Transaction Announcement and Market Rumour: What Differs

An announcement has an author who can be identified, a date, and figures somebody is accountable for. A rumour has a figure and none of the other three. A rumour can be correct and an announcement can be incomplete, so the difference is not accuracy. The difference is that only one of them can be checked, corrected or held against anyone later.

What three questions can be put to any figure?

Somebody says the hall for a cousin's wedding costs Rs 4,80,000/-. Before a single other plan is built on that number, there are three things worth knowing, and whether the number is right is not one of them. Who said it. When they said it. And whether that person can be gone back to and asked what the figure covered. A hall quoted at Rs 4,80,000/- with the catering inside it and a hall quoted at Rs 4,80,000/- with the catering outside it are two different evenings.

If the figure came off a printed slip with the hall's name on it and a date at the top, all three questions have answers. The slip can be carried back in. If it came from a neighbour who heard it from somebody at a function, none of the three has an answer, and there is nobody to walk back to. Only one of them can be taken back to its source, so the slip and the neighbour may quote the same number and still not be the same kind of thing.

The same three questions carry straight into a transaction. Every figure ever written about a purchase arrives in one of those two conditions. A figure either has attributionThe link between a figure and the person or body that put it out, so that it can be traced back to a source and asked about. or it does not. The three questions above settle which, in the order they are worth asking.

Question one, who wrote it

An author is a person or a body that put the figure into the world and can be named. Not the publication that repeated it, and not the person who forwarded it on. Only the parties to a purchase were in the room where the number was agreed, so the author of a figure about a purchase is one of them. Everyone else in the chain is repeating.

Authorship is a yes or no question with no middle. A figure attributed to unnamed people close to the transaction has no author, in exactly the way a bill with no shop name on it has no shop. Such a figure reads as though it has an author, and that is what makes the question worth being firm about.

Question two, when it was written

A date does two jobs. A date fixes what was known at the moment the figure was stated, and it makes the figure comparable with anything that came afterwards. A price stated in an announcement and a price paid at completion may differ, and the only way to see by how much and why is to have both figures with their dates attached. Without the date, the second figure looks like a contradiction of the first rather than a movement from it.

Question three, who is accountable

AccountabilityBeing answerable for a figure: somebody can be asked to explain it, and can be held to it if what happens later does not match. is the question the other two are really building towards. Somebody can be asked to explain the figure, or nobody can. When Harivansh Packaging Limited states a number about its purchase, the company has put its own name against it, and if the number turns out to be described wrongly, the company is the one who has to deal with that. When a figure circulates with no author, there is nothing to hold and nobody to ask.

Notice what these three questions are not. The three questions are not a test of whether the person is honest, or clever, or well informed. The test is whether the figure is connected to anything. A figure that fails all three is not a bad figure, and not a figure at all in the sense required, only a numeral in circulation. A table that lets one in carries a row nobody can trace afterwards.

What makes something an announcement rather than a report?

An announcement is a statement made by a party to the transaction, on a date, in a place where it can be found again. Three conditions, and each one is doing work.

Only a party to the transaction was present when the figure was set, so the first condition is about position rather than about expertise. A very well informed observer who was not in the room is still reporting. The distinction is not about how good anybody is at their job. The distinction is which of them can say what the number was measured on from direct knowledge rather than from inference.

The second condition, the date, has already been covered. The third is the one that gets treated as a formality and is not. Being findable again means the statement sits somewhere fixed. In eight months, when the purchase has completed at a different figure, both versions can then be put side by side. A statement that cannot be retrieved cannot be checked against what actually happened, and checking it against what actually happened is most of what makes it worth anything. A promise that cannot be looked up again is a conversation, not a record.

What the announcement in this purchase actually said

Harivansh Packaging Limited announced the acquisition of 100 per cent of Sundarban Polymers Private Limited. The statement carried four figures, and this is the part worth slowing down for: every one of the four arrived with a sentence beside it saying what it was measured on.

The enterprise value was Rs 1,320 crore, being 10.0 times Sundarban Polymers' earnings before interest, tax, depreciation and amortisation (EBITDA) of Rs 132 crore for the last full yearThe most recent complete twelve month period the accounts cover, as against a part year or a period still running.. The equity value was Rs 1,140 crore before completion adjustmentsThe changes to the price that the agreement itself defines, computed on the actual position at completion rather than on the position assumed at signing., being the Rs 1,320 crore less the Rs 180 crore of net debt that Sundarban Polymers carried at signing. A further Rs 60 crore was payable only if Sundarban Polymers reached Rs 145 crore of EBITDA in the first year after completion. And the shares acquired were 100 per cent, so the whole of the business passed.

