Completion Accounts vs Locked Box: Which Date Decides

Completion accounts and a locked box are the two ways a document fixes the economics of a sale. Completion accounts settle the price after completion from accounts prepared on that date, so the final number arrives late. A locked box takes an earlier balance sheet, prices from it, and then never revisits the price. Each obliges the paper to define quite different things.

What does this comparison rest on?

Both mechanisms were named already, in the walkthrough of the definitive agreement. Neither is being introduced here. Setting the two mechanisms against each other properly is the job that remains.

One transaction runs underneath the whole comparison. Harivansh Packaging Limited was buying every share in Sundarban Polymers Private Limited. The enterprise valueWhat the whole business is being valued at before separating what the lenders are entitled to from what the shareholders are entitled to. was Rs 1,320 crore, and after taking off the target's own net debtBorrowings less the cash sitting against them. A single figure for what the business owes the world beyond its trading obligations. of Rs 180 crore, the equity valueThe amount that actually reaches the people selling the shares, once the debt in the business has been accounted for. Never the same as the enterprise value. that reaches the sellers was Rs 1,140 crore. The Rs 1,140 crore equity value is the number both mechanisms are run against.

The Harivansh purchase used completion accounts. Two line items adjusted at completion. Working capitalThe money tied up in running the business day to day: stock on the floor and money customers still owe, less the money the business still owes its own suppliers. was pegged at Rs 96 crore and came in at Rs 108 crore. The gap lifted the price by Rs 12 crore. Net debt was assumed at Rs 180 crore and turned out to be Rs 195 crore. The net debt leg pulled the price back by Rs 15 crore. Net effect minus Rs 3 crore, and the sellers received Rs 1,137 crore. Every figure that follows is one of those, or is worked out from them.

One more thing carries over from taking the definitive agreement apart. A defined term carries only the meaning the paper gives it and carries no other. Both mechanisms are made almost entirely of definitions, so that rule matters more here than almost anywhere. A mechanism made of definitions fails in exactly one way: something the two sides both assumed was obvious turns out never to have been written down. Neither of these mechanisms is a calculation with a document wrapped round it; each one is a document with a calculation inside it, and the document is the part that decides.

What is each mechanism actually fixing, and on which date?

Everything else about the two mechanisms is downstream of one genuinely different thing, so start there.

Completion accounts value the business by what was actually there on the day the transaction completed; a locked box values it by an earlier balance sheet and then declines to look again. That is the whole of the difference. Two mechanisms, one question, two answers to it: which day's version of the business is the one being paid for?

Consider the sale of a second-hand car. Buyer and seller shake hands on a price on a Monday, and for reasons to do with paperwork the buyer actually takes the keys three weeks later. In those three weeks the seller keeps driving it, puts fuel in, puts another eight hundred kilometres on the clock, and one tyre gets worse. Is the price shaken hands on a price for the car as it was on Monday, or a price for the car as it is on handover day? Both answers are perfectly reasonable. The two answers are just different. If the two sides never decide which one they meant, they find out at handover, and handover is the worst possible moment to find out.

Which day's version of the business is being paid for is the entire question, and a transaction document has to answer it in writing. Completion accounts answer it with the handover day, and then have to build the machinery to work out what the car was worth on handover day. A locked box answers it with the Monday, and then has to build different machinery to make sure nobody strips the car in between. Neither answer is soft. Each one commits the paper to a different pile of work.

Notice what is not different. The business is the same business under either mechanism. The buyer is buying the same shares, the same customers, the same factory. Nothing about the target changes because of which mechanism the two sides picked, and that is precisely why it is so easy to treat the choice as procedural. The choice is not procedural. The choice moves money.

What does the paper have to define for completion accounts to work?

Completion accounts are a promise to do arithmetic later. The promise is a perfectly good one, but it is worth exactly as much as the instructions that come with it, and the instructions have to be written before anybody knows what the answer will be.

Six things have to be settled in the document, and the reader of a draft should be able to find all six.

The line items that adjust. On this transaction there were two, working capital and net debt, and that is a common enough shape without being a rule. The point is that the paper says which. An adjustment mechanism that adjusts for whatever seems fair is not a mechanism.

The meaning of each of those items. Working capital sounds like a settled idea until two accountants have to agree a figure for it on one particular Tuesday. Does a disputed customer balance count? Does stock that has been sitting for two years count at full value? Every one of those is a real choice, and every one of them is worth money.

