The Deal Timeline: The Sequence and Its Pressure Points

A deal timeline is the sequence of milestones with dates attached and a named person answerable for each one. A timeline is a plan rather than a forecast. Much of what it schedules is done by parties outside the transaction. On the Harivansh Packaging Limited purchase the term sheet and completion sat twenty two weeks apart, nine of those weeks being the conditions period and the other thirteen running up to signing.

Anybody looking at a transaction for the first time wants a number. How long does a deal take, three months or nine, and can somebody please just say. No honest number exists, and saying so is not evasion. The absence of an honest number is the single most useful thing anyone can be told about a transaction timetable. The reason no honest number exists is the same reason timetables fail, and once that reason is visible any timeline can be read and the place it will break can be named.

Somewhere ordinary first. A household is planning a wedding for a daughter, and there is a date everybody has agreed on. The work divides in two. Printing the cards, ordering the sweets, booking the band, cleaning the house, getting the clothes stitched: every one of those moves faster if more cousins are pulled in and more money is spent. Then the rest. Which dates the hall actually has free. When the other household's elders will confirm. How long the tailor's own queue is in the marrying season. No amount of effort inside the house shortens any of those. The hall, the elders and the tailor will take what they take.

A wedding plan that treats those two lists as the same kind of work is a plan that will be missed, and it will be missed in exactly the places nobody in the house can fix. So it is with a transaction. Every deal timeline contains two kinds of week that look identical in writing and behave nothing alike, and telling them apart is most of what reading a timeline actually is.

Harivansh Packaging Limited, an invented listed maker of rigid and flexible packaging, is buying 100 per cent of Sundarban Polymers Private Limited, an equally invented and unlisted maker of flexible packaging films. The transaction was struck at an enterprise value of Rs 1,320 crore. Taking off Sundarban Polymers' net debt of Rs 180 crore leaves an equity value of Rs 1,140 crore, and that equity value is the sum that actually reaches the sellers. Devyani Kulkarni is chief financial officer of Harivansh Packaging Limited and Ashwin Rege leads the transaction team. On this purchase, twenty two weeks ran from term sheetThe short document that sets out the agreed shape of a deal, price, structure and main terms, before the long agreements are drafted. to completion, of which the conditions period was nine. The twenty two weeks are this transaction's own and say nothing whatever about how long any transaction takes.

What is a deal timeline, and what is it not?

A timeline is three things at once, and a document missing any one of them is not a timeline. The first is a list of milestonesA point in a piece of work that is either reached or not reached, with nothing in between. A milestone is an event, not a task: it takes no time itself., in order. The second is a date against each of them. The third is a named person answerable for each one. When a date is at risk, that person's job is to say so out loud rather than leave a committee to discover it late.

Now the part that matters more than the definition. A timeline is a plan. A timeline is not a forecast, and the difference is not pedantry.

A forecast is a claim about the world: it will rain on Thursday. The claim is either right or wrong, and being wrong about it is a fact about the forecaster's judgement. A plan is an intention with a person behind it: the drafting will be finished by Thursday. Missing a plan is not a discovery that the world was misjudged. A missed plan is an event, and events on a plan are managed. The date moves, what moved it is stated, and what else it moved is checked.

Teams defend dates they should be moving because they have quietly reclassified their plan as a forecast, and a missed forecast feels like being caught out. Watch it happen in any transaction that is running late. Saying so aloud feels like admitting an error rather than doing the job, so the date stays on the paper long after everyone in the room privately knows it has gone. The cost of that silence is not the missed date. The cost is that every other date downstream is still being worked to as though nothing has changed, and three or four other workstreamsOne strand of work on a transaction, run by its own group of people: diligence, drafting, funding, conditions and so on, each proceeding in parallel. keep spending against a schedule that no longer exists.

One more thing a timeline is not. A timeline is not the agreement. Nothing on a timeline binds anybody to anything. A date can be moved by the person answerable for it after a conversation. The obligations in the transaction cannot be moved that way at all: they sit in signed paper, and what that paper contains is set out under the transaction documents. Confusing the two produces a very specific bad afternoon: somebody treats a slipped internal date as a breach, and somebody else treats a real obligation as a date that can be talked about.

