The Disclosure Schedule: Carving Exceptions Out of the Warranties

A disclosure schedule is the seller's list of facts put on the record so that they cannot later be claimed as breaches of the warranties. A warranty says the business is one way. A disclosure says except for this. In money terms, a disclosed fact is a fact the buyer has agreed to carry, so the warranty stops reaching it and the claim disappears.

The warranties themselves are settled ground, and so is the place where the schedules sit inside the agreement. Missing so far is the document that quietly undoes parts of them, one paragraph at a time, without changing a single thing about the business being bought.

Three terms carry the money once a claim is made, and all three are settled under the warranties and the claim limits rather than here. The transaction runs a package of claim filters. A de minimisA floor amount. Anything smaller than it drops out before the counting even begins, so it is never part of what gets tested. of Rs 1.14 crore sets the floor below which a claim is not counted. A basketAn aggregate threshold. Counted claims must get past it before any money moves at all, so a real loss underneath it recovers nothing. of Rs 11.37 crore is the level counted claims must reach before anything is paid. A cap of Rs 227.40 crore is the ceiling on everything the seller can ever be asked for, and an escrowMoney parked with a neutral holder for a fixed stretch after completion, available if a claim later succeeds and not spendable by the seller meanwhile. of Rs 113.70 crore is money parked so that a successful claim has somewhere to be paid from. Once a claim gets past the basket, two ordinary drafting conventions give two different answers on the very same claim: a tipping basketOnce the threshold is reached, the whole claim becomes recoverable, including the part underneath the threshold. makes the whole claim recoverable, and a deductible basketOnce the threshold is reached, only the part of the claim sitting above the threshold is recoverable. The part underneath stays with the buyer. pays only the part sitting above the threshold.

What is a disclosure schedule doing to the warranty schedule?

A flat and a list handed over on the same afternoon have the same shape. The lease says the flat is in good condition. Attached to it is a sheet of existing marks: the scratch on the bedroom door, the chipped tile behind the sink, the tap that drips. At move-out, nothing on that sheet can be taken out of the deposit. Everything not on that sheet can. The lease and the list are one arrangement. The lease read alone describes a perfect flat. The list read alone describes a ruin.

A disclosure schedule works exactly that way against the warranty schedule. The warranties are a run of numbered statements the seller makes about the business: the tax returns were filed, no customer is in dispute, the licences are current, nothing is outstanding against the company. The disclosure schedule is delivered alongside them and is keyed to those same numbers. Against warranty nineteen it says here is what is not true, or here is the qualification. The two documents are two halves of one sentence, and a reader who has been through the warranties without the disclosures has read only half of it.

The shape of the two documents is why an associate handed the paper on a Friday and asked for a summary by Monday gets the summary wrong so often. The warranty schedule is the readable one. The warranty schedule is written in confident sentences and looks like the answer. The disclosure schedule is a keyed list of exceptions with no narrative in it, so it looks like an appendix. The disclosure schedule is not an appendix. The disclosure schedule is the second half of every sentence in the document that matters.

Which base is every threshold struck on?

This transaction carries two candidate bases and only one of them is correct. Settle the base before a single rupee is discussed, and start where the bridge starts. Sundarban Polymers Private Limited, an invented manufacturer, reports earnings before interest, tax, depreciation and amortisation (EBITDA) of Rs 132 crore. The price was struck at 10.0 times that, putting enterprise value at Rs 1,320 crore. Strip out the Rs 180 crore of net debt sitting inside the company and the headline equity value comes to Rs 1,140 crore. Then the completion adjustments ran. Working capital came in above its agreed level and pushed the price up by Rs 12 crore. Net debt came in at Rs 195 crore rather than the Rs 180 crore assumed, and pulled the price down by Rs 15 crore. Netting those two legs leaves minus Rs 3 crore, and the sellers were actually paid Rs 1,137 crore.

Every threshold in this transaction is computed on the Rs 1,137 crore that changed hands, never on the Rs 1,140 crore the bridge produced, and the base is named beside each percentage so the two can never be confused. The record confirms the choice without being asked. Adding the Rs 60 crore earn-out to Rs 1,137 crore brings the ceiling on what the sellers can receive to Rs 1,197 crore, exactly the figure the record carries. The same addition run through the headline instead gives Rs 1,200 crore, overstating the ceiling by the whole Rs 3 crore the adjustments took out. A second route settles it too: adding back the Rs 195 crore of net debt actually there at completion reaches the maximum enterprise value of Rs 1,392 crore that the record also carries.