The sentence beside each figure is the figure's basisThe words that say what a figure was measured on: which quantity, over what period, and on what assumption.. The basis is not commentary and it is not padding. Rs 1,320 crore on its own is a numeral. Rs 1,320 crore at 10.0 times Rs 132 crore of EBITDA for the last full year can be taken apart, put back together, and set against another purchase measured the same way, so it is a figure that can be worked with. An announcement is worth reading not because its figures are more accurate but because its figures arrive attached to the sentences that make them usable.

What is a market rumour, considered as a structure?

The word carries a dismissive reading that gets in the way. Take it off first. A rumour is not a lie, and treating it as one leads a reader straight past the real difficulty. Structurally, a rumour is a figure in circulation with no identifiable author, and that is the whole definition. A rumour carries a number, and it carries nothing else.

Suppose a figure of Rs 1,400 crore is circulating about this purchase before anything is announced. Consider what can be done with it. The figure can be read. There is nobody to ask what it was measured on. Whether it is an enterprise value or the cheque to the sellers cannot be told. Whether somebody has put the conditional Rs 60 crore inside it or left it out cannot be told. There is no date on it, so whether it was current a week ago and has since been superseded cannot be told.

The honest part, which is where it gets difficult

A figure in circulation is often substantially right, and that is precisely what makes it dangerous rather than what makes it safe. Look at the Rs 1,400 crore against what was eventually announced. The announced enterprise value was Rs 1,320 crore, so the circulating figure sits 6.06 per cent above it. The maximum enterprise value of Rs 1,392 crore is what this purchase reaches if the conditional amount is paid in full. Against that maximum, the circulating figure sits 0.57 per cent above. The circulating figure was close. On a different day, with a different circulating figure, it would not have been.

Here is the trap in one sentence. The comparison can only be made after the announcement exists. At the moment the circulating figure was the only thing available, there was no way of telling a 0.57 per cent version from a 40 per cent version, and that moment is the only moment at which the question mattered. Afterwards, when the check is finally possible, the announcement is there and the check is no longer needed.

The household version is familiar. A neighbour says the flat downstairs sold for about Rs 90 lakh. Some months later the registered figure turns out to be Rs 88 lakh, and the neighbour looks well informed. The neighbour being right once says nothing about the next thing the neighbour says, and it would certainly not have justified negotiating a sale off the earlier number. Being right once is a fact about one figure. Being checkable is a property of a source.

So the honest position on a circulating figure is neither trust nor scorn. The position is that the figure has one property, its value, and none of the three properties that would make it usable. Saying that a rumour is usually roughly right may well be true, and it does not help the reader who has to use the figure now.

Why is accuracy the wrong test for telling the two apart?

Both sides are now defined, so the two can be put side by side. Start with the test almost everybody reaches for first, and watch it fail.

Using accuracy as the test requires knowing the answer already, and the answer is exactly what a reader does not have at the moment of reading. That is not a subtle objection but the whole objection, and once it is stated the test collapses. If the circulating Rs 1,400 crore could be checked against the truth, the circulating figure would not be needed. The only situations in which accuracy is available as a test are the situations in which no test is needed.

There is a second failure, and it runs the other way. An announcement is not automatically accurate in the sense a reader wants. An announcement is accurate about what it says, and what it says may be incomplete. The announcement in this purchase was complete about the four figures it carried. At its date nobody knew what the completion adjustments would eventually do, so it was silent about them. The final position moved: the equity value went from Rs 1,140 crore to Rs 1,137 crore once the working capital and net debt adjustments defined in the agreement were computed on the actual position. The Rs 3 crore is a real movement, and the announcement did not predict it. An announcement is a statement of what is agreed, not a forecast of what will be computed.

So accuracy gives the same reading on both sides of the comparison. A circulating figure can be right. An announcement can be incomplete. A test that answers the same way for both sides has not separated them, and a test that separates nothing is not a test.

The test that does work

Replace it with the working question: can this be traced back to somebody who is accountable for it. The traceability question always has an answer, it has one at the moment it is needed, and it separates the two cleanly every time. The question has one more property the accuracy test lacks. Traceability survives being wrong. An announcement that turns out to describe something differently from how it eventually happened is still traceable, still dated, still attached to a party who can be asked what changed. A circulating figure that turns out to be perfectly accurate is still none of those things, and no comparison can be built out of it.

Hence the three questions come first and the definitions come second. The definitions describe two kinds of statement. The questions are the working procedure.

Which figure does an announcement lead with, and why that one?

