What Limited Liability Actually Excludes

A liability does not disappear because of how an enterprise is organised. It is allocated — and somebody is standing underneath it when it lands.

Ask a leadership team what the group's entity structure is for and the answers arrive readily: tax efficiency, asset protection, a clean line between the trading business and the shareholders, an acquisition that came with its own company and was never merged. Ask the narrower question — which exposures does this structure exclude, and which does it leave exactly where they were — and the room goes quiet. Someone offers to check with the accountant.

The quiet is the subject. Entity architecture is one of the few enterprise-level decisions made once, by an external adviser working from the facts of that year, then left alone while the business changes beyond recognition. It is reviewed rarely, and usually for the first time after it has already failed.

Limited liability does not extinguish an obligation. It allocates one. When something goes wrong the cost lands somewhere: on an operating entity, on a parent, on directors personally, on an insurer, on a counterparty who accepted a contractual risk, or — where none of those can pay — on the person who suffered the harm. Structure determines the order in which those parties are reached and how far the reach extends. That is valuable, and it is not the same thing as protection.

ERANORTH is not a law firm and does not provide legal, tax or financial advice. This article addresses the general proposition only. It asserts no rule of Australian law, and every point that would require legal grounding needs verification by a qualified practitioner against the enterprise's own circumstances before it is relied upon [FACT CHECK REQUIRED].

The Strategic Context

Groups accumulate entities the way houses accumulate keys. One for a joint venture, one for a new jurisdiction, one because a financier wanted an asset ring-fenced, one that arrived inside an acquisition and was kept because dissolving it looked like effort with no reward. Each addition is approved by somebody; none triggers a review of the whole. The register is then maintained by the company secretary as a compliance artefact — filings, officers, dates — and rarely presented to the board as what it also is: a map of where consequences are meant to stop.

This matters commercially, not only legally. Structure influences the cost of capital, because financiers price the recourse available to them; it influences insurability, because underwriters ask which entity does what; and it determines whether a business line can be sold or closed without disturbing the rest. A structure that has drifted rarely announces itself — it shows up as friction in negotiations nobody traces back to its cause.

What the Structure Is Assumed to Do

Three assumptions do most of the damage, and all three are held sincerely.

The first is that the corporate veil is a wall. It is better understood as a default position — the starting point from which liability is traced — which a range of ordinary commercial and statutory mechanisms commonly displace. Obligations voluntarily assumed by signature sit outside it. Duties attaching to a person rather than an entity sit outside it. Statutory regimes designed to reach individuals precisely because the entity would otherwise end the inquiry sit outside it [FACT CHECK REQUIRED].

The second is that structure and insurance do the same job. Structure decides who is liable; insurance decides who funds the payment. An enterprise can be well structured and entirely unfunded.

The third is that because a structure was established by professionals, it is currently correct. Advice is accurate as at its date, on the facts it was given, for the purpose it was asked about. Nothing that has happened since updates it.

Reframing the Issue

Replace "are we protected?" with two better questions.

Who is protected, from what class of claim, and by which feature of the structure? A team that can answer this entity by entity has a risk-allocation model. One that answers with the name of the holding company has a diagram.

Which of those protections has the enterprise since traded away, and what did it receive in return? This is the more important question, and the more uncomfortable one, because it is answered by signatures rather than by structure charts. Every guarantee and indemnity given since formation moves exposure back across a boundary the structure was built to draw.

The Guarantee Ledger Nobody Maintains

Across a decade a growing business signs a great deal: leases with parent or personal guarantees, facilities with cross-guarantees between group members, equipment finance, supplier trade accounts, performance security on customer contracts, indemnities buried in service agreements. Each is a rational trade — better pricing, a contract that would otherwise have gone elsewhere — and each transfers a slice of exposure back over the line.

The structure chart in the board pack shows the architecture as designed. A consolidated schedule of guarantees and cross-obligations would show it as it operates. Where the two disagree, the second is the accurate description of the enterprise.

Consider a deliberately hypothetical case. A group with four trading entities presents a clean structure to its board. Asked to list every live guarantee, the finance director finds eleven instruments, three of them given by the holding company for a business the group sold two years earlier and never obtained a release for. That is not a legal failure discovered late. It is a governance question nobody asked.

The remedy is unglamorous and cheap: an obligations register kept alongside the entity register — who signed, for which entity, in favour of whom, capped at what, triggered by what, expiring when, released or not.

Obligations That Attach to a Person Rather Than a Vehicle

A second category sits outside the structure by design. Duties owed by directors, exposure arising from continuing to trade while insolvent, and statutory penalty regimes in domains such as work health and safety, environmental protection and tax administration commonly attach to individuals as well as entities. Where that is so it is generally deliberate: a regime aimed only at a company can be defeated by allowing the company to fail. The scope, defences and operation of any such regime are jurisdiction-specific and must be confirmed by a practitioner rather than inferred from general commentary [FACT CHECK REQUIRED].

The governance implication does not depend on the detail. Exposure of this kind is not a function of shareholding and is not bounded by the balance sheet of the entity concerned. It follows conduct, knowledge and the condition of the company at the time — so it concentrates exactly where a board is under most pressure: trading through distress while a refinancing is pursued, integrating an acquisition, responding to a serious incident. Which makes appointment to a dormant subsidiary as an administrative convenience something other than a nominal act, whatever the person accepting it assumes.