Now look at what the base choice does to the four filters. Twenty per cent of Rs 1,137 crore is Rs 227.40 crore; twenty per cent of the headline is Rs 228.00 crore. Ten per cent gives Rs 113.70 crore against Rs 114.00 crore. One per cent gives Rs 11.37 crore against Rs 11.40 crore. Nought point one per cent gives Rs 1.137 crore against Rs 1.140 crore. Both print as Rs 1.14 crore, and Rs 30,000/- vanishes at the second decimal.

None of those four gaps is a coincidence. Each one is that filter's own percentage applied to the Rs 3 crore by which the two bases differ, so the cap gap of Rs 0.60 crore is 20.0 per cent of Rs 3 crore, the escrow gap of Rs 0.30 crore is ten per cent of it, and the basket gap of Rs 0.03 crore is one per cent of it. The pattern is forced arithmetic rather than a set of separate results that happen to agree. The pattern also says something useful about checking somebody else's work: if two analysts' caps differ by Rs 0.60 crore, neither of them has made an arithmetic error. The two analysts have used different bases.

What does one paragraph do to the money?

Here is the whole mechanism in one move. The seller makes a statement. The statement is not true of some particular fact. Without a disclosure, that untruth is a breach, the breach converts into a claim, and the claim is tested against the filters. With a disclosure, the buyer was told, so the statement was never presented as covering that fact at all. There is no breach, so there is nothing to test. The loss does not travel to the seller. The loss stays where it landed, and the buyer now has the business and the problem inside it.

Disclosure is the cheapest lever anywhere in the transaction paper. The seller changes nothing whatsoever about the business and removes a liability by writing a paragraph. The alternatives cost a great deal more. To remove the same exposure by fixing the underlying problem, the seller would have to spend real money and real months. To remove it by cutting the price, the seller gives up rupees at completion in a way everybody can see. To remove it by disclosing costs the seller a sentence and an afternoon of argument, and the headline transaction figure does not move at all.

The invisibility of the move is what makes the disclosure schedule so easy to under-read. Nothing about it shows up in the headline. The Rs 1,137 crore stays Rs 1,137 crore. The cap stays Rs 227.40 crore. Only the set of things the cap could ever be reached by has quietly shrunk.

How far does the arithmetic actually move?

Take the three worked claims this transaction carries and walk them through the package, then put a disclosure in front of the biggest one. All three meet the same filters, and those filters were measured against the adjusted base of Rs 1,137 crore.

A claim of Rs 0.8 crore comes in first. The claim sits under the Rs 1.14 crore de minimis, so it is not counted at all. A claim that small does not fail; it never enters. A claim of Rs 8 crore comes in second. The Rs 8 crore clears the de minimis comfortably and is counted, and then it stops short of the Rs 11.37 crore basket. A buyer with a real, provable loss recovers nothing, and the package is doing exactly what it was drafted to do rather than misfiring. A claim of Rs 15 crore comes in third and is the only one of the three that gets past both filters.

The Rs 15 crore recovery depends entirely on a convention sitting elsewhere in the same document. Under a tipping basket the whole Rs 15 crore is recoverable. Under a deductible basket only the part above the threshold is recoverable. Rs 15 crore less Rs 11.37 crore leaves Rs 3.63 crore, being Rs 3,63,00,000/- in full. The deductible answer is 24.2 per cent of the tipping answer, so the choice of convention cuts the recovery by 75.8 per cent, rounded in the record to 76 per cent. The base matters here too. The same subtraction run on the headline basket of Rs 11.40 crore gives Rs 3.60 crore instead, understating the recovery by Rs 3,00,000/-. The understatement is one per cent of the Rs 3 crore base difference, the identical gap the basket itself moved by.

Now suppose the fact underneath that Rs 15 crore claim appears in the disclosure schedule. Nothing else changes. The de minimis is still Rs 1.14 crore, the basket is still Rs 11.37 crore, the cap is still Rs 227.40 crore, the escrow of Rs 113.70 crore is still funded. But there is no breach, so there is no claim, so there is nothing to test against any of it. The Rs 0.8 crore claim was never counted. The Rs 8 crore claim was counted and pays nothing. The third one no longer exists. Not one of the three claims pays a single rupee, and the total moved from Rs 15 crore or Rs 3.63 crore down to nothing because of one paragraph.

One more convention is worth naming, and it does not rescue the buyer. The worked claims here are each tested on their own. A document can instead require the counted claims to be added together before the basket is tested. On that convention the two counted claims add up to Rs 23 crore between them. Rs 23 crore clears Rs 11.37 crore easily and pays Rs 23 crore on a tipping basket or Rs 11.63 crore on a deductible one. Disclosing the biggest one away leaves Rs 8 crore on its own, and Rs 8 crore no longer clears the basket. The answer is nil again. Under every combination of the two conventions, the disclosure takes the recovery from something to nothing, so the buyer cannot draft its way out of this one.

