Deal Narrative vs Investment Case: Told or Underwritten

A deal narrative is the account of a purchase given to everyone outside it: what is being bought, and why. An investment case is the reasoning that had to justify the price: which sources of value were being counted on, and what the buyer must achieve to earn them. One is told. The other is underwritten, and only the second can ever be shown to be wrong.

Both documents describe one transaction, so a reader meeting them for the first time usually assumes they are the same thing said at two lengths. The two documents are not the same, and the difference is not about tone or detail. The difference is about what each one asserts. A narrative asserts facts and intentions. A reasoning paper asserts amounts against dates. Facts and intentions cannot fail. Amounts against dates can, and that single asymmetry is the whole of the difference.

What purchase are both documents describing?

One purchase supplies every figure below, and its shape is worth having before either document is opened. Harivansh Packaging Limited, an invented listed maker of rigid and flexible packaging, is buying all of Sundarban Polymers Private Limited, an unlisted maker of flexible films. Sundarban Polymers already supplies several of the customers Harivansh Packaging sells to.

The sellers received Rs 1,140 crore. The Rs 1,140 crore is what is left once the target's own net debtBorrowings with cash netted off, so a business sitting on a large cash balance is treated as owing less than its loan balance says. of Rs 180 crore comes out of an enterprise valueThe whole business priced as one thing, lenders and shareholders together, before anybody's debt has been taken out of the figure. of Rs 1,320 crore, and that enterprise value is the one struck at ten times earnings before interest, tax, depreciation and amortisation, or ten times Sundarban Polymers' EBITDAEarnings before interest, tax, depreciation and amortisation. A profit measure taken high up the ladder, before financing costs and before anything is written off for assets wearing out. of Rs 132 crore. The enterprise value is the price of the whole business. The Rs 1,140 crore is the money that actually left the buyer and reached the sellers. Write the two figures down separately and keep them separate. Confusing the two is the single most common mistake made with transaction figures, and an announcement that quotes only the larger number is not lying, it is simply being read carelessly.

New borrowing of Rs 1,000 crore carries most of that payment, at 9.0 per cent, a rate this transaction contracted for itself and never a statement about what borrowing costs in India. The other Rs 140 crore came straight off the buyer's own cash balance, emptying that balance. And there is one more number worth having. The buyer paid Rs 1,140 crore for a business whose books show a net worth of Rs 320 crore, and the difference is Rs 820 crore of goodwillWhat sits on the buyer's balance sheet when the price paid runs above the book value of what was bought, before any part of it is assigned to identifiable items. before anything is assigned to identifiable items. The treatment of that Rs 820 crore afterwards is an accounting question, settled where accounting is taught.

Now the part that shapes everything that follows. Harivansh Packaging turns Rs 3,180 crore of revenue into Rs 477 crore of EBITDA. Sundarban Polymers turns Rs 880 crore into Rs 132 crore. Dividing each of those gives the same answer twice: 15.0 per cent, and 15.0 per cent again. The effect on the combined business is the first thing to settle.

The revenues add to Rs 4,060 crore. The EBITDA adds to Rs 609 crore. Two identical fractions added together give the same fraction back, so the second divided by the first is 15.0 per cent with nothing left over. The most convenient story anybody could tell about this purchase, that a good business bought a better one, is arithmetically unavailable here. The record carries matching margins, so the easy explanation cannot be reached. Every account of this transaction has to point somewhere other than at profitability on revenue.

What is a deal narrative, and who is it written for?

A deal narrative is the short account of the transaction that goes to everybody who is not inside it. Employees hear it in a town hall. Customers hear a shortened form from a salesperson. Lenders read it in a covering note. The market reads it in an announcement. The account says what is being bought, from whom, roughly for how much, and why the buyer thinks the two businesses belong together. A narrative is usually a paragraph, occasionally several, and almost never longer.

Here is the temptation to resist. Because the narrative carries no arithmetic, it is easy to treat it as the soft version, the marketing wrapper around the real thinking. A transaction that nobody can explain in two sentences is a transaction that nobody will help to deliver, so a narrative is not a lesser document. Think of a shopkeeper who has just taken over the shop next door. Ten people work for her. On Monday morning she has to say something, and if what she says is confused, or if she says nothing at all, then for the next six months every person in both shops is guessing. Staff guess about their jobs. Staff guess about whose till procedure wins. Staff guess about which supplier stays. A clear two-sentence account does not make the takeover work, but the absence of one reliably makes it worse.

