Worked structure ten · Structural Risk Review
A change of control the chain does not survive.
The transaction moves an interest somewhere near the top of a structure. Beneath it sit contracts signed years earlier by other people for other reasons, and several of them characterise that movement as a change of control. No two of the definitions are the same, and not one of them is negotiable once the money is drawn.
The critical path is the order of the steps, because the order decides whether the trigger fires at all.
Worked structures · fifteen rooms
- 01An existing platform
- 02An ineligible asset
- 03An investor class
- 04A narrow exit
- 05A repeatable structure
- 06A co-investment
- 07A continuation
- 08Unavailable security
- 09The seed terms
- 10A change of control
- 11Drawn elsewhere
- 12An in-kind distribution
- 13A change of domicile
- 14An investor's perimeter
- 15A strategy, no vehicle
01 · The transaction as it arrives
Nobody in the room intends a change of control. Five documents record one anyway.
Four arrivals. One is a column in a table written before the chain exists, one is a renewal that could have removed the provision and put it back unread, one is a signed transaction with a condition nobody has priced, and the fourth is the disguise the other three travel in.
The rarest of the four, and the only arrival where the definitions can still be met by design rather than by consent. Whether a future movement of the interest is read as a change of control is a column in a table here, and it costs a column.
The one moment when the triggering provision is on the table anyway and the counterparty holds nothing it can withhold. It is treated as a renewal of terms already agreed, and the provision goes back in unread because it went in last time.
The transaction exists, the counterparty has been told, and the question is no longer whether a consent is needed but what it costs. What is wanted here is not design. It is a statement of what each outstanding consent can do to the transaction, including the ones that can end it.
Almost never "does this trigger anything". It arrives as a lender waiver to be obtained, a landlord consent that has been sitting for weeks, or a single unclosed line in a diligence list. The structural question sits behind whichever of those came through the door, and it is the same question in all four.
The transaction in this class is not exotic. It is the ordinary movement of an interest inside a private-market structure: a stake sold at the top, a new investor admitted, two vehicles collapsed into one, a manager reorganised for succession. None of those events is about the asset, and every one is read by documents beneath it, drafted by people who will never hear the transaction described.
No manager, counterparty, asset, value or date appears in this room. What transfers from one transaction to the next is the shape of the problem, and the shape is what is set down here.
02 · The structural problem
One transaction, and six documents decide separately what it was.
A change of control is not something a transaction does. It is a characterisation applied to it afterwards, by contracts drafted at different times by different parties to protect different things. There is no reason for their definitions to agree, and they do not.
The provisions are not one provision. A structure of any age carries a facility, a licence attaching to what the asset does, a lease or concession under it, a joint venture agreement beside it, and long contracts that carry its revenue. Each holds its own change-of-control clause, written to protect the thing its author cared about.
A facility defines control to protect a credit judgement about who stands behind the borrower. A licence defines it to protect a public judgement about who is fit to hold a permission. A lease protects a counterparty's choice of who occupies its ground, a joint venture agreement protects a partner's right to choose its partner, and a revenue contract protects a customer who committed on the strength of who was on the other side. Five purposes, and no committee ever sat that would have reconciled them.
Where the definitions come apart
They come apart at the level. Some clauses measure control at the direct holder of the shares, some look through to ultimate beneficial ownership, and some attach to the manager rather than to the vehicle, so a transaction at fund level trips nothing and a reorganisation inside the manager trips everything. A structure built to stay above one clause's reach is frequently sitting inside another's.
They come apart at the threshold. Control is a percentage in one document, the power to appoint a majority of a board in the next, and the ability to direct the affairs of a person in the third, which is a test rather than a number.
They come apart at the clock. Some provisions fire on completion. Some fire on the agreement to complete, which means signing rather than closing. Some fire on the notice, and a notice cannot be recalled. Some fire on a public announcement, which the transaction team may not think of as a step at all.
Nobody inside the structure holds all five documents. The facility sits with the finance team, the licence with the operating company, the lease with whoever manages the asset, the joint venture agreement with counsel from an old acquisition, and the revenue contract with a commercial team that has never read the fund's constitution. Each counterparty holds its own clause and reads the transaction the moment it is told.