What are the milestones, and how is a span measured?

Seven milestones, named the same way every time throughout this sequence. Approach and confidentiality. Indicative offer and term sheet. Confirmatory diligenceThe detailed checking a buyer does after the shape is agreed, to confirm that what it was told about the target holds up against the target's own records.. Documentation. Signing. The conditions period. Completion.

Read that list once more and notice that it is a sequence of events, not a sequence of tasks. Signing takes no time; it is a moment. The conditions period takes time but nobody works at it in the way that diligence is worked at. Milestones and workstreams therefore have to be drawn separately, and a plan that lists only one of them is missing half the picture.

Now the arithmetic discipline that every span has to obey. A spanA stretch of elapsed time between two named points. A duration with only one end named, or neither, is not a span and cannot be added to anything. is a stretch of time between two named points. A number of weeks with no endpoints attached is not a span at all. An unanchored duration floats free, and three readers in the same meeting can read it three different ways without any of them noticing they disagree.

Take the sentence that appears on real transaction timetables all the time: the conditions period is nine weeks. Nine weeks of what, running from when? On this purchase it runs from signing in week thirteen to completion in week twenty two. Read the same nine weeks from the term sheet instead and completion lands in week nine, thirteen weeks earlier than it actually did. The date the last condition was filed is not in this record. Read the nine weeks from there and they cannot be placed on the calendar at all.

A timeline quoting durations without endpoints is a list of hopes, and it is indistinguishable at a glance from a timeline that has been thought through. That is what makes it dangerous rather than merely sloppy. Nobody reads a beautifully laid out table of weeks and thinks to ask what each row is measured from.

How does the twenty two week span divide, and against what base?

Two spans are locked in this record and every other figure is derived from them by arithmetic that can be checked. Twenty two weeks ran from term sheet to completion. The conditions period, from signing to completion, was nine of them.

So the span from term sheet to signing is twenty two less nine, or thirteen weeks. Notice that this figure is a subtraction rather than an observation. Nobody timed the thirteen weeks; they fall out of the two numbers that were timed. The distinction is worth saying plainly. Derived figures acquire the same authority as measured ones the moment they are typed into a table, and a reader who does not know which is which cannot tell where the record actually ends.

Now the share, and here is where most timeline commentary goes wrong. Nine over twenty two is 40.9 per cent. The base is the twenty two weeks from term sheet to completion, and that base has to travel with the number in the same sentence, every time. Change the base and the same nine weeks says something completely different. Nine over the thirteen controlled weeks would be 69.2 per cent, a true division of two real numbers on this transaction that means nothing anybody would want to know. The other side of the split is thirteen over twenty two, or 59.1 per cent. Adding 40.9 and 59.1 gives 100.0, and that sum is the check that the two parts are parts of the same whole.

A share without its base named is not a share of anything, and on a timeline this is the error that turns a real finding into a slogan. Somebody in a meeting says forty per cent of the deal was spent waiting. Forty per cent of what? Of the twenty two weeks after the term sheet, or of the whole stretch counting back to the first approach, a stretch this record never dated? The first is checkable and the second is not available at any price.

Confirmatory diligence and documentation sit inside the thirteen weeks, and this record does not split them. The missing split is not sloppiness in the record, and it should not be filled in with a plausible guess. The two workstreams overlap, deliberately and unavoidably, and the overlap is the first place this transaction comes under strain. Any table that shows diligence finishing neatly in one week and drafting starting in the next is describing a transaction that has never happened.

Which of those weeks can a buyer actually move?

One distinction changes how a transaction is planned, and most timetables do not draw it at all.

Controlled timeTime on a plan that responds to effort: putting more people, money or hours against the work genuinely finishes it sooner. is time that responds to resource. Confirmatory diligence is controlled time: put four more reviewers into the data room and the reading genuinely finishes sooner. Documentation is controlled time: put another drafting session in on Saturday and the mark-up comes back faster. Controlled time behaves the way work behaves everywhere else, so every manager has an instinct for it. More hands, less elapsed time, up to the point where the hands start colliding.

