Writing a Transaction Case Study: Say Which Price First

A transaction case study states which price it is describing before it describes anything else. One purchase can carry a headline price, an amount actually paid, a maximum permitted and several correct multiples, so an account that opens with one figure and no basis has already lost the reader. The structure runs price, bridge, adjustments, contingent portion, multiple, then what the figures cannot settle.

Which number will the reader actually keep?

Consider the last account of a purchase read all the way through. A week later, what is left of it? Almost certainly not the structure of the payment, not the conditions, not the adjustments. One number. Possibly the name of the buyer. The surviving number is the honest measure of what a written account transfers, and it is a much narrower channel than most writers assume they are working with.

The same thing happens in ordinary life. A neighbour sells their shop and the street settles within a day on a single figure. Nobody on the street asks whether that figure was before or after the loan on the premises was cleared, whether the stock was counted separately, or whether part of it arrives next year only if the new owner keeps the same suppliers. The figure travels; the qualifications stay behind. A case study is read by somebody who will remember exactly one number from it, and the writer, not the reader, decides which number that is.

The writer decides it by position. Not by repetition, not by bolding, not by putting it in a heading. The figure in the opening sentence arrives before the reader has any structure to file the later figures into, so it is the one that survives. A case studyA written account of one completed purchase, setting out what was paid and on what basis, so that a reader can follow the figures rather than take them on trust. that opens with a number and no basis has already spent its one transferable slot, and everything careful that follows is being written for a reader who has stopped needing it.

The opening requirement of the form is therefore not a stylistic preference. The which-price ruleThe requirement to say which of a purchase's several correct prices a sentence is using, stated at the point the figure appears rather than once at the top of the write-up. says that before a figure is written, the sentence names which of the purchase's several prices it is. Not somewhere in the write-up. At the figure. Putting the largest figure first is a choice with an effect, whether or not the writer intended one.

Harivansh Packaging Limited, an invented maker of packaging, acquired Sundarban Polymers Private Limited, and that single purchase offers a writer at least five defensible opening figures. Every one of them is arithmetically correct. Every one of them describes something different. The writer who has not decided between them in advance will reach for whichever one came to hand first. In practice that tends to be the largest. The largest is usually the one already quoted elsewhere.

Why does one purchase carry several correct prices?

Three lines are enough. Every one of them is set out in the underlying valuation material, and none of it needs rebuilding. Enterprise value covers the business including the borrowings that come with it. Equity value is what the sellers actually receive. The completion adjustments move that equity value by the difference between what the agreement assumed and what was there on the day. And a conditional amount may follow later, if a stated condition is met.

Consider buying a second-hand delivery van from a small operator. The advertisement says one figure. The amount actually handed over is that figure less the loan still sitting against the vehicle. The seller has to clear that loan out of the buyer's money. Then on the day, the fuel in the tank and the spare tyres in the back get counted and the price moves by a few thousand rupees either way. And there is an agreement that if the vehicle passes its next inspection without work, a little more falls due. Four numbers. All of them true. Nobody who has bought a van is confused by this, and yet the same structure written up in a finance note routinely produces a single figure with no label at all.

All of these figures are correct, and a write-up has to name which one it is using at every point where a figure appears, rather than declaring the convention once at the top and hoping it holds. Naming it once at the top does not work. A reader who quotes the write-up will quote one sentence rather than the whole of it, and that sentence has to survive being lifted out. The whole problem fits in one line: figures travel out of their context, and the only defence is to attach the context to the figure rather than to the document.

Here are the five figures this one purchase carries, each with what it names.

Look at the spread. From Rs 1,137 crore to Rs 1,392 crore is Rs 255 crore of difference between two entirely correct descriptions of the same purchase, and that is 22.4 per cent of the amount that actually changed hands. A reader who picks up the wrong one of those five has not made a small error. The reader has taken away a figure nearly a quarter adrift from the money that moved, and nothing they were told was untrue.

What structure does the write-up follow?

