Deal Terms: What Turns a Headline Into Money Paid

Deal terms are the clauses that decide what a headline price turns into: how the price is fixed, what counts as debt and as working capital, how much is held back and for how long, who carries a risk that surfaces later, and what has to be true on the day money moves. The number is agreed first and defined afterwards, and the definitions are where it changes.

The prefilled figures are this transaction as recorded. The eleven entries return an enterprise value of Rs 1,320 crore, an equity value of Rs 1,140 crore, an adjusted price of Rs 1,137 crore, and Rs 1,023.30 crore crossing on completion day once Rs 113.70 crore goes into escrow. With the earn-out earned the sellers reach Rs 1,197 crore. The headline stands Rs 296.70 crore, or 22.48 per cent, above what crosses on the day.

What do deal terms actually decide?

A seller agrees to sell a second-hand scooter for Rs 60,000/-, and the price is settled. Then the buyer notices the insurance lapses next week, so the seller carries the Rs 1,400/- renewal. The seller is keeping the spare battery, worth Rs 800/-, and that comes off too. The buyer has been caught before, so he will pay the last Rs 2,000/- only once the registration transfer comes through. Nothing about the scooter changed, and yet what left his account on the day was Rs 55,800/-, with Rs 2,000/- following weeks later.

A price is a headline, and the terms are the machinery that turns that headline into a specific amount reaching specific people on a specific day. Every part of that machinery below carries its own rupee figure.

Harivansh Packaging Limited is buying 100 per cent of Sundarban Polymers Private Limited, at 10.0 times the target's earnings before interest, tax, depreciation and amortisation (EBITDA) of Rs 132 crore, an enterprise value of Rs 1,320 crore. Rs 1,320 crore is the figure an announcement carries. The enterprise value is not the amount reaching the sellers, not the amount leaving the buyer's account on completion day, and not the maximum the buyer can end up paying, and each of those three amounts is produced by a clause rather than by the multiple.

The multiple is agreed early, in a term sheetA short document setting out the main commercial points before the long agreements are drafted, usually not binding on price., by the people whose names are on the transaction, in a meeting everyone remembers. The definitions are agreed weeks later, in schedules, by different people, usually late at night, in a document nobody presents to a board. The multiple and the definitions are agreed at different times by different people, and that is exactly why the two so often disagree.

How does a headline multiple become money in a bank account?

The price mechanismThe name practitioners give to this cluster of clauses when they are read together as one calculation. is a sequence of defined steps, four of them here, each a paragraph somebody wrote, argued over and signed. Run in order, they rebuild every number in an announcement from the outside.

1. Apply the multiple to an earnings figure10.0 times Sundarban Polymers Private Limited's EBITDA of Rs 132 crore gives an enterprise value of Rs 1,320 crore. The multiple is applied here, not built.
2. Deduct net debt to reach the equity valueRs 1,320 crore less the target's net debt of Rs 180 crore is Rs 1,140 crore. That figure, not the enterprise value, is what the sellers are paid for their shares.
3. Adjust for the position on completion dayWorking capital came in at Rs 108 crore against an agreed normal of Rs 96 crore, so the price rises by Rs 12 crore. Net debt came in at Rs 195 crore against the assumed Rs 180 crore, so it falls by Rs 15 crore. Rs 1,140 crore plus 12 less 15 is Rs 1,137 crore.
4. Hold part of it backAn escrow of 10.0 per cent of Rs 1,137 crore, being Rs 113.70 crore, stays behind for eighteen months. Rs 1,023.30 crore crosses on the day.

Every one of those four steps is a definition somebody drafted, and only the first is ever discussed in public. The multiple gets a paragraph in the announcement; the other three get a schedule, a mechanism clause and a separate agreement with a third party. By rupees moved that order is close to backwards, and the ranking below puts a figure on how backwards.

What counts as debt, and why is that argued line by line?

Almost every purchase of a private business is agreed on a debt-free cash-freeA pricing basis treating the buyer as taking the business without its borrowings and without its cash, with both settled at completion. basis: the business is priced, and the borrowings and the cash are settled separately. A debt-free cash-free basis sounds administrative. Settling those two columns is the most valuable argument per hour in a transaction.