Uncontrolled timeTime on a plan that does not respond to effort, because the work is being done by somebody outside the transaction who sets their own pace. does not respond to resource at all. The work in it is not being done by anybody inside the transaction. Three of the four things in this transaction's conditions period behave that way, and they are worth taking one at a time.

A regulatory approval proceeds at the pace of the authority granting it. There is real work for the buyer here, and it is front loaded: preparing the application properly, answering questions quickly, creating no delays of its own. But the deciding is somebody else's, on somebody else's schedule, and no amount of resource inside Harivansh Packaging Limited alters that. How long an approval takes is a matter for the authority granting it.

A counterparty consentPermission from an outside party, often a customer or supplier, that a contract may pass to the new holder. That party decides in its own interest and on its own timetable. proceeds at the pace of a business that has its own interests, its own board calendar and no particular urgency about the buyer's transaction. Two of Sundarban Polymers Private Limited's contracts change hands on this purchase and each of those counterparties has to agree. Each counterparty will get to it when it gets to it, and one that sees an opportunity in taking its time may take rather more of it.

A material adverse changeA condition satisfied by the absence of some serious deterioration in the target between signing and completion. Nothing has to be done to satisfy it; it is a state of the world. condition is not a task at all but a state of the world that has to hold. Nobody works at it, nobody can accelerate it, and the only thing that resource can do is watch. Adding people to a condition that is satisfied by nothing happening is a category error, and it is made more often than might be expected by people who are simply trying to help.

Adding people shortens one kind of time and does absolutely nothing to the other, so the first question on a late transaction is never how many more people but which part is late. Getting that question the wrong way round spends real money making the controlled weeks faster while the transaction sits exactly where it was.

Put the two together on this purchase. Roughly 40.9 per cent of the elapsed span from term sheet to completion sat in the part that nobody inside the transaction could accelerate by working harder. Not because the team was slow. Because that is where those weeks live. And say the necessary sentence in the same breath: twenty two and nine are this invented transaction's own elapsed weeks, and the 40.9 per cent is a fact about them and about nothing else in the world.

The household version is immediate. A no objection letter from the housing society is needed before the sale of a flat can be registered. Every paper can be ready by Tuesday. The society's committee meets when it meets. Hiring a second lawyer does not move the committee's calendar by a single day, and everyone who has stood in a queue for a licence renewal knows this in their body even if they forget it the moment it is drawn as a bar on a slide.

Where do the pressure points sit?

A team asked where a transaction is likely to go wrong usually hands back a list of tasks: diligence is hard, the drafting is long, the approval is uncertain. Every one of those is true and none of them is where the trouble actually shows up.

A pressure pointA place on a plan where two pieces of work meet and have to hand something over. The strain shows up at the handover rather than inside either piece of work. sits on a join. A join is the place where one workstream has to hand something to another, or where two of them want the same thing at the same hour. On this transaction there are three worth naming.

The first join is diligence into drafting. The agreement is being written while the facts it is written against are still arriving. The overlap is not a failure of sequencing. Waiting for diligence to be entirely finished before a word is drafted would add the whole of one to the whole of the other, so the overlap is the only way a transaction can be run in thirteen weeks rather than twenty six. The price of the overlap is that a finding landing in week nine can send the drafters back over ground they thought was settled, and the cost of that rework is invisible on any plan that shows the two workstreams as separate rows.

The second join is drafting against the conditions workstream. Preparing an approval application and chasing two counterparty consents are not clerical jobs. The application and the consents need the same senior people who are sitting in the drafting sessions, and they need them in the same weeks. A plan that shows both proceeding happily in parallel has quietly assumed two of everybody. The second join produces the most familiar complaint on any transaction: everything is urgent, and one person has to be in three rooms.

The third join is signing itself. At that moment the work stops being something the buyer does and becomes something the buyer waits for. Nothing about the team changes and everything about the nature of the remaining weeks does. Teams that have run hard for thirteen weeks often keep running for the first fortnight after signing, generating activity against conditions that do not respond to activity, and then discover that the useful thing was to have prepared the application properly before signing rather than to chase it afterwards.