A transaction does not produce a figure. A transaction produces several, all correct, describing different things. Which one goes at the top of the announcement is chosen by somebody, and the choice is not neutral.

Two of them matter here, and both were worked in full above. The enterprise value of Rs 1,320 crore describes the whole business, including the borrowings that came across with it. The equity value of Rs 1,140 crore describes the cheque, the amount the sellers actually received before adjustments. The gap between them is Rs 180 crore, Sundarban Polymers' net debt at signing, and naming the gap in the same breath as the price is what keeps the two figures from being read as one.

The larger figure describes the business and the smaller one describes the cheque, and an announcement leading with the larger one has made a choice rather than reported a fact. There is nothing improper in the choice. Enterprise value does not depend on how much debt the acquired business happened to be carrying, so it is the figure that lets one purchase be set against another, and that is a real argument for putting it first. But it is an argument, and the sellers' side of the table would lead with a different first figure if they were writing.

The everyday version sits in any property conversation. A flat is described as a Rs 1 crore flat. Whether that means the price agreed, the price plus the registration and the stamp duty, or the price plus the two years of work the new owner is planning, depends entirely on who is speaking and what they are trying to convey. Nobody is lying in any of the three versions. The three versions are three correct figures about one flat.

The discipline this asks of a reader is small and it is absolute. Never accept a figure about a purchase without establishing which of the several correct figures it is. A reader who takes Rs 1,320 crore as the amount the sellers received has not been misled by anybody. The reader has simply not asked, and the announcement did not force them to.

How does a figure lose its basis one retelling at a time?

Here is the mechanism, and it is worth naming rather than treating as carelessness. A figure leaves an announcement with its basis attached. Each retellingAny second-hand version of a figure: a report, a summary, a note passed on, or a row in somebody else's table. drops a little of it, for perfectly ordinary reasons of length and emphasis. By the third repetition the figure is travelling alone. Call it basis decay: a slow loss of the attached sentences, with no false statement made at any step.

Watch it happen to one figure from this purchase.

Stage zero, the announcement

Rs 1,320 crore of enterprise value at 10.0 times the Rs 132 crore of EBITDA earned in the last full year. Rs 1,140 crore of equity value before completion adjustments. Up to Rs 60 crore more, conditional on Rs 145 crore of EBITDA in the first year after completion. Four things are knowable from this: which of the enterprise values it is, which EBITDA it was struck on, whether a conditional amount sits inside it, and whether the figure is the business or the cheque.

The first retelling

A report says the purchase values Sundarban Polymers at Rs 1,320 crore, up to Rs 1,392 crore including the earn-outAn amount payable only if something happens after completion, so it is a permission to pay rather than a payment already made.. Both figures are correct. The Rs 1,392 crore is the maximum enterprise value, being the adjusted equity value of Rs 1,137 crore plus the conditional Rs 60 crore plus the Rs 195 crore of net debt actually there at completion. Nothing false has been said.

The equity value has gone. A reader now has two enterprise values and no cheque, so the natural reading of Rs 1,320 crore is that it is what the sellers received. The sellers received Rs 1,140 crore before adjustments and Rs 1,137 crore after. The gap between what the reader believes and what happened is Rs 183 crore, and it was created by an omission rather than by a statement.

The second retelling

A summary says the purchase was worth up to Rs 1,392 crore at about 9.6 times. Still arithmetically correct: Rs 1,392 crore divided by Rs 145 crore is exactly 9.60 times. But look at what has happened to the denominator. The denominator has silently become the Rs 145 crore that the conditional payment is conditioned on, rather than the Rs 132 crore that Sundarban Polymers actually earned. The purchase now reads 4.0 per cent cheaper on the multiple than the announcement described it, and nobody has changed a price.

The third retelling

A comparison table records the purchase at 9.6 times, with no note. Nothing false was said at any of the three steps, and the entry in that table cannot now be compared with anything. Basis decay is a mechanism, not the fault of any individual. Every person in the chain shortened a true statement. The result is a row that looks like data and behaves like noise.

One last thing about that row. The Rs 132 crore Sundarban Polymers earned is the denominator most people would assume. A later reader who takes 9.6 times and multiplies it back by that figure gets about Rs 1,267 crore. Rs 1,267 crore appears nowhere in this transaction. The amount is not the announced enterprise value, not the adjusted one, not the maximum. The retelling did not merely lose information. The retelling manufactured a figure that never existed.

What did the three retellings actually destroy?