The transaction is one event. It is characterised five times, by five documents, and not one of the five was drafted by anybody now in the room.
03 · What binds
Five of the six can be bought. The sixth is not for sale and it is asked last.
Six. Five were signed by the manager or its predecessors when agreement was the cheap answer. The sixth was signed by nobody: it is a condition of a permission rather than a term of a contract.
The facility, under which a change of control is a prepayment event or an event of default answered on the lender's timetable. The security package, under which shares in a charged entity cannot move until a chargee releases them. The permission attaching to what the asset does, which requires a change in the person controlling the holder to be approved before it happens. The lease, ground interest or concession beneath the asset, carrying a consent requirement and sometimes a termination right. The joint venture agreement, carrying pre-emption, drag and put rights that decide who ends up owning the asset. And the long revenue contract, carrying the line the underwriting rests on and a right to walk away from it.
Each one, by source and by what it does when it fires
Signed at the financing on the lender's paper, with a definition drawn to catch movement anywhere above the borrower. It fires as money. The remedy is a waiver, priced by a party already holding a signed transaction.
The share charges granted alongside the facility. This one wants a release rather than a consent, and a release is an act rather than an opinion. Until it is given the shares cannot move at all.
Not a contract. A condition attaching to a licence or approval held by the entity that operates the asset, requiring advance approval of a change in its controller. The one constraint here no party can move, and the longest clock.
Granted by a party with its own reasons, usually before the manager was involved. Consent not to be unreasonably withheld is a standard enforced by a process rather than a phone call. Where the right is termination instead, the asset is at stake.
Signed at the acquisition, when the partner was being persuaded rather than asked. Pre-emption, drag and put rights do not price the transaction. They redirect it, on terms fixed years ago.
The offtake, the long supply arrangement, the operating contract. Its clause is short and it terminates. Where the model rests on that contract, this is not a consent problem. It is the valuation.
Five of these produce a price, and a price can be negotiated by somebody who knows it before the other side does. The sixth produces a lapse, and a lapse is not negotiated at all.
04 · The architectures considered
Four ways to move control, and the order decides which of them is still a choice.
Four architectures, and they are not alternatives on a single date. Three exist only before a particular step has been taken, and that step is quieter and cheaper than any of them.
Transfer and then ask, which is the default and the most expensive. Sign conditional on every consent, which is honest and tells every counterparty at once. Move the interest at a level none of the definitions reaches, which works only where the definitions were drawn narrowly and is what a look-through clause exists to prevent. Or remove the trigger before the transaction exists, by replacing the contract that carries it while the counterparty has nothing to withhold.
The fourth route disappears without anybody noticing, because it depends on the transaction having been foreseen at a moment when there was no transaction. A facility refinanced in the ordinary course, a revenue contract renewed on schedule, a lease regeared for its own reasons: each is an occasion on which a change-of-control clause can be narrowed for the cost of asking, and each passes unused.
What each route reaches, and what closes it
Reaches completion fastest and needs no design. Closed by nothing, which is why it is chosen. Every counterparty is then approached as the holder of a right over a transaction that has already happened, and each prices accordingly, learning from the one before.
Reaches a completion clean on every document, and keeps the transaction reversible until the last consent lands. Closed by time, and by the fact that a condition tells every counterparty the transaction needs it. One discretionary consent then holds the whole timetable.
Reaches a change of economic ownership without meeting any test, where the tests are drawn at the direct holder. Closed by a look-through definition, a beneficial ownership test, any clause attaching to the manager, or an anti-avoidance provision. Read on the documents as they stand, never assumed.
Reaches the cleanest position of the four, because there is nothing to consent to. Closed by the passing of the occasion: a refinancing completed last year on the old wording, a contract renewed unread. The only route negotiated with a party that does not yet know it holds anything.
The same problem drawn as a sequence. The first four steps are taken inside the house and can be taken again on different facts. The fifth cannot.
- 01 The definitions read Every document in and beside the chain read for its own definition of control: level, threshold, remedy, clock. Nothing has left the building.
- 02 The trigger set drawn Which of those definitions the transaction as proposed actually meets, and at which level. Where the answer is none, there is no consent to ask for.