Six parts, in this order. The order is doing work: each part hands the next one a figure it needs, and the account reads as one continuous movement from what was agreed to what was paid. Nothing in the list below is a mechanism, and none of the steps explains why a number is what it is. Each is a placement instruction, and the check underneath is what a draft is read against.

1. State the price and the basis it was struck onEnterprise value Rs 1,320 crore, being 10.0 times the Rs 132 crore of earnings before interest, tax, depreciation and amortisation (EBITDA) for the last full year on the reported basis.
   Checking: does the first figure in the write-up carry the thing it was struck on, in the same sentence?
2. Run the bridge to what the sellers receivedLess net debt assumed of Rs 180 crore, giving an equity value of Rs 1,140 crore payable to the sellers.
   Checking: has the enterprise value figure been visibly separated from the amount payable?
3. Report both completion adjustments at their own sizePlus Rs 12 crore on working capital, minus Rs 15 crore on net debt, a net of minus Rs 3 crore.
   Checking: are both gross movements in the write-up, and not only the net of the two?
4. Set out the contingent portion with its conditionRs 60 crore payable if EBITDA reaches Rs 145 crore in the first year after completion.
   Checking: does the condition sit in the same sentence as the amount it attaches to?
5. Give the multiple on every denominator that applies10.09 times at completion, 10.55 times on what was earned, 9.60 times on what the payment turns on.
   Checking: does each multiple name the EBITDA figure it was divided by?
6. List what the figures cannot settleThe open questions, written out as a list at the end rather than dispersed as hedging.
   Checking: can a reader see, without inferring anything, what the write-up has not answered?

The sixth part is required rather than optional, and its absence is what separates a case study from an advertisement. The claim sounds harsh until the alternative is examined. An account with parts one to five and nothing after them is complete, accurate and internally consistent, and it quietly implies that every remaining question has been settled. The account has not settled those questions. Not mentioning a question is not the same as closing it, and a reader has no way to tell the difference between a question that was considered and closed and a question that was never raised.

Why report both movements when the net is only Rs 3 crore?

Part three of the structure is where a careful writer most often takes a shortcut that looks like tidiness. The agreement fixed a working capital pegThe normalised level of working capital the agreement fixes in advance, so the price moves by the difference between that level and the actual level measured at completion. of Rs 96 crore. Actual working capital at completion came in at Rs 108 crore, so the price moved up by Rs 12 crore. Separately, net debt was assumed at Rs 180 crore and came in at Rs 195 crore, so the price moved down by Rs 15 crore. Plus Rs 12 crore and minus Rs 15 crore net to minus Rs 3 crore, and the equity value moves from Rs 1,140 crore to Rs 1,137 crore.

Minus Rs 3 crore on Rs 1,140 crore is 0.26 per cent. The movement is genuinely small, and the temptation is to write one clause: the completion adjustments reduced the price by Rs 3 crore. The one-clause version is true, and it destroys information that cannot be recovered from anywhere else in the account.

A reader given only the net of minus Rs 3 crore cannot recover either of the two movements behind it, and a reader given both gross adjustmentsEach completion movement reported at its own full size, before it is netted against the movement running the other way. can rebuild the entire part from scratch. From plus Rs 12 crore and minus Rs 15 crore, a reader can work out the peg, the actual, the assumed net debt and the actual net debt if any two of them are given. From minus Rs 3 crore, a reader can work out nothing at all, and cannot even tell that two separate things happened.

There is a household version of this that makes the asymmetry obvious. An electricity bill for the month comes in Rs 200/- lower than last month. Useful? Barely. Split it: Rs 900/- less was used because the weather turned, and the tariff change added Rs 700/-. Same net figure, and now there is something to act on next month. The net said almost nothing precisely because two forces of very different size cancelled inside it.

How is the multiple reported when more than one is right?