Set out what the announcement supported, using only the figures established above. Three enterprise values are correct for this purchase. Rs 1,320 crore as announced, on the Rs 180 crore of net debt assumed at signing. Rs 1,332 crore once the completion adjustments defined in the agreement were computed, taking the equity value to Rs 1,137 crore against actual net debt of Rs 195 crore. And Rs 1,392 crore as the maximum, if the conditional Rs 60 crore is earned in full.

Two EBITDA denominators are correct as well. Rs 132 crore is what Sundarban Polymers earned in the last full year. Rs 145 crore is what the conditional payment turns on, 9.8 per cent above what Sundarban Polymers earned. Three numerators and two denominators give six multiples, and every one of the six is arithmetically correct.

The six run from 9.10 times to 10.55 times. The spread is 1.44 turns on the exact figures, or 1.45 if the two rounded endpoints are subtracted as printed, and the rounding is worth saying out loud rather than hiding. The table entry preserves one of the six and preserves nothing whatsoever about which one, so the single most useful thing about the figure has been discarded while the figure itself survives.

Put the spread into rupees to feel it. Applied back to the Rs 132 crore Sundarban Polymers earned, 9.10 times and 10.55 times describe enterprise values Rs 190.3 crore apart. Rs 190.3 crore is more than the entire net debt the target carried. The gap is the size of a small transaction, and every rupee of it is reporting rather than price.

Now put the circulating figure back beside the table entry. The comparison is instructive and slightly uncomfortable. The Rs 1,400 crore that was circulating before any announcement is not more wrong than the 9.6 times entry, only unusable in a different way. The 9.6 times entry has an author somewhere and once had a basis; the trouble is that the basis is no longer attached. The Rs 1,400 crore never had one at all, and there is nobody to ask what it was measured on. Both end up in the same place, a number that cannot be used, and they got there by different routes.

What is to be done with a figure that has no author?

The procedure is short enough to carry in the head. Three checks, in order, and one instruction if any of them fails.

The first check is the author: not the publication that carried it, but the party to the transaction who put the figure out. The second is the date, fixing what the figure was describing and when. The third is which figure it is, meaning which of the several correct values, struck on which denominator, with or without any conditional amount inside it. Three checks, each of which either has an answer or does not.

If any one of the three is missing, the number is unusable rather than approximately right, and treating an unsourced figure as approximately right is the habit that ruins a table of purchases. Not roughly correct. Not a reasonable starting point. Not something to put in with a question mark beside it. Unusable, until the missing check can be answered from the announcement itself.

The rule will feel severe the first few times, and the severity is the point. Every softer rule collapses under pressure. Allow in one figure that is probably about right and there is no way to stop, and the row allowed in looks identical afterwards to the rows sourced properly. There is no visual difference in a finished table between a row taken from an announcement and a row taken from a summary of a summary. The decision therefore has to be made when the row goes in, and it cannot be revisited later.

And if all three checks answer, the row goes in with the answers recorded beside it. Which value it is. Which EBITDA it was struck on. Whether the conditional amount is inside the numerator. The three notes cost one line, and they are what makes a table something the next reader can use. The third retelling carried none of them.

How do the people around a transaction use this in practice?

The reading procedure is not an academic exercise. Four kinds of reader run it, and each of them runs it slightly differently.

An analyst building a comparison set produces nothing but rows, so they run it hardest. For them the third check does the most work: which of the several correct values is this, on which denominator. Devyani Kulkarni, chief financial officer of Harivansh Packaging Limited, would expect the same discipline of anyone inside the company setting this purchase against others, and Ashwin Rege, who leads the transaction team, would be the one asked where a row came from.

A lender assessing a borrower who has just bought something runs it for a different reason. The lender needs to know what left the borrower's bank account, so the equity value matters and the enterprise value is background. A lender who reads Rs 1,320 crore as the cash outflow has overstated it by Rs 180 crore, and a lender who then also treats the conditional Rs 60 crore as certain has overstated the certain part of it again. The third check asks which figure it is, so the three checks catch both.

An investor reading about a purchase in a listed company they hold runs the procedure to know whether they are reading the company or reading somebody's summary of the company. Everything downstream of the original statement has already begun losing its basis, so the most useful single habit is to find the original before forming any view. A filingA document a company lodges with an exchange or a registry, which stays retrievable afterwards at a fixed place. stays retrievable, which is exactly what makes the original worth the extra few minutes.

And the household version runs every day without the vocabulary. Somebody says what the neighbours got for their flat. The questions are who told them, when, and whether the figure included the parking and the society transfer charges. If the answer to any of those is a shrug, all that has been learned is that the figure exists, and nothing else, and no flat should be priced off it. The household questions are the same procedure, applied to a smaller number.