So the narrative has a real job and it does it by pointing outward. Its audience is plural and it has to survive being repeated by people who were never in the room. A narrative is therefore short, avoids conditionals, and prefers a reason over a number. A reason travels. A number invites a follow-up question that the person repeating it cannot answer.

Notice what follows from that design. A document written to be repeated by strangers is a document that has been deliberately stripped of everything a stranger could get wrong. And the things a stranger could get wrong are precisely the things that could later be checked.

What is an investment case, and what does it commit somebody to?

An investment case is the internal paper that had to stand behind the amount actually handed over. The paper has four slots and is only an investment case when all four are filled. Which sources of value are being counted on. How much each of them is worth, in rupees. By when. And who inside the buyer is answerable if the amount does not arrive. An investment case names an amount, a date and a person, and a narrative names none of the three. The amount, the date and the person are the whole distinction, and everything else follows from them.

Each of the four slots does separate work. Take them one at a time. The source of value is a mechanism, not a hope: one purchasing contract instead of two, one selling team calling on customers both businesses already share, one despatch warehouse instead of a pair. The amount converts the mechanism into rupees a year, and forces somebody to size it rather than assert it. Value expected in year one and value expected in year five are different promises about the same mechanism, so the date turns a claim into a test. And a figure with nobody behind it quietly stops being anybody's problem, so the name is what makes the paper survive the meeting it was written for.

The household version is uncomfortable and worth sitting with. Somebody buys a second-hand car to start driving for a ride-hailing app. The narrative is that the car will pay for itself. The investment case is that the car costs Rs 4,20,000/-, the loan on it costs Rs 4,900/- a month, the driving brings in Rs 18,000/- a month before fuel and Rs 11,000/- after, and the household will know within four months whether those figures hold. The first version cannot fail. The second version can fail in the fourth month, in public, with a number attached. Failing in public is not a weakness of the second version. Failing in public is the only reason to write it.

Now the awkward part. Nothing in the record of this purchase publishes what Harivansh Packaging's investment case actually said. No source of value is written down. No amount is attached to one. No date, no name. The record publishes, in full, everything the case would have had to answer for: the price, the funding, the combined position and the earnings result. Only one of the two documents was published, so the comparison runs on how each is built and who each is built for, and never line by line on what each of them says. The absence is the ordinary condition of anybody reading a transaction from outside it, and learning to work inside that condition is most of the skill.

Where do the two say exactly the same thing?

Most treatments of this comparison set the two documents against each other from the first line, and that gives a false picture. On a well run transaction they agree about far more than they disagree about. Both documents describe the same buyer and the same seller. Both name the same price. Both give the same intention: that these two businesses are worth more held together than held apart. The reason in the narrative was lifted out of the reasoning paper in the first place, so both usually even give the same headline reason.

The two documents are not opposed, and a reasoning paper that flatly contradicts the account being given outside almost always means somebody changed their mind and did not tell the people repeating the old version. A contradiction of that kind is a communication failure rather than a document design failure, and it shows up in a very ordinary way: a salesperson repeats a reason that the internal paper stopped believing in two months ago, a customer asks a sharp question, and nobody in the room knows which answer is current.

The agreement is what makes the disagreement legible, so hold it clearly. Both documents assert that this purchase is a good idea. The two differ in what they are willing to be held to when they say it. One says it is a good idea and stops. The other says it is a good idea by this much, in this way, by this date, and here is who will be asked about it. Same claim, radically different exposure.

Where do they come apart, and which of the two moves?

The two come apart at exactly the point where a figure would have to be produced. A narrative never committed to a number, so it can absorb any number put in front of it. The narrative never mentioned profit per share, so a narrative told that the purchase reduced profit per share simply shrugs. A reasoning paper told the same thing must give somewhere: either an amount in it was wrong, or a date in it has moved, or the paper concedes that the first year was always going to look like this.