What are the two kinds of entry, and what does each show?

Entries in a disclosure schedule come in two shapes, and the difference matters to a reader for a completely practical reason before it matters to anybody for a legal one.

A specific disclosure names a fact and keys it to a numbered warranty. A specific disclosure is short and it points somewhere. Because a seller does not spend negotiating capital disclosing something it is relaxed about, the specific entry is the most informative paragraph in the entire document. Every specific entry is a small confession of where the seller thinks the trouble is. A banker with an hour and no legal training can still read the list of specific entries and come away knowing what worried the other side.

A general disclosure works the other way round. The general entry points at a body of material and treats everything inside that body as disclosed. The lease analogy again: instead of listing the scratch and the chipped tile, the landlord writes that the flat is let in the condition shown in the photographs already emailed to the tenant. One line of that kind does an enormous amount of work, and how much depends entirely on how many photographs there were and how carefully they were examined. A general entry says nothing about the business. A general entry says how much reading the buying side agreed to treat as already done.

The reading habit that follows is simple. Specific entries answer what is wrong. General entries answer how much the buyer is assumed to have absorbed. Both are disclosures and both remove exposure, but only one of them supplies a place to look next.

When does the schedule arrive, and what does an update do?

The transaction ran 22 weeks from term sheet to completion, of which the conditions periodThe stretch between the day both sides sign and the day the business actually changes hands, used to satisfy the agreed conditions to completion. was nine weeks. Signing therefore fell at week thirteen, and the last nine weeks are 40.9 per cent of the whole elapsed time. The delivery question is about those nine weeks.

Suppose the schedule is delivered at signing and cannot be touched afterwards. Because the disclosure schedule closed at week thirteen, the warranties are repeated at completion against a document that no longer moves. Anything the seller learns during those nine weeks stays exactly where it is, with the seller. Now suppose instead that the paper permits an update at completion, sometimes alongside a bring-downA repeat of the same promises on the day of completion, so the buyer is told whether they are still true at that moment rather than only on the day of signing. of the warranties. Anything discovered in those nine weeks can be written in, and once written in it is no longer a breach. The nine weeks have moved from the seller to the buyer.

One sentence about whether the schedule may be updated hands an entire nine week window from one side of the transaction to the other, and it takes about a minute to find in the document. The asymmetry between how little it takes to read and how much it decides is why it belongs near the top of anybody's checklist.

There is a fairness argument on both sides, and the arithmetic does not settle it. A seller running a business for nine more weeks will genuinely learn new things, and being made to carry every one of them can look harsh. A buyer who priced the business on what it was told at signing did not price the nine weeks. The argument gets resolved by a drafting choice rather than by the arithmetic, and skipping that one sentence is not skipping a technicality.

Why does the price get reopened here?

A disclosure that arrives late is not a formality. A late disclosure is new information about the business, delivered by the person selling it, at the point where the buyer has already spent months and money and has an internal approval built on a number. The buyer has three responses available and no fourth. Accept it and carry the exposure for nothing. Price it and ask for the consideration to move. Or ask for an indemnityA promise to put a specific named cost back into the buyer's pocket if it arises, rather than a promise that something is true. against that particular item, leaving the price alone and carving the one exposure back out separately.

Which of the three a buyer reaches for depends on size, and size is why the earlier arithmetic matters. On this transaction the Rs 15 crore claim is 1.32 per cent of the Rs 1,137 crore paid, 3.14 per cent of the acquirer's own EBITDA of Rs 477 crore, and 6.60 per cent of the cap, whose Rs 227.40 crore would cover a claim that size 15.16 times over. The three readings are the same rupee figure measured against three different things, and a buying team will reach for whichever one supports the position it wants to take.

On most transactions the disclosure schedule is the last document finished. Both sides are still arguing about it when everything else is settled. Lateness means something different once that order is understood. A material entry landing two days before signing is not somebody being disorganised. The entry is a negotiating position, taken at the moment the other side has the least room to walk away from it.

How is a hundred-sheet schedule read in an hour?

A schedule the thickness of a phone directory arrives, with a question about what is in it before the call at four. The instinct is to open at sheet one. The instinct is wrong, and it will spend the whole hour on the entries keyed to the warranties nobody was ever going to sue on.

A vegetable seller closing up for the evening does not count the coriander first. She counts what carries the money. Read a schedule the same way. Start from the warranties capable of producing the largest claims on this particular business. On a packaging manufacturer that means the customer contracts, the environmental and operating permissions, the tax position and any litigation. Get the numbers of those warranties. Then read only the entries keyed to those numbers. Then, separately, read every general entry. A general entry is not keyed to anything and will not turn up in a numbered search.