Part five is the shortest part of the structure and the one that most often goes out wrong, because a multiple looks like a single fact and is not. The Harivansh Packaging purchase carries a contingent portionAn amount payable only if a stated condition is met after completion, so it is neither part of what was paid on the day nor safely left out of the account. of Rs 60 crore, payable if Sundarban Polymers reaches EBITDA of Rs 145 crore in its first year after completion. Rs 145 crore is 9.8 per cent above the Rs 132 crore it earned. The moment a conditional payment exists, so does a second denominator, and both of the resulting multiples are true.

Work through the four figures. The headline was struck at 10.0 times: Rs 1,320 crore over Rs 132 crore. At completion, the sellers received Rs 1,137 crore and the business carried Rs 195 crore of net debt, so enterprise value at completion is Rs 1,332 crore. Divided by the Rs 132 crore earned, that is 10.09 times. If the contingent amount becomes payable, the maximum enterprise value is Rs 1,392 crore. Divided by the Rs 132 crore earned, that is 10.55 times. Divided by the Rs 145 crore the payment is conditioned on, it is 9.60 times.

Where more than one multiple is correct, the write-up gives the one it is using and names the other in the same sentence, with the denominator stated for each. Giving both readings is double reportingGiving both correct readings of the same measure in one sentence, each with the denominator it was struck on, rather than picking one and leaving the other unmentioned., and it is one extra clause. Not a paragraph, not a footnote, not an appendix. One clause of about a dozen words, sitting where the figure sits.

Omitting that clause is how a purchase gets recorded as cheaper than it was. Notice which of the two a writer will reach for if they are only going to give one: 9.60 times reads lower than 10.55 times, and a writer who is fond of the purchase will find the lower one arriving under their hand without any conscious decision. The reader then carries away a purchase that looks like it was done under ten times, when the figure struck on what the business actually earned is above ten times.

What belongs in the section that settles nothing?

Part six is a list, and the discipline is that it is written as a list. Not as qualifications scattered through parts one to five, not as a closing sentence beginning with the word however, and not as a general acknowledgement that the future is uncertain. A named cannot-settle sectionThe named part of a write-up that lists the questions the figures do not answer, written as a list at the end rather than as hedging dispersed inside the other parts. with items in it, each of which a reader could in principle go and investigate.

For this purchase the list has five items. Whether Rs 1,140 crore was the right price. Whether any synergy arrives, and what it costs to achieve. Whether the Rs 60 crore contingent amount is earned. What Harivansh Packaging Limited would otherwise have done with the money. And how the combined business actually performs. Every one of those is a genuine question about this purchase, and not one of them is settled by any figure in parts one to five.

A reader who knows what is open can go and look, and a reader given a guess has been given something to repeat, so listing what is open is more useful to a reader than any speculation about it. The distinction is worth sitting with. A guess about whether the synergy arrives is not a small service to the reader. A guess is a liability, and it will be quoted back with the writer's name on it and without the hedge that was attached to it. A list of open questions cannot be misquoted in the same way. There is no claim in it to detach.

There is also a practical benefit that has nothing to do with honesty. Writing the list forces the writer to notice which of the open questions are actually answerable with a little more work and which are not. Whether the contingent amount is earned becomes knowable in about a year. Whether the price was right may never become knowable at all. Knowable in a year and never knowable are different kinds of open, and a reader who can see the difference is better served than one who was handed a confident sentence about both.

What does the finished write-up actually look like?

Here is the whole purchase written to the six-part structure. Read it as a specimen rather than as a description of a specimen: this is the output the structure produces, and it is short.

Count what the discipline actually cost. Against a careless version of the same account, the specimen above adds four clauses: the basis the 10.0 times was struck on, the net debt deducted in the bridge, the two gross movements instead of the net, and the second denominator on the multiple. Four clauses, perhaps forty words. Forty words are the entire price of the form.

Which purchases get written up, and what does that leave out?

Now the part that makes this more than a template, and it concerns the shelf of case studies available to learn from rather than the one being written. Purchases get written up when they end clearly. A purchase that turns out very well gets written up because there is a story. A purchase that turns out very badly gets written up because there is a lesson, and often because somebody wants the lesson attached to somebody else. A purchase that turns out roughly as expected, with the business performing about as the buyer thought and the contingent amount partly earned, gets written up by nobody at all.