Now the direction of drift, the part people miss. The account given outside gets tidier, shorter and more confident every time it is retold. The reasoning behind the price stays exactly where it was written unless somebody deliberately sits down and rewrites it. Shortening is not dishonesty. Shortening is what repetition does to any sentence. Qualifications fall off first, being the hardest part to remember and the least satisfying part to say. Then the range collapses to its midpoint. Then the midpoint becomes the fact.

Watch this happen in an ordinary setting. A man tells his cousins he bought a shop because the rent was low and the footfall looked promising, though he would need a year to be sure. Six months later he is telling people he bought it because the footfall was good. A year later it has become the shop everyone knew was going to work. Nothing was falsified at any step. Each retelling just dropped the least convenient clause, and there was never a written version to check the retelling against. The investment case is that written version. A document cannot drift the way a memory drifts, and not drifting is its entire mechanical advantage.

What in each one can actually be checked?

In the narrative, only the facts: what was bought, from whom, and for how much. The three facts can be taken to the agreement and to the accounts and settled in an afternoon. Everything after the word because is beyond checking, not because it is false but because it asserts nothing measurable. A sentence saying the purchase strengthens the buyer in flexible films has no quantity in it, no date, and no threshold below which somebody would agree it had failed. There is nothing there to test.

In the reasoning behind the price, everything is checkable in principle. Each source of value has an amount, and an amount can be looked for in next year's numbers. Each amount has a date, and a date passes. Each assumption is stated, and an assumption can be compared with what actually happened. The asymmetry between a claim that can be tested and one that cannot is the entire comparison, and it explains why anybody trying to work out whether a transaction is on course reads the internal reasoning and treats the outward account as background.

People collapse one more distinction constantly, so it is worth holding. Uncheckable is not the same as false. A claim that a purchase strengthens a position may be completely true and still be unverifiable by an outside reader, from where that reader sits, with the documents that can be obtained. The disciplined move is to mark it as unverified and carry on, rather than either believing it or dismissing it. A reader who treats every unverifiable sentence as spin ends up as badly informed as a reader who treats every one as fact.

What do the two look like on this purchase?

Set them side by side and let the figures decide which one carries weight. Here is the account, written the way an account is written.

Harivansh Packaging Limited, a listed maker of rigid and flexible packaging, has agreed to acquire 100 per cent of Sundarban Polymers Private Limited, an unlisted maker of flexible films that already supplies several of the same customers. On combination the two businesses carry revenue of Rs 4,060 crore, being Rs 3,180 crore and Rs 880 crore, and EBITDA of Rs 609 crore, being Rs 477 crore and Rs 132 crore, a margin of 15.0 per cent. The purchase brings customers who currently buy from both businesses under one supplier.

Every sentence of that account is true, every figure in it reconciles, and not one word of it commits anybody to anything. Read it again and try to name what would have to happen for it to be wrong. There is no such thing. The margin figure looks like a commitment and is not. The margin is arithmetic on figures that already exist rather than a claim about figures that do not exist yet.

The reasoning behind the price has to do far more, and it has one result waiting for it on the day the transaction completes. Work the combined earnings per shareProfit after tax spread over the shares in issue, so one shareholder's slice can be read off directly. and this is what comes out.

Three things in that table need saying out loud rather than leaving in a cell. First, the interest. Rs 1,000 crore at 9.0 per cent is Rs 90 crore of interest, and at an effective tax rateThe tax charge for the year divided by profit before tax, which is the rate actually borne rather than any headline rate. of 25.0 per cent the cost that actually reaches shareholders is Rs 67.5 crore. Second, the share count does not move. Cash and borrowing paid for this purchase rather than new shares, so every rupee of the change comes from the profit line and none of it from the denominator.

Third, and this one matters more than it looks. Rs 61 crore is not what the ladder produces. Worked through, the ladder gives Rs 61.35 crore: Rs 98 crore of operating profit, less Rs 16.2 crore of interest on the target's own borrowings at 9.0 per cent, taxed at a quarter. Rs 61 crore is the whole-number figure this record locks, it is the one used here, and it is named as rounded every time it appears. On the exact Rs 61.35 crore the chain gives Rs 218.85 crore, Rs 12.16/- a share and a fall of 2.7 per cent, so the purchase reduces profit per share on either figure and only the size of the headline moves. No exact reconciliation should ever be built on the 2.9 per cent, and none is built here.