A disclosure schedule is an index, not a narrative, and reading it in printed order is the reliable way to spend an afternoon and learn nothing. The order the entries were typed in describes the drafting process. The order of the claims they could remove describes the transaction.

What does a thin schedule show?

Four sheets arrive where forty were expected, and something relaxes. Relaxing is premature. A thin schedule has two completely different explanations and the document itself cannot say which one is in hand.

The first explanation is that there was little to disclose. A clean business, run properly, with the paperwork in order and nothing hiding in it. The clean-business reading is genuinely good news. The second explanation is that little was looked for. A short diligence process, a seller who did not go hunting, a buying side that did not ask, and a schedule that is thin because the search was thin rather than because the business is. The thin-search reading is not news at all, but silence.

The check that separates the two sits entirely outside the schedule: a thin schedule after a thorough process is evidence, and the same thin schedule after a short one is an absence of evidence wearing the same clothes. So the question to ask on receiving four sheets is not about the four sheets. The question is how long diligence ran, how many people worked on it, what they were asked to look at, and what they were told they did not need to.

The same trap sits well beyond a transaction. A household that has never had a medical scare and a household that has never had a check-up both report no known conditions. The report is identical and the information content is not remotely the same.

What does this transaction's own record actually contain?

A record can carry the whole machinery of a disclosure schedule and not one line of its contents, and that is the state this transaction is in.

The record for the purchase of Sundarban Polymers Private Limited by Harivansh Packaging Limited, an invented listed acquirer, carries the machinery in full. The record carries the warranties, the claim filters struck on Rs 1,137 crore, the two drafting conventions, the escrow and its eighteen month hold, the nine week conditions period and the completion adjustments. The record carries no disclosed item at all. There is no entry against warranty nineteen, and no warranty nineteen either; the numbers in the first drawing above show how the keying works rather than what this schedule contains.

The arithmetic above is locked by the record, and the disclosed facts set in front of it were built so the machinery could be watched moving money. The Rs 15 crore claim is real in the sense that the record locks it as one of the three worked claims. The disclosure of that claim was added on top, so a recovery could be watched going to nothing.

Two things follow. The arithmetic travels into other work and the content of any entry does not. And a description of a document is not the document. When a description carries vivid specifics, ask whether those specifics came from the document or from whoever wrote the description.

How does somebody on the buying side actually use this?

Three different people pick this document up for three different reasons, and none of them reads it the way a lawyer does.

The associate on the transaction team is looking for the delta. She already has a summary of the warranties. She reads the specific entries against the six or seven warranties that carry the money and asks, for each one, whether it changes anything her team has already told the investment committee. Most entries do not. The two or three that do are the whole reason she opened the document.

The lender financing the buyer reads it as a leverage question rather than a legal one. The purchase is funded with Rs 140 crore of the buyer's own cash and Rs 1,000 crore of new borrowing, so a lender cares about anything that will consume cash at the borrower after completion. A disclosed exposure is exactly that: an amount the borrower will pay and cannot claim back. Measured against the buyer's own EBITDA of Rs 477 crore, an item the size of the Rs 15 crore claim is 3.14 per cent of one year's earnings, small enough to be absorbed and large enough to be asked about.

The analyst on the outside, months later, never sees the schedule at all. The analyst sees a goodwill figure, a subsequent write-off, or a provision appearing in a later set of accounts with no explanation attached. A disclosed exposure does not disappear when it leaves the warranty; it simply stops being somebody else's problem and starts being a cost inside the buyer. Knowing that this document exists, and that it can move a Rs 15 crore item from one side of a transaction to the other in one paragraph, is what turns an unexplained later charge from a puzzle into a hypothesis worth testing.

For a household, the same move is familiar and much smaller. A buyer takes a second-hand two-wheeler, and the seller mentions, in passing, that the battery is on its way out. He has just disclosed. The buyer can still buy it, but can no longer return next month saying he was not told, and the Rs 4,000/- battery is now his. Nothing about the vehicle changed when the seller said it. The change is in who carries the battery.

When a control earns its place

Whether an interactive control earns its place depends on the shape of the relationship it is asked to show, and disclosure is a small lesson in that.

A control is useful when the underlying relationship is continuous and its shape is not obvious. Disclosure is neither. A fact is disclosed or it is not, so the input has exactly two states, and a slider across two states is a bar that sits in one of two positions while pretending to be a dial. Both states are already worked above: the Rs 15 crore claim recovers Rs 15 crore on a tipping basket or Rs 3.63 crore on a deductible one, and once the fact is disclosed both of those become nothing.

The genuinely continuous relationship in this area is a different one: how a claim of varying size behaves as it crosses a de minimis and then a basket. The claim-size relationship is a real curve with real kinks in it, and it belongs under caps, baskets and limits.