor's cheque funded the uses along with everybody else's money. Two of those uses converted into nothing. So the sponsor is behind by the amount that converted into nothing, and it is behind by that amount on day one, before a single valve has been sold under the new arrangement.

As a proportion, the structure starts 4.33 per cent behind. Rs 52,00,00,000 over Rs 12,00,00,00,000 is 4.3333 per cent. Everything the business goes on to do, every rupee of earnings growth and every rupee of borrowing repaid, first has to make up that 4.33 per cent before the sponsor is level with where it started. 4.33 per cent is a small number against the size of the transaction and a large number against a five year hold, and both of those things are true at once.

Who holds what on the day after completion?

Four parties, four different things, and only one of those things has a value that can move. Setting them out side by side is the fastest way to see what the structure has actually done.

The lenders hold Rs 13,00,00,00,000 of claims on the business, in two ranked layers: Rs 9,00,00,00,000 senior and Rs 4,00,00,00,000 subordinated. The lenders' claim is fixed. If the business does brilliantly they still get Rs 13,00,00,00,000 plus their coupons; if it does badly they get less. There is no version of the future in which they get more.

The sponsor holds equity of Rs 12,00,00,00,000 in a new holding company that sits above the business. Sponsor equity is the only claim in this structure whose value moves with how the business performs, in either direction. Rs 1,00,40,00,000 of that Rs 12,00,00,00,000, being 8.37 per cent, is held by the target's senior management, who put back half of what they received for their own shares rather than taking all of it in cash. The arithmetic of that arrangement, and what it is actually for, is covered separately.

The company's former shareholders hold Rs 20,08,00,00,000 in cash and no further claim on anything. Whatever happens next, good or bad, has nothing to do with them. And the former lenders hold nothing at all: their Rs 6,00,00,00,000 was repaid in full at its carrying amount as use two, and they received no premium. A change of control pays a premium to shareholders and never to lenders. The absence of a premium for lenders is also the reason the enterprise value premium came out smaller than the share price premium.

How does the buyer's own operating case differ from the company's forecast?

Exactly two, and the discipline of that is worth more than the changes themselves. Revenue and EBITDA are unchanged in every one of the five years, so nothing about this structure rests on an operating improvement the buyer has not been shown to make. Revenue still adds Rs 1,20,00,00,000 a year, EBITDA still sits at 24.0 per cent of revenue, and the margin still does not move.

The first change is capital expenditure, held flat at Rs 90,00,00,000 a year for the whole hold periodThe number of years the buyer expects to hold the business before selling it., against the company's own case of Rs 1,34,80,00,000 in Year 1 rising to Rs 1,54,00,00,000 in Year 5. The reason is specific and nameable: the sponsor defers the third valve line for the whole period. Over five years the company's own case spends Rs 7,22,00,00,000 and the sponsor's case spends Rs 4,50,00,00,000, a difference of Rs 2,72,00,00,000.

The second change is the movement in net working capital, held at Rs 12,00,00,000 a year against Rs 18,00,00,000. The sponsor targets a tighter cash conversion cycle. Over five years that is Rs 60,00,00,000 against Rs 90,00,00,000, a difference of Rs 30,00,00,000.

Add the two and the sponsor's case frees up Rs 3,02,00,00,000 more cash across the hold than the company's own case would have. Neither item touches the profit and loss account, so neither changes the tax charge; both are cash items below the EBITDA line, and that is exactly why a sponsor reaches for them. Deferring the third valve line carries an implication of its own: a business that spends Rs 2,72,00,00,000 less on capacity over five years is a slightly different business at the end of it, and whoever buys it next will form their own view about that.

The reason this matters for everything that follows is traceability. Because revenue and EBITDA are untouched, every rupee the structure eventually produces can be pointed at. The sponsor underwrote its own case, it named the two lines it was changing, and it changed nothing else.

What did this structure return, and what does that figure not establish?

Over a five year hold, exiting at the same 8.50 times, this invented structure produced a money multipleMoney returned divided by money put in, ignoring how long it took to get it back. of 2.40 times and an internal rate of return of 19.14 per cent. Both are outputs of the structure rather than inputs to it. Where the return came from, how it splits between earnings growth, repayment of borrowing, any change in the multiple and the fees, and how much the whole answer moves if the entry price moves, are all worked out separately.

Both figures come with a boundary attached, and a number that arrives without one does more harm than good.

A money multiple does not show that the transaction was well judged. Two figures computed over an invented five years on an invented company, using an entry price and an exit multiple both chosen by whoever built the example, cannot establish that. A structure returned 19.14 per cent over five years on an invented company, and a return of 19.14 per cent says nothing about whether the transaction was a sensible thing to do or whether the price paid was a sensible price.

The two figures do not give the range around the answer. Both the entry multiple and the exit multiple were assumptions. Change the exit multiple and the whole figure moves; change the entry multiple and it moves in the other direction. A single return figure with no range attached is the least honest way to present the arithmetic of a transaction, and the decomposition and the range are covered together, separately.

And they say nothing about the borrowing along the way. The senior loan sweeps every spare rupee, so the amount outstanding falls year by year and the interest bill falls with it. The smaller interest bill feeds back into how much cash is available to sweep next year. The circularity, and the year by year schedule that resolves it, are covered separately.

One thing about the arrangement does travel, though, and it is worth carrying: entry and exit are at the same 8.50 times on purpose. With the multiple pinned, no part of the return can come from the market simply deciding to pay more for the same business, so every rupee of it has to come from somewhere a reader can point at. Pinning the multiple is a modelling discipline rather than a prediction, and it is the reason this particular example is worth learning on.