The basis the accounts are drawn up on. The conventions that govern how anything is measured and reported are settled by the Institute of Chartered Accountants of India, not by the buyer's team. The document says which basis applies, and what happens where that basis leaves a choice open. The document does not restate the conventions.

Who prepares them and who reviews them. One side draws the accounts up and the other side looks at what they did. The sentence naming the preparer does more work than it appears to do. Which side prepares is taken up below.

How long each side has. A drafting period for the preparer, a review period for the reviewer. Without these the mechanism has no end, and a mechanism with no end is not a mechanism either.

The route out if the two sides disagree. And this last one is not optional. A mechanism that produces its number after the money has already moved needs a route for a disagreement, and a document without one has created an argument rather than a mechanism. Think about what the position looks like without it. The money is gone. The shares have moved. The buyer says the working capital figure should be Rs 101 crore, the seller says Rs 108 crore, and there is nothing in the paper that can make either of them stop saying it. The usual answer is expert determinationHanding one narrow disagreement to an agreed third party whose answer both sides have promised in advance to accept, so the argument has somewhere to end., where a named third party settles the specific point and both sides have agreed beforehand to live with the outcome. Whether the document uses that or something else is a drafting question. Whether it uses something is not.

What does the paper have to define instead for a locked box?

A locked box takes the opposite route. The price was settled on a balance sheet that already existed, and nothing that happens afterwards is read for price purposes, so there is no arithmetic after completion at all. A locked box looks simpler. A locked box is not simpler. A locked box is earlier.

Three things have to be settled, and all three have to be settled before anybody signs anything.

The locked box date. A specific date, named in the paper, and there is a balance sheet drawn up at that date. Everything about the price is derived from that one document, so it had better be a document both sides have actually looked at.

The balance sheet at that date. Not an estimate, not a management pack, an actual set of numbers both sides are prepared to be bound by. Under completion accounts the buyer's diligence is followed by a second look at completion. Under a locked box there is no second look, so the first look has to be the whole of it.

The definition of value leaving the business between that date and completion. The third one carries the real weight, and it is the reason a locked box is not the easy option it looks like from a distance. The shopkeeper analogy is exact here. A buyer agrees to take on a small shop, price fixed on the stock and the cash as they stood on the first of the month. Handover is six weeks away. In those six weeks the shopkeeper is still standing behind the counter, still taking money out of the till, still paying suppliers, still paying himself. Some of that is completely normal and was expected when the shop was priced. Some of it is not: a large payment to a relative, a car moved out of the business, a dividend declared the week before handover. The document has to draw a line between those two categories, in advance, in words, covering things nobody has thought of yet.

The second mechanism does not remove the work, it moves it: the work goes from preparing accounts afterwards to defining what may and may not leave the business beforehand. One mechanism does its hard thinking with a calculator after completion. The other does its hard thinking with a dictionary before signing. Both are hard. The two mechanisms are hard at different times and they use different people.

When does each side actually know the final number?

Here is where the difference stops being conceptual and starts showing up in somebody's job.

Under a locked box both sides know the final number at signing; under completion accounts nobody knows it on the day the money moves, and it settles only once the accounts have been prepared and agreed. An estimated amount changes hands at completion, and the true amount is worked out afterwards, and the difference is trued up. So for a stretch of time after the transaction has visibly happened, the price of the transaction is still an open question.

The Harivansh purchase took twenty two elapsed weeks between the term sheet and completion, and nine of those were the conditions periodThe weeks that run from putting names to the agreement until completion actually happens, spent working through whatever the agreement made completion depend on.. Twenty two weeks is this deal's own elapsed time and nothing more. So signingThe day the two sides put their names to the agreement. On most transactions signing is not the day anything moves. The gap between signing and completion is what matters. came at week thirteen and completion at week twenty two. Under a locked box the number is settled at week thirteen. Under completion accounts it is not settled at week twenty two either. How much later it settles is decided by the paper, and no general answer exists. A reader who wants to know how long the tail runs on a particular transaction has to ask.

Why should an analyst care about a number arriving late? Because a number that is not final is a number nobody can use. The buyer's finance team cannot close its accounting for the purchase on a price that may still move. The seller cannot tell anybody what the sale actually realised, cannot distribute the proceeds cleanly, and cannot close the file. A wedding caterer who sends the final bill three weeks after the wedding has not done anything wrong, but everybody who was at that wedding has been unable to settle up until it arrived. Multiplied by a few hundred crore, that is the position both sides are in.