The enterprise value was struck on EBITDA and does not move, so when the net debt definitionThe list, written into the agreement, of which items count as borrowings and which as cash for the price adjustment. A schedule of named items, not a concept. pulls one more item into the debt column, the equity value falls by exactly that amount. Every rupee argued into the net debt column comes straight out of what the sellers receive, and the value of the business has not changed at all. It is a transfer between the two sides of the table, disguised as an accounting question.

The obvious items take minutes: bank borrowings are debt, cash is cash. Then come the items that are debt-like without being borrowings, and each is a separate argument:

None of those has an answer arithmetic can supply. Each is settled by negotiation, and each settles a specific number of rupees. Ten items at Rs 1 crore each is Rs 10 crore off the price, or 0.88 per cent of the Rs 1,140 crore equity value. The percentage looks small written down, and Rs 10 crore is large to whoever gives it up.

Why is there a working capital target, and who decides what normal is?

A small sweet shop changes hands. The price assumed stocked shelves and nothing owed by the regular customers. The seller, knowing the date, runs the stock down for three weeks and collects every outstanding bill early. On handover day the shop is worth what the paper says and is unusable without a week of purchases. The buyer puts back what was taken out and pays twice for it.

The stripped shelves are the entire reason a working capital targetA normalised level of stock, receivables and payables, agreed in advance, against which the actual position on completion day is compared. exists. The agreement fixes a normal level and the actual level on the day is measured against it. Here the normal is Rs 96 crore against an actual Rs 108 crore, so the price rises by Rs 12 crore; net debt was assumed at Rs 180 crore against an actual Rs 195 crore, so it falls by Rs 15 crore. The two net to minus Rs 3 crore, and the price moves from Rs 1,140 crore to Rs 1,137 crore, a quarter of one per cent. A buyer who had checked only one of the two would have been wrong by Rs 12 crore or by Rs 15 crore rather than by Rs 3 crore, so the smallness of the net is exactly why both are computed and never a reason to skip either.

The whole difficulty sits inside the word normal. Normal has to be agreed in advance from a period both sides accept, and a business with a season does not have one number that describes it. Set it from a low month and the buyer pays for a position the business rarely holds; set it from a high month and the seller is charged for stock that was never really there. The months chosen are therefore worth more than they look.

Two mechanisms exist for the measurement. Completion accountsAccounts drawn up after the transfer date, used to settle the final price some weeks later. measure the position after the day and settle the difference afterwards, and this transaction used them. A locked boxThe balance sheet is fixed at an earlier dated set of accounts and the buyer carries the economics from that date, with no adjustment afterwards. fixes the position at an earlier balance sheet instead, and the buyer carries the economics from that date with nothing settled afterwards. Which of the two a transaction uses is settled under the transaction paperwork.

What do a holdback, an escrow and a retention buy?

A wedding hall costs Rs 2,00,000/-. The customer pays Rs 1,80,000/- before the event and the hall keeps Rs 20,000/- for a fortnight in case something is broken. Nobody expects a breakage. The retention exists because chasing a caterer for Rs 20,000/- three weeks after everyone has gone home is not really a remedy.

A holdbackPart of the price the buyer simply does not pay on the day, retained in the buyer's own hands until a stated date or event., an escrowMoney placed with an independent third party under an agreement setting out who gets it and on what conditions. and a retention are one idea wearing three names. Money that does not move on completion day, held so that a claim discovered later has something to be paid out of. The difference between them is who holds it: a holdback stays with the buyer, an escrow sits with an independent third party under its own agreement, and a retention is the general word for the amount itself.

On this transaction the escrow is 10.0 per cent of the adjusted Rs 1,137 crore, or Rs 113.70 crore, held for eighteen months. The buyer gets a fund it can actually reach, and a promise becomes money already sitting somewhere neutral. The sellers pay for that with the use of Rs 113.70 crore for a year and a half after selling the business. The percentage and the period are a negotiation about that cost.

What is an earn-out, and what does it do to the years after completion?

An earn-outPart of the price paid after completion, conditional on the acquired business reaching stated figures within a stated period. exists because the two sides disagree about the future and neither can prove its case. The seller says the business is about to step up; the buyer says show me. The earn-out pays for the step up only once it appears.