Pressure points sit on joins rather than inside tasks, and that is exactly why a plan that lists tasks never shows where the transaction will actually hurt. Finding them on any timeline means reading not the rows but the places where the rows touch.

One related idea is worth having to hand, the critical pathThe chain of work where any delay pushes the finish date out, because nothing in that chain has spare time to absorb it.. Work that has schedule floatSpare time around a piece of work, so that it can start late or run long without moving anything that follows it. can run long without moving anything that follows it; work on the critical path cannot. Completion is defined as the moment the last condition is satisfied, so on this purchase the conditions period is on the critical path by construction. Nothing else can absorb a week of it.

What moves when a milestone slips?

A slipA milestone reached later than the plan said. The word describes the movement of a date, not any judgement about whose fault it was. is a milestone reached later than the plan said. Everybody knows that a slip moves the end date. Almost nobody works out what else it moves, and on a transaction there are two entirely different animals wearing the same word.

A slip before signing moves signing and everything after it. The thirteen controlled weeks become fourteen or sixteen, and because the conditions period runs from signing, the whole nine weeks shifts to the right with it. Completion moves by exactly the amount of the slip.

Signing has already happened, so a slip inside the conditions period cannot move it. The papers are executed and both parties are bound. So the nine weeks becomes ten or twelve and completion moves, but the thirteen weeks in front of it are fixed forever.

A prediction made before the arithmetic is worth more than reading the answer.

Work it. The conditions period goes from nine weeks to twelve. The span from term sheet to completion goes from twenty two weeks to twenty five. The share is now twelve over twenty five, or 48.0 per cent, against 40.9 per cent before.

Both the numerator and the denominator moved, so a slip in uncontrolled time does not merely lengthen a transaction, it changes the shape of it. The transaction is now one where nearly half of the elapsed span sat after signing, when walking away is no longer easy for either side, against two fifths before. A reader tracking only the completion date sees three weeks of delay. A reader tracking the shape sees that the balance of the transaction between doing and waiting has shifted by seven points, and that is a different kind of transaction to be running.

Now put the two kinds of slip side by side, and hold the completion week constant so that the comparison is fair. A three week slip before signing takes signing to week sixteen and completion to week twenty five, with the conditions period still nine weeks. Nine over twenty five is 36.0 per cent. A three week slip inside the conditions period takes completion to the same week twenty five with signing still at week thirteen, and twelve over twenty five is 48.0 per cent.

Identical delay. Identical completion week. Twelve percentage points apart on the measure that matters, and in opposite directions from where the transaction started.

The two slips are not comparable as better and worse. The first delays a transaction that either side can still walk away from; the second extends a period in which both are bound and only the long-stop date sets any limit at all. Which of those is preferable is a question about position, not about arithmetic. A buyer who has become nervous about the target would far rather be in the first. A seller who wants the money would far rather be past signing. Nobody can rank them without saying whose side they are standing on.

How is a timeline rebased without losing the plot?

RebasingResetting the dates on a plan to what is now believed, rather than continuing to work to dates everyone knows are gone. is what happens when a date has been missed and there is no honest way to hold the ones that follow it. Rebasing is three steps and none of them is optional.

The date moves to what is now believed. The reason for the move is said, in one sentence, at the time it moves. A milestone almost never travels alone, so what else the move touched is checked as well.

The middle step is the one that gets skipped, and the middle step is the one that carries all the value. Recording the reason at the moment of the move costs about thirty seconds. Reconstructing the reason four months later costs an afternoon and produces something nobody trusts. Everyone involved now knows how the transaction turned out, and remembers accordingly.

Here is the rebase history of this purchase, laid out as it would sit in the transaction team's own file. The history is illustrative rather than locked. Every row reconciles to the two spans that are locked, and that reconciliation is the test any rebase log has to pass.

Read the rows and check them yourself. Eleven plus eight is nineteen. Thirteen plus eight is twenty one. Thirteen plus nine is twenty two, and twenty two is where the transaction actually landed. The total slip was three weeks, two of them in controlled time and one in uncontrolled time, and the two moves had entirely different characters even though they read the same way in a status report.