And it is not only a reporting nuisance. There is an asymmetry hidden in there. Until the accounts are agreed, one side is holding money that may have to come back, and the other side is waiting for money that may never arrive. Money that may have to come back is a real financial position, sitting on two sets of books, created entirely by the choice of mechanism.

Who does the work afterwards, and who checks it?

Under a locked box the answer is nobody. Nothing is drawn up after completion because nothing after completion is read.

Under completion accounts a great deal happens, and most readers underestimate it. Somebody has to produce a set of accounts for one specific date, on an agreed basis, covering agreed line items, within an agreed period. Then the other side reviews it, within its own period. Then either they agree, or they do not and the route in the paper gets used.

Here is the part worth carrying away. Whoever prepares the accounts makes the first set of judgements, and the reviewing side then finds itself arguing against a completed document rather than building one. Those are not the same job and they are not equally comfortable.

Think about how it feels in the room. The preparer decides, item by item, where the line falls: this receivable is collectable, that one is provided against; this stock is good, that batch is written down; this accrual belongs in the period, that one does not. Each of those is defensible. Each of them is also a choice, and choices made by a party with a financial interest in the answer tend, on average, to fall the same way. Nobody has to be dishonest for that to happen.

The reviewer arrives afterwards holding a finished document. To move a number, the reviewer has to identify the specific judgement, explain why it was wrong, and get the preparer to accept it or take it to whatever route the paper named. Every one of those steps costs time and credibility, and the reviewer has to spend that on each item separately. Meanwhile the preparer only has to hold still.

The sentence naming who prepares is therefore not an administrative detail buried in a schedule. The sentence is a real allocation of advantage, and the paper allocates it explicitly. In a draft, that sentence is the one to find. The sentence is usually short and it is never neutral.

Who keeps the money the business makes in the gap?

Most readers skip the question of who keeps the gap, and it decides more money than any other part of the comparison.

Between the locked box date and completion, the business keeps trading. The business sells things. The business collects cash. The business generates profit, or it loses money. Somebody gets the benefit of that stretch, and the two mechanisms give it to different people.

Under a locked box the buyer takes the economics from the locked box date, so profit earned after that date is the buyer's even though the buyer did not hold the shares yet; under completion accounts the seller keeps the economics right up to completion. That is not a technicality and it is not a rounding item. The choice of mechanism decides who gets the money the business made in the gap.

The everyday version is a flat with a tenant in it. A buyer agrees to take it in March, with the price struck on the position in March, and actually gets the keys in June. Who gets the rent for April and May? If the answer is the buyer, then the March price was a price for the flat plus three months of its rent, and it should have been priced that way. If the answer is the seller, it was not. The flat is identical in both cases. The cheque should not be.

A locked box makes the buyer the economic beneficiary from the early date onwards while the seller continues to be the legal shareholder until completion. Economic benefit and legal ownership are pulled apart on purpose. Once that is seen, a lot of locked box drafting stops looking strange: the leakage definition exists because the seller is still in the chair, running a business whose upside now belongs to somebody else, and the paper has to say what the seller may still take out for itself.

Under completion accounts the two things stay together. The seller runs the business for its own account until completion, keeps what it earns, wears what it loses, and hands over a business whose position at that moment is then measured and paid for. Completion accounts are the tidier arrangement conceptually, and the one that needs the machinery afterwards.

On a live transaction the value of the gap is built from the target's own monthly numbers, and it is built before the two proposals are compared rather than after.

Which risks does each one move, and on to whom?

By now the honest way to read the comparison should be visible. Neither mechanism is safe and neither is dangerous. Each of them takes a specific exposure off the table and puts a different one on.

Completion accounts take away the risk of paying a price struck on a position that had already changed by the time the money moved. The exposure removed is a real one: a business can lose stock, build up receivables, draw down cash or take on borrowings in the weeks between signing and completion, and completion accounts price what is actually there.

Completion accounts put two things on the table in exchange. The first is that the final number is unknown on the day of completion. The second is that the accounts themselves are arguable. Somebody prepares them, somebody else reviews them, and both have money riding on the outcome.

A locked box takes both of those away. There is nothing to prepare, so there is nothing to argue about, and the number was settled at signing. A locked box puts a single, specific exposure on the table instead: value can leave the business between the two dates in a way the leakage definition did not catch. Not fraud, necessarily. Just something nobody thought to write down, in a definition that had to be written before anybody knew what would happen.

Neither mechanism removes risk and both of them relocate it. The comparison is about where the risk goes rather than about how much of it there is. Anyone who calls one of them the safer option has either not read the leakage definition or has not read the dispute route, and usually neither.