Here a further Rs 60 crore is payable if Sundarban Polymers Private Limited reaches EBITDA of Rs 145 crore in the first year after completion. The threshold sits Rs 13 crore above the Rs 132 crore it earned, or 9.8 per cent higher. The maximum price to the sellers is therefore Rs 1,197 crore, and on the actual net debt of Rs 195 crore at completion the maximum enterprise value is Rs 1,392 crore.

Rs 1,392 crore is 10.55 times the Rs 132 crore the business actually earned. The same Rs 1,392 crore is also 9.60 times the Rs 145 crore the payment is conditioned on. Both are correct, and they are answers to different questions. A note quoting one multiple for a transaction with an earn-out has not yet said which EBITDA it used, and until it does the multiple carries no information.

A second effect is about behaviour rather than arithmetic. For the length of the earn-out period the sellers are influencing a business whose measured result decides their own pay. A cost deferred past the measurement date lifts the figure, and so does a postponed hire. None of that is bad faith. Anybody with a bonus formula in front of them does the same. The measurement therefore has to be drafted harder than the amount. An earn-out with an undefined baseline is a dispute with a date on it.

The definitions worth arguing over are short and specific: which costs are included in the measured EBITDA, whether central charges from the buyer count against it, who may commit spending in the period, what happens if the buyer reorganises the business so the measured unit no longer exists, and on what accounting basis the figure is struck. The accounting basis decides whether earnings arriving through the combination, rather than from the target on its own, count towards the Rs 145 crore.

What do warranties and the indemnity actually allocate?

A warrantyA statement of fact about the business, given by the seller, which the buyer can claim against if it proves untrue. is a statement of fact about the business, and the seller stands behind it: the accounts are properly drawn, the machines are held free of any charge, there is no litigation other than what is disclosed. An indemnityA promise to reimburse the buyer, rupee for rupee, for a specific identified risk, usually one both sides already know about. is a promise to pay for a specific risk that both sides have already identified: the tax demand for a named year, the boundary dispute on a named plot.

A warranty allocates the unknown and an indemnity allocates the known, and swapping them puts a risk everybody had already found into the column meant for surprises. The two carry different limits. Warranty claims run into a package of caps, thresholds and time limits. The parties are not guessing about an indemnity, so it is usually written to pay from the first rupee.

Now the limits, all struck on the adjusted price of Rs 1,137 crore, and all of them arithmetic:

Work three claims through it. A claim of Rs 0.8 crore is below the de minimis of Rs 1.14 crore, so it is never counted, and it does not even help other claims reach the basket. The counted total has not passed Rs 11.37 crore, so a claim of Rs 8 crore is counted and recovers nothing. A buyer with a real, provable Rs 8 crore loss recovers zero, and that is the package working as drafted rather than failing.

Then a claim of Rs 15 crore arrives, and two ordinary drafting conventions answer it differently. Under a tipping basket the whole Rs 15 crore is recoverable once the threshold is passed. Under a deductible basket only the Rs 3.63 crore above the Rs 11.37 crore threshold is paid, a difference of 75.8 per cent of the claim. A reader who knows the claim, the loss and the limits still cannot compute the recovery without reading which basket was drafted.

What has to be true before the money moves?

A purchase is agreed on one day and completed on another. In between, both sides have committed, nobody has paid, the seller still runs the business, and a list of things has to become true. The items on that list are the conditions to completionThe items written into the agreement that must be satisfied or waived before the transfer and the payment take place., and on this transaction there are three.

The first is a regulatory approval. The regulator and company law set which approval, which body and what has to be filed, not the parties, and those requirements change. The condition does one thing: until the approval is obtained, no money moves and no shares transfer.

The second is the absence of a material adverse change: the business the buyer completes on must still resemble the business it agreed to buy. The clause is short, heavily negotiated and almost never used. Material is defined narrowly enough that ordinary bad news does not reach it.

The third is consent from two counterparties whose contracts change hands. A supply agreement or a lease may end, or need permission to continue, when control changes. If those counterparties matter to the earnings the buyer priced, their consent is worth part of the price.

On this transaction twenty two weeks ran from term sheet to completion, of which the conditions period was nine. A conditions period is only as short as the slowest approval inside it, so the twenty two weeks here belong to this transaction alone. What matters is the shape: signing and completion are different days, and two further clocks run well past the day everybody calls the end.

Which clauses move the most money per word?