So the buyer is paying a price that reduces its own profit per share in the first year. The fall is dilutionA fall in the profit attached to each share already in issue, whether because more shares now exist or because what was bought brought in less profit than the funding cost., and on the locked figures it is 2.9 per cent. There is a clean way to see why, and it depends on putting both halves on the same base. The target brings in Rs 61 crore of profit after tax. Measured against the Rs 1,140 crore that was paid for it, the return is 5.35 per cent. The money used to pay costs Rs 67.5 crore after tax, and struck on that same Rs 1,140 crore that is 5.92 per cent. The gap between them is Rs 6.50 crore a year, or 0.57 of a percentage point, and spread across 18.00 crore shares it lands on Rs 0.36/-. Rs 0.36/- taken off Rs 12.50/- leaves Rs 12.14/-, exactly where the table ended.

Two numbers are now sitting next to each other and it is worth naming the tension between them rather than resolving it. The buyer trades at 12.58 times its own EBITDA on an illustrative share price of Rs 300/-, and it bought this business at 10.0 times. On a multiple comparison the purchase looks cheap. On the funding arithmetic it reduces profit per share. Both statements are true and they are answers to different questions. A multiple compares two prices. The earnings test compares an income against what the money to buy it costs. Neither of those two readings is the answer, and an account that picks one has quietly changed the question.

Now put the account and the reasoning back beside each other. The comparison gets no sharper than this. The fall from Rs 12.50/- to Rs 12.14/- does not touch the account given outside at all. The account never mentioned profit per share, so an earnings result cannot contradict it, and it will be repeated word for word in the year after completion exactly as it was written before. The reasoning behind the price cannot ignore the same result for a moment. Justifying that price against that funding cost is the only thing it was ever written to do. The immunity of the outward account is not a strength, it is precisely what makes it weak evidence about anything.

What does it cost when the account outruns the reasoning?

Sooner or later the buyer has told the outside something the internal reasoning never supported. Usually nobody lied. A phrase that was carefully hedged in the paper arrived in a slide deck without its hedge, somebody liked how it sounded, and it hardened. The gap surfaces at the first moment a person with standing asks for the number behind the sentence. A lender asks it during a covenant review. An analyst asks it on a results call. A new finance director asks it in week three of the job, and week three is when it hurts most. A new person has no investment in the old wording.

The reputational damage is real, but it is not the expensive part. An account cannot be missed, so the expensive part sits inside execution, where a team delivering against an account rather than against a reasoning paper has no way of telling whether it is on track. Sit with that. If the target is that the two businesses come together well, then no week is a bad week. Nothing is behind. Escalation needs a threshold and there is none, so nothing can be escalated. Every conversation about progress becomes a conversation about how people feel it is going, and the person who speaks most confidently wins it.

The street-level version is a food stall that expands to a second cart. If the plan is that two carts will do better than one, the owner cannot tell in March whether it is working. If the plan is that the second cart must clear Rs 1,800/- a day by the end of the second month, then on the last day of the second month there is a fact, and the fact either arrives or it does not. The number is not the point. The date attached to the number is the point. The date converts a hope into something a person can act on while there is still time to act.

Why does the reasoning so often get written backwards?

Writing the reasoning backwards is the failure that produces most of what goes wrong, and it is almost never intentional. The contents of a paper written under time pressure are the first thing to settle.

The price came out of a negotiation that ran under pressure. The internal paper has to exist for a meeting. And the fastest way to write it is to find sources of value that total what has already been agreed, then present the total as the justification. The result looks exactly like an investment case. The paper has the four slots filled in. The paper has a table. And every figure in it was chosen to reach a total rather than estimated on its own and then added, so the paper functions as an account.

What happens to each one after completion?

The account given outside has a long, comfortable afterlife. The account gets repeated in results calls, in induction packs, in internal decks, and in the first paragraph of every subsequent strategy paper. And it gets shorter each time, in the way already described, until three years later it is a single clause that everybody in the building believes and nobody can source. The shrinking is not a scandal, but the normal decay of a sentence that was never written down in a form anyone was expected to check.