Structure and Insurance Fail in Opposite Directions

Over-reliance on either produces a symmetrical error.

Rely on structure where insurance was the right instrument and the enterprise discovers that the liable entity cannot pay. The obligation has not been avoided; it has been exported — to a customer, an employee, a supplier, a community. Those consequences arrive faster than the legal ones, and counterparties who watch a group let an obligation land on someone else adjust their terms with every other part of it.

Rely on insurance where structure was the right instrument and the enterprise discovers what the policy actually describes. Cover carries limits, exclusions, notification conditions and an insurer's own solvency. It responds to the wording, not to the assumption in the boardroom — and structural change compounds this as named-insured schedules go stale.

A thinly capitalised entity is also a disclosure rather than a shield. Experienced counterparties read the structure, price the recourse and ask for the parent guarantee — which is how the guarantee ledger grows, one reasonable concession at a time. The same logic applies to risk pushed outward by contract: a transfer to a supplier is worth only that supplier's capacity to pay, and some things handed across the boundary cannot be recovered by any clause once they have gone [Related article: What Did You Hand Your Contract Manufacturer That You Cannot Get Back?].

The Architecture Was Designed for a Business You No Longer Run

Structures are built for a business as it was. The triggers that should force review are ordinary growth events: entering a new jurisdiction, adding a materially different channel, taking on external debt, completing an acquisition, moving from selling products to selling outcomes, or accumulating a new class of asset such as data or intellectual property.

Two exposures show how far the drift can run. The first concerns assets the enterprise never owned. Access to customers held through a platform, a marketplace or an intermediary is a dependency rather than an asset, and no structure has anything to say about it; the risk is loss of access, not a claim, and it is answered by ownership of the relationship [Related article: Do You Own Your Route to the Customer, or Rent It?]. The second concerns harm originating in a systematised decision rather than an individual act. Where a specification executes identically across a whole population, attribution concentrates on the legal person who approved it instead of diffusing across many actors — a question about evidence rather than about which entity holds the contract [Related article: Why One Machine Failure Costs More Than a Thousand Human Ones].

Decision Framework

The table is a structuring device, not legal advice; every row requires practitioner confirmation for the enterprise's own jurisdiction and facts [FACT CHECK REQUIRED].

Three governance tests sit alongside it. The purpose test: for every entity, state in one sentence why it exists, what it holds and what it is exposed to; those failing it carry cost and directorships in exchange for nothing. The signature test: who may bind the group to a guarantee, at what value, and where is the record. The clean failure test: take the entity most likely to fail and walk the consequences through the group — cross-defaults, shared services, shared insurance, change-of-control clauses. If it cannot fail without taking others with it, the separation is nominal, and the board should learn that before a lender does.

From Strategy to Execution

The immediate work needs no external spend: compile the obligations register, write the one-sentence purpose statement for each entity, and reconcile the schedule of named insureds against the current group. Most enterprises can finish all three within a quarter, and the exercise surfaces much of what a legal review would later charge to find.

The medium-term work is to make entity architecture a governed decision rather than a historical artefact — a named owner, usually the chief financial officer or general counsel; review triggers tied to transactions and market entries; and a delegation of authority treating a guarantee as a decision requiring approval rather than a routine condition of trading. A structural consequence assessment inside the approval paper for acquisitions and financings puts the question while the deal is still negotiable.

The long-term positioning is structural optionality: the ability to separate, sell, close or ring-fence a business without unpicking a decade of cross-obligations. Enterprises that can exit cleanly can also enter more boldly, because the downside of a new venture is bounded in fact rather than in theory.

Signals to Monitor

Guarantees granted as a routine condition of ordinary contracts, with no escalation and no register. An entity list containing companies nobody present can explain. Insurance renewals where the named entities no longer match the group. Financiers, landlords or major customers asking for parent guarantees more often than they used to — the market is repricing your structure and telling you so. And the sentence "we assumed that entity was separate" surfacing in a meeting, which means the assumption has already been tested by someone else.

Questions for the Leadership Team

1. For each entity we own, can we state in one sentence why it exists, what it holds and what it is exposed to?
2. Can we produce this week a complete schedule of every guarantee, indemnity and undertaking the group has given, with amounts, triggers and expiries?
3. Which protections created by our structure have we since signed away, and what did we receive for each?
4. If our most exposed entity failed tomorrow, what else would be pulled in — and who has tested that rather than assumed it?
5. Where are we relying on structure to do work only insurance can do, and the reverse?

Closing Perspective

A corporate structure is a set of promises about where consequences stop. Those promises hold only to the extent that nobody has undone them by signature, that duties attaching to people rather than vehicles are recognised as separate, and that the design still matches a business which has changed since it was drawn.

None of this argues for less structure. Well-designed entity architecture is one of the few instruments that lets an enterprise take a risk in one place without putting everything else behind it. The argument is that it is a live governance decision, not a completed piece of professional work filed after formation.

The choice available to a board is narrow. It can find out what its structure excludes deliberately, at a time of its own choosing, for the cost of a few days of work and a practitioner's fee. Or it can find out at the moment of failure, when the answer is written by a counterparty, a regulator or a liquidator — and the only question left is who was standing underneath.

About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.

Copy link Print