Put every clause worked above on one base, the Rs 1,140 crore equity value, and rank them by rupees. One per cent of that base is Rs 11.40 crore, and that conversion makes any percentage in the papers readable in money. Half a turn of the multiple on EBITDA of Rs 132 crore is Rs 66 crore, and a quarter turn is Rs 33 crore.

The length of a clause and its value are unrelated, and the shortest definition in the papers is very often the most expensive one. A single word, tipping or deductible, moves Rs 11.37 crore on one claim, and the net debt schedule moves the price rupee for rupee with no ceiling on how many items go into it.

Read the last column rather than the third. A term settled in the open is visible, contested and defended. A term settled in a schedule is conceded by whoever is most tired at the time.

What does the buyer's balance sheet look like the day after completion?

The terms and the funding meet on the same morning. Harivansh Packaging Limited funds the purchase with Rs 140 crore of its own cash and Rs 1,000 crore of new borrowing at its own contracted 9.0 per cent. Borrowings move from Rs 740 crore to Rs 1,740 crore and cash goes to nil, so the acquirer's own net debt moves from Rs 600 crore to Rs 1,740 crore.

A lender asks for the leverage, and a covenant is written on the answer. There are exactly two honest answers, and they differ because they measure different sets of companies.

The standalone reading is the higher of the two, and most readers expect the opposite. The reason is worth holding onto: Sundarban Polymers arrives carrying Rs 180 crore of net debt against Rs 132 crore of its own EBITDA, so consolidating it brings in proportionally more earnings than debt and pulls the group ratio down rather than up.

A third pairing is not a reading at all. Put the standalone Rs 1,740 crore over the combined Rs 609 crore and the answer is 2.86 times. The numerator and the denominator describe different sets of companies, so 2.86 times must never be quoted: it sits 0.30 turns below the consolidated reading and 0.79 turns below the standalone one, and matches neither. A leverage ratio that does not say which basis it is struck on has left its most-read number undefined.

How does a transaction team actually use this?

Devyani Kulkarni, the chief financial officer of Harivansh Packaging Limited, reads the agreement as a list of amounts rather than as a legal document. Beside every defined term she writes what it is worth in rupees at the current draft, and beside every limit the percentage converted into money on the Rs 1,137 crore base. By the time the draft reaches the board, the net debt schedule and the multiple sit side by side.

Ashwin Rege, who leads the transaction team, uses the ranking to decide where to spend the two things a transaction is always short of: hours, and goodwill with the other side. Knowing the basket convention is worth Rs 11.37 crore and the de minimis Rs 1.14 crore per claim tells him which to trade away when something has to be conceded.

A lender reads the same paper differently. Its own definition of net debt sits in the facility agreement rather than the purchase agreement, and the two are not required to match. Where they differ, the buyer can satisfy one and breach the other on the same set of facts. The two documents are therefore read together rather than in sequence.

An analyst outside the transaction has the same tool and the harder job. Given an announcement carrying Rs 1,320 crore, the published net debt and any disclosed adjustment mechanism, the bridge rebuilds from outside, and a disclosed earn-out puts a range rather than a point on the price. Most of the money sits in what an announcement does not say, and anyone who can run the four steps above can read it there.

The sellers use it in the opposite direction. For them the escrow is a delay rather than a protection, and the earn-out is a period during which their pay depends on a measured figure. The measurement definition and the escrow release date decide when the sale is actually finished, so a seller who has understood the mechanism negotiates them at least as hard as the headline.

What the calculator cannot settle

Four things, and no arithmetic settles any of them.

The calculator cannot settle whether the purchase was a good idea. Merit depends on what the money would otherwise have done and on what the combined business goes on to achieve, so every figure above is knowable and the merit is not.

The calculator cannot settle what a normal cap, basket, escrow period or earn-out looks like. The percentages here are this transaction's own commercial terms, each one negotiated rather than taken from a standard.

The calculator cannot settle what any approval requires, how long any filing takes, or which regime applies. Approvals, filings and regimes are set by the regulator and by company law, and all three change.

The record never fixes the six items missing from that net debt schedule, so the calculator cannot settle what they should have been worth. The arithmetic is fixed: Rs 33 crore won on a quarter turn, and an average of Rs 5.5 crore across six items to give it back. A definition silent on an item cannot be priced, and a figure put on that silence would be a guess wearing the clothes of a reading.