Purchases get written up when they end clearly, well or badly, and the ordinary ones in the middle are the majority of what happens and almost none of what is available to read. The gap in the middle is a selection effectThe distortion that appears when the examples available to read are not a fair draw from the examples that exist, so what is visible is systematically unlike what is typical., and it is doing quiet damage to anybody who builds their sense of how purchases work out of the written record.

The same shape appears in an entirely ordinary setting. Anybody asked about restaurants will describe the outstanding meal and the food poisoning. The eleven adequate dinners in between produced no story and left no trace, even though they are what eating out mostly is. The reviews available to read are not a fair draw from the meals that were eaten, and everybody knows this about restaurants while forgetting it about transactions.

The consequence is not that the dramatic examples are wrong. The dramatic examples are usually accurate. The consequence is that the mechanism worth learning ran in all of them, including the ordinary ones, and it is far easier to see where no dramatic outcome is competing for attention. In a purchase that ended spectacularly, every reader attributes everything to the ending. In a purchase that ended at about the expected place, the bridge and the adjustments and the denominators are all still there, doing exactly what they do, and nothing is drowning them out.

The same point is the argument for writing up an ordinary purchase even though nobody asked for it. The write-up will be less interesting to read and more useful to learn from, and the six-part structure is what makes an unremarkable purchase worth reading at all: it gives the reader something to follow when there is no story.

Why does the write-up stop short of a verdict?

A case study written to this structure does not conclude. There is no final paragraph saying the purchase was sensible, or expensive, or a mistake. The refusal irritates readers, who came looking for exactly that paragraph. The reason for it is specific rather than a general caution about opinions.

The arithmetic can be recomputed by anybody holding the figures, and the merit cannot be recomputed by anyone at all. Every number in parts one to five is checkable: given the Rs 132 crore, the Rs 1,320 crore, the Rs 180 crore, the peg and the actuals, a reader lands on Rs 1,137 crore and 10.09 times without having to trust the writer. Checkability is the entire value of the form.

Merit is not like that. Whether Rs 1,140 crore was the right price depends on what Harivansh Packaging Limited would otherwise have done with the money, and on what the combined business goes on to achieve. Neither of those is contained in any published figure, and neither becomes available by thinking harder about the figures that are published. A verdict in the last paragraph is not the arithmetic taken one step further. A verdict is a different kind of claim, resting on things the write-up has not established, and attaching it to five parts of checkable work borrows credibility that the verdict has not earned.

Something else happens to the write-up the moment a verdict appears in it. Everything above the verdict starts reading as support for it, whether or not it was assembled that way. The bridge stops being a bridge and becomes evidence. The reader stops checking and starts agreeing or disagreeing. A write-up with no conclusion keeps the reader in the position of somebody following an arithmetic chain. Only from that position can the writer's mistakes be caught.

What question does a writer put to a draft?

The writing testA question a writer puts to their own draft before releasing it, aimed at the single figure a reader is most likely to carry away from the write-up. is one sentence, and it is the practical close of everything above. The question put to the draft is this: a reader will remember exactly one number from it, so which number will it be, and is that the number the writer meant them to keep?

Two things make it a useful test rather than a slogan. First, it has a definite answer, found by looking at the opening sentence, so it cannot be waved away. Second, it converts every rule above into a single check. If the figure in the first sentence has no basis attached, the draft fails. If it is the maximum rather than the amount that moved, the draft fails. If the multiple in it has no denominator, the draft fails. One question, and the whole form is in it.

Ashwin Rege, who leads the transaction team at Harivansh Packaging Limited, and Devyani Kulkarni, its chief financial officer, will both be quoted from whatever gets written about this purchase. Which figure travels with their names attached is settled entirely by whoever writes the first sentence, and it is settled before anybody reads part six.