- 03 The transaction reshaped The step where the order still pays. What moves, at which level, in how many parts, in what sequence. Two shapes reach the same outcome, and one of them meets fewer definitions.
- 04 The removable triggers removed Contracts that can be replaced, refinanced or regeared before the transaction is visible are dealt with here, against a counterparty holding nothing. Later, the same conversation has a different price.
- 05 The consents requested One round, one description, every remaining counterparty at once. The first request cannot be withdrawn and made again on better facts, and each party asked now knows it holds something the transaction needs. One-way
- 06 Signing, completion, notice Whichever of the three the documents actually name is the moment their clauses fire. Read at step one it is a condition. Discovered here it is a breach with a date already on it. One-way
Steps one to four cost paper and can be taken again on different facts. Step five cannot: the first approach tells a counterparty that the transaction needs it, and that knowledge is not returnable. The first four steps exist to shorten the list that reaches the fifth, and the shortening has to happen before anybody outside is told.
05 · The critical dependencies
Three, and all three have to be true before anybody outside is asked anything.
Each is a condition of the order rather than a task inside it. A dependency here is not something that has to be finished. It has to be true on the day the first approach is made, because that approach cannot be taken again.
That the trigger set is complete. Owned by counsel, read from the documents themselves rather than from anybody's memory of them, and drawn against the transaction as proposed rather than a description of it. A missed provision is the same problem, found after the leverage has gone, by a counterparty who found it first.
That the transaction has one description, and every counterparty receives that one. Owned by the manager. Requests describing the same transaction differently produce a comparison the manager does not control, and counterparties compare. Where a document requires particular information, the difference is one of detail and never of substance.
That the permission question is answered first, though it is answered last by every transaction team that has not met one before. Owned by the manager through advisers licensed to answer it. Its clock belongs to somebody outside the transaction, it does not shorten under pressure, and an unapproved change in controller is not repaired by an application made afterwards.
What each costs to hold, and what it costs to restore
The first costs a reading. Every document in and beside the chain is put against six questions: what it calls control, at which level it measures, what threshold it uses, what it does when it fires, on what event it fires, and whose signature ends it. The answers go in one table, maintained at each renewal, which is when they change.
Restoring it later is a different exercise, because the reading is then done under a timetable by people who already know what answer they need. A trigger map drawn while a transaction is live is drawn by parties who want it to be short.
The second costs a document and the discipline to use it. The third costs an early conversation with the advisers who hold the permission question, at a point where there is nothing to disclose except a possibility, which is the only point at which it is free.
A consent that was always going to be given is still worth asking for in the right order. The order is the only part of the request the manager still owns.
06 · The architecture that survives
The transaction is drawn against the definitions, not the definitions read against the transaction.
What the constraints leave standing is not a preference among the four routes. It is a rule about the order the work is done in, and a transaction shaped so that as few counterparties as possible ever have to be asked anything.
The trigger map is drawn first, from the documents, before the transaction has a shape. It is not a list of clauses. For each document it carries what that document calls control, the level at which it measures, the threshold, the remedy, the event on which it fires, and the name of whoever has to sign. Six columns, and the sixth turns the map into a plan.
The transaction is then shaped against the map. What moves, at which level, in how many parts, in which order. Where two shapes reach the same commercial outcome and one meets fewer definitions, that is the difference between a transaction with two counterparties in it and a transaction with five.
Where a trigger cannot be designed around, the contract carrying it is replaced or refinanced before the transaction is visible, on the next occasion that was going to arise anyway. Where it cannot be removed either, the consent is asked for in one round, against a completion condition, with the fallback written into the transaction documents rather than improvised when the answer comes back.
What the map does that a list does not
A list of clauses tells the manager how many parties are involved. The map tells it which clauses fire on the same event, which fire earlier, and which can be made not to fire by a change in the shape rather than by a conversation. Three clauses that all fire on completion are one problem with one date. Three that fire on signing, on notice and on announcement are three problems.
The map also states, for each trigger, what happens if the answer is no. That column is the reason the map exists. A structure whose only plan is that every counterparty says yes is not a structure. It is an expectation with completion mechanics attached.