Now the part that repays a second look. The conditions share was eight over nineteen in version one, or 42.1 per cent. In version two it was eight over twenty one, or 38.1 per cent. In version three it was nine over twenty two, or 40.9 per cent. The share fell and then rose, and the conditions period never once got shorter. The controlled part grew in version two and the denominator grew with it, so the share fell. A share can move because of something that happened to the other part entirely, and that is the second reason a percentage on a timeline is unreadable without its base and both its components.

A timeline rebased four times with no reasons attached cannot be learned from, and afterwards it is indistinguishable from a transaction that was never in control. That second half is the sharper cost. The learning is lost, and that is a shame. But the record is also gone, and there is a difference between a transaction that moved three times for three specific and defensible causes and a transaction that simply drifted. Without reasons written at the time, nobody can tell which of those was being run, including the people running it.

What does a timeline built backwards from a wanted date do?

Somebody wants completion by a particular date. There is a board meeting there, or a financial year end, or a window in which the announcement would land well. So the timeline is drawn from the right hand edge inwards: put completion at the wanted date, then space the milestones backwards until they reach today.

The method feels responsible. The plan has a deadline, it has milestones, and every week is accounted for. The plan is also wrong on the first day, and wrong in a very particular place.

Watch the mechanism. Working backwards, the planner starts with the parts their own people do and can therefore size honestly: diligence and drafting. On this purchase that is thirteen weeks and the team knows it. Whatever is left over after the wanted date has absorbed those thirteen weeks is then handed to the conditions period, not as an estimate of anything but as the remainder of a subtraction.

Suppose the wanted completion is week eighteen. Thirteen weeks of controlled work leaves five. So the plan gives the regulatory approval, the two counterparty consents and the material adverse change condition five weeks between them, or 27.8 per cent of the plan's span. On the transaction that actually ran, those same conditions took nine weeks and were 40.9 per cent of the span. The plan allotted 55.6 per cent of the time the work would need, and it did so without anybody making a judgement about the approval at all.

Nobody involved believed the approval would be faster, and that is the strange, important part. Ask any of them privately and they would have said the same thing. The plan simply had nowhere else to put the weeks, so they went where there was space rather than where they were needed.

A timeline built backwards compresses precisely the time that nobody in the transaction can compress, so it is wrong from the first day in the one place where effort cannot rescue it. The controlled weeks then run more or less to plan, which makes the plan look credible for three months, and the transaction stops dead in the uncontrolled ones.

Completion arrives at week twenty two, four weeks after the promised eighteen. And in the report that goes up, the four weeks are described as a delay in approvals.

How does a lender, an analyst or a board actually read a timeline?

Four readers, four different first questions, and none of them is the expected one.

A lender committing Rs 1,000 crore of new borrowing against this purchase does not read the timeline for the completion date. The lender reads the timeline for how long its own commitment has to stay alive. A commitment sits inside paperwork that has its own expiry, and a slip eats into that. So a lender's first question is what fraction of the remaining span is uncontrolled. In that part a borrower's promise to try harder is worth nothing. On this purchase 40.9 per cent of the span from term sheet to completion sat there.

An analyst covering Harivansh Packaging Limited reads it as a question about which reporting period the target's earnings first appear in. A four week slip that crosses a period end moves a chunk of revenue and profit from one set of results into the next, and every model built on the earlier assumption is now wrong in a way that has nothing to do with the businesses at all. Analysts therefore ask, tediously and correctly, whether completion is expected before or after a period end rather than on which day it is expected.

A board member reads it for one thing: which of the remaining weeks are ours. The question sorts a transaction report into things the board can affect by asking for more resource and things it cannot affect by asking for anything. A board that does not draw that line spends its time applying pressure at the one point in the transaction where pressure has no purchase. The pressure is uncomfortable for everybody and changes nothing.

Ashwin Rege, running the transaction, reads it for the joins. He is not looking at whether diligence will finish. He is looking at the week when the drafting sessions and the consent chasing will want the same three people, the week where the plan silently assumed two of everybody.

The household version of all four is the same reflex. When a relative says the house will be ready by Diwali, the useful question is not whether they are working hard. The question worth asking is which of the remaining jobs depend on somebody outside that household, and how many of them there are. Their effort cannot reach that part of the answer.