What happens when the same transaction is run both ways?

Now put the two mechanisms against one set of facts and let the arithmetic finish the argument.

Harivansh Packaging Limited agreed an equity value of Rs 1,140 crore for Sundarban Polymers Private Limited. Rs 1,140 crore is what the two sides negotiated, and it is the starting point under either mechanism.

Under completion accounts, the route this transaction actually used. The document set a working capital peg of Rs 96 crore, and completion day showed Rs 108 crore sitting there, which lifted the price by Rs 12 crore. Against an assumption of Rs 180 crore of net debt, completion day produced Rs 195 crore. The difference cut the price by Rs 15 crore. Net effect minus Rs 3 crore, and the sellers received Rs 1,137 crore.

Notice what Harivansh Packaging did not know on the day it signed. The buyer did not know the final price. The buyer knew it to within a range, and the range was not narrow: either leg on its own was worth Rs 12 crore or Rs 15 crore, and the two could have moved in the same direction as easily as against each other.

Under a locked box on the same facts. The price is Rs 1,140 crore. Rs 1,140 crore at signing, Rs 1,140 crore at completion, and neither the Rs 108 crore of working capital nor the Rs 195 crore of net debt sitting there on completion day changes it by a single rupee. The two completion figures still exist. Nobody reads them for price purposes.

So the same underlying business produces two different cheques, Rs 1,137 crore and Rs 1,140 crore, and the Rs 3 crore between them is not a mistake by either side but the price of choosing a date.

The Rs 3 crore invites exactly the wrong conclusion. The gap is only Rs 3 crore because the two legs happened to point in opposite directions, and on this same transaction a locked box would have differed from completion accounts by Rs 12 crore had only the working capital leg moved, or by Rs 15 crore had only the net debt leg moved. Nobody arranged that offset. The offset was luck, and the importance of a mechanism is not measured by a net figure that luck happened to produce.

Take four points between Rs 150 crore and Rs 240 crore of actual net debt, all worked from one expression. At Rs 150 crore of actual net debt, completion accounts pay Rs 1,182 crore. At Rs 180 crore, exactly the figure the transaction assumed, they pay Rs 1,152 crore. At the Rs 195 crore that actually turned up, they pay Rs 1,137 crore. At Rs 240 crore they pay Rs 1,092 crore. The locked box price is Rs 1,140 crore at every one of those four points, and at every point in between them.

How does a reader actually use this in the first ten minutes?

Most people who have to read one of these clauses are not drafting anything. A reader has been handed a document, or a summary of one, and somebody wants to know what it means before lunch. Four kinds of reader are not all looking for the same thing, and the ten minutes look different for each.

An associate summarising the paper for a committee reads for the date first. Which mechanism, and what date do the economics pass on. The date the economics pass on changes what the price means, and a summary that gives the price without giving the date has told the committee something that is technically true and practically useless. Then the dispute route, then the preparer, then the numbers.

An analyst modelling the buyer cares about when the number becomes real. Under a locked box the consideration is a known figure from signing and can go into a model as one. Under completion accounts it cannot, and the honest treatment is a range with a stated basis rather than a point estimate dressed up as one. The analyst also has to know whether any part of the consideration sits in an escrowMoney held by a third party under agreed instructions, so that neither side can simply take it while the thing it was set aside for is still open.. Held cash and paid cash behave differently in a forecast.

A lender to the buyer is doing something more specific. A lender has committed to fund a purchase price, and under completion accounts that price can move after drawdown. So a lender reads the adjustment mechanism as a question about the size of its own facility: how far can the number go, in which direction, and is there headroom for the worst case rather than the expected one. On this transaction the two legs offset to Rs 3 crore, but a lender sizing against Rs 3 crore rather than against Rs 15 crore has sized against an outcome, not against a mechanism.

A seller planning what to do with the proceeds reads for finality. A locked box says the figure now. Completion accounts say a figure now and the real one later. The seller cannot commit the last slice of the proceeds to anything until the accounts are agreed. A household that has sold a flat and is buying another one understands this instantly: money that has not finished arriving cannot be promised.

Across all four, the same discipline holds. Read the date the economics pass on, then read what happens if the two sides disagree, and only after those two read the price. It takes about ninety seconds and it is the difference between summarising a document and describing a number.

Harivansh Packaging Limited and Sundarban Polymers Private Limited are invented.
Educational material. Not advice on any investment, tax, budget or market position.