The reasoning behind the price has exactly two possible afterlives and there is no third. Either somebody tracks its amounts against what actually happened, at the dates it named, and writes down where each one landed, or the document is quietly abandoned within a year and nobody mentions it again. A commitment nobody checks cannot be missed, and a commitment that cannot be missed is an intention. So a reasoning paper that nobody ever revisits was never functioning as an underwriting, whatever it was called on the day it was approved.

The word underwriting earns its place at this point. To underwrite something is to stand behind it in a way that costs the author if it fails. A paper that names an amount, a date and a person, and is then read again on that date by somebody with the standing to ask about it, is an underwriting. The same paper, filed and never reopened, is an account of what people intended in a particular week. The words in the two papers are identical. The separation lies entirely in what happens afterwards.

What can a reader outside the transaction actually see?

The account given outside, in full: publishing it is the whole reason it exists. Of the reasoning behind the price, almost nothing: only whatever a listed buyer chooses to say or is required to say, and that requirement is settled under Indian markets and regulation. So an outside reader is permanently working with one complete document and one that is mostly absent, and the useful discipline is not to guess at the missing one.

Here is the move that works. Take the published account apart line by line. Sort each line into what can be checked against figures the transaction itself produced and what cannot. Check the first group. Mark the second group as unverified, and leave it there. Marking a claim unverified is a finished piece of work, not a failure to finish one, and treating it as false is just as much an error as treating it as fact.

Which of the two documents to reach for depends entirely on the question being asked, and reading the wrong one is how a sentence ends up being quoted as though it were a commitment. For what happened and for how much, the outward account is the right document and it is sufficient. For the obligations the buyer now carries, the outward account is the wrong document and no amount of careful reading will make it the right one. And for whether the purchase was a good idea, neither one settles it.

How does a lender, an analyst or an employee use the difference?

Three people meet these two documents from three positions, and watching how each of them handles the gap is the fastest way to make this practical.

The lender who put up the Rs 1,000 crore reads the reasoning and treats the account as background noise. Reading the reasoning is not scepticism, it is arithmetic. Borrowings move from Rs 740 crore to Rs 1,740 crore on the buyer's own books, and against its own EBITDA of Rs 477 crore that is 3.65 times, where it was 1.26 times before. Put the target's Rs 180 crore of net debt alongside it and the consolidated position is Rs 1,920 crore against combined EBITDA of Rs 609 crore, or 3.15 times. Two honest measures, and the lender has to be told which one any covenant is struck on. None of that can be read off an account that never mentions a rupee of borrowing. The lender therefore wants the amounts and the dates, and will keep asking until somebody hands them over.

The equity analyst is in a harder position and mostly cannot fix it. The published account, the reported figures and whatever the buyer says on a call are all that is available. The free floatThe part of a company's shares not held by its promoters, and therefore the part that can ordinarily change hands in the market. of Harivansh Packaging is 42.0 per cent, so the people affected by the answer are numerous and none of them was in the room. The disciplined analyst writes down the three checkable lines, checks them, writes down the unverifiable ones as unverifiable, and then asks the buyer for one specific thing: the amount, the date, and who is answerable. Whether that gets answered is itself informative.

The employee inside Harivansh Packaging has the most immediate stake and the least attention paid to it. If Devyani Kulkarni, the chief financial officer, and Ashwin Rege, who leads the transaction team, have circulated the account and not the reasoning, then in month four nobody on the integration team knows whether closing one warehouse was ever worth Rs 1.30 crore a year or whether that figure was somebody's rounding. The team will still work hard. The team will just work hard on whatever seems most urgent, and whatever seems most urgent is not what the price assumed. An investment case that never reaches the people delivering it has been written for the meeting rather than for the work.

Why the distinction is a threshold rather than a dial

A control on this subject would have to assert something the argument denies. A dial running from told to underwritten would draw a continuum between the two documents, when the actual position is a threshold: a paper either names an amount, a date and a person, or it does not, and there is no halfway state to slide through.

A control on the sources of value would be worse. A control of that kind would put numbers on an investment case this record does not publish, and a reader moving the slider would come away with the impression of having seen the buyer's reasoning. And a control on the funding rate would suggest the buyer could have chosen a different one, when 9.0 per cent is what this transaction contracted. Each of those would have to assert something untrue in order to exist.

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