Where none of this was done and a notice has already gone out, the design work is over and what remains is a negotiation. What we add then is a full account of what each outstanding consent can do to the transaction, so the manager knows which requests it should be prepared to lose.
A change-of-control provision is not a risk somebody put into the structure. It is a right the structure granted years ago, and the only thing still in the manager's gift is when the holder learns it has one.
07 · The implementation framework
Six steps, and five of them happen before anybody outside the house is told.
Written as dependency statements rather than as tasks, because here the sequence is not how the work is organised. The sequence is the work.
Read every document for its own definition of control. Draw the trigger map from those readings. Shape the transaction against the map. Remove what can be removed while it is still cheap to remove. Ask, once, in one round, for what is left. Complete, and serve the notices the documents require rather than the ones the transaction team assumes.
The signatures then belong elsewhere, as they should. The manager's counsel construes each provision and states what it requires and on what event. Its tax advisers take the treatment of each candidate shape, because two routes to one outcome are two transactions with two answers. The permission question goes to advisers licensed to answer it. The lender answers the change-of-control question under its own facility, and the chargee decides what it will release. Our work is the trigger map, the comparison of the four architectures, and the order in which the steps are taken.
The sequence, and what each step depends on
- 01 The documents gathered and read Depends on the facility, the security documents, the licence conditions, the lease, the joint venture agreement and the revenue contracts being in one place. Frequently the hardest step here, and it holds no counterparty.
- 02 The trigger map, drawn Depends on step one and on counsel construing rather than summarising. Six columns per document, and a seventh stating what happens if the answer is no.
- 03 The shapes compared against the map Depends on the manager having stated the outcome it wants rather than the mechanics it assumed. Each shape is scored by how many definitions it meets and which parties they bring in.
- 04 The removable triggers removed Depends on an occasion existing: a refinancing, a renewal, a regear. Where one is due it is brought forward. Where the last one passed unused, this step is unavailable.
- 05 The single round of requests Depends on all four above and on one agreed description. The one-way step. Every remaining party is asked at once, so none of them learns the shape of the transaction from somebody else's answer.
- 06 Completion, against the clocks Depends on the map naming, for each document, whether the event is signing, completion, notice or announcement. The sequence follows those four, not the closing checklist.
Nothing on this list is difficult. The list is worth something in this order and nothing in any other, because in any other it is a record of what was asked too early.
08 · What this case generalises to
Wherever a right was granted to somebody outside the room, the transaction is designed around the definition or priced by its holder.
Four families share the shape of this one. In each, a document signed for one purpose reads a later event it was never drafted against, and its holder reads it before anybody inside the structure does.
An operating business, a licensed platform, a concession, a regulated subsidiary. The permission is held by an entity nobody thinks of as part of the transaction, and its conditions reach up the chain to whoever controls it. The response is to answer that question before the shape is fixed, because it is the one constraint whose timetable belongs to somebody else.
The credit was underwritten on who stands behind the borrower, so the facility defines control broadly and deliberately. The financing documents are read at the start of the design rather than the end of it, and the next refinancing is treated as the moment to narrow the definition rather than as a rate conversation.
Joint ventures, club deals, co-investment vehicles with governance rights, an early investor whose consent rights were the price of coming first. A partner's change-of-control right does not price the transaction, it redirects it, on terms fixed when the partner was being persuaded. Know before the approach what that partner is entitled to do, because it is entitled to do it whatever anybody intended.
A stake sold in the manager, a succession, an internal reorganisation, a key-person event. These are not fund transactions and are not treated as transactions at all, which is why they are the most dangerous member of the family. Definitions attaching to the manager sit in documents nobody thinks to read, because the documents are about the asset and the event is about the house.
The engagement that answers this class is a Structural Risk Review: the trigger map, the comparison of the four architectures to the same depth, and the order in which the steps are taken. It stops where the signatures start.
Nobody in the chain intended a change of control. The chain does not read intention. It reads the definition it was given, on the date the definition was met.
Written as a type · no party, no value and no date · stated as at August 2026
Write to us.
Complex transactions fail at the interfaces between otherwise workable components. We resolve the structural complexity between investment intent and transaction execution.
Write to us