The structure you choose for a growth bet decides, in advance, whether you will ever be able to stop it.
A growth proposal usually reaches the board with the number already settled, and the discussion that follows is about whether the amount is right and the return assumptions credible. It is rarely about the question that will matter most eighteen months later: which balance sheet absorbs the loss if the bet does not work, and who holds the authority to declare that it has not.
That question is normally answered by default rather than by decision. The venture is funded from the operating business because that is where the cash sits, staffed from it because that is where the capable people are, and introduced to its customers because those relationships already exist. None of these were deliberate choices, yet together they attach the failure of the new activity to the health of the old one.
There is an alternative most executives can describe and fewer apply with discipline: keep the proven business intact, incorporate a distinct vehicle for the speculative activity, capitalise it deliberately, and admit any outside investor only to the new risk. The structure is unremarkable. What matters is what it does to the decisions you can still take once the early enthusiasm has worn off.
The Strategic Context
The proven core is not simply a source of money. It is a particular kind of asset: reasonably predictable cash conversion, a credit standing built over years, customer relationships that tolerate the occasional failure because the record is otherwise good, and a workforce that stays because the place feels stable.
A speculative growth vector has close to the opposite profile — cash conversion unknown, timing unreliable, evidence base thin, and the most likely single outcome that it does not work. That is not a criticism of growth ventures. It is the definition of one.
When both sit on the same balance sheet, the risks do not average out in any comforting way. The stable asset becomes collateral for the unstable one's learning process. Borrowing capacity built by the core is consumed by the venture. Attention the core requires to stay predictable is redirected to the thing that is interesting. Most consequentially, the venture's failure arrives as a shock to the enterprise rather than as a result from an experiment.
Enterprises are meanwhile carrying more simultaneous bets than their operating rhythm was designed to absorb, which makes this a risk allocation question rather than a company formation one.
What the Structure Debate Usually Gets Wrong
The first misreading is that this is a legal and tax question, properly delegated to advisers who return with a recommendation. Advisers are essential to the execution, but what they execute is a governance decision about reversibility, and it belongs to whoever will later have to stop the venture.
The second is that separation exists to protect the owners. Asset protection is real, but as a rationale it understates the point. The stronger function of a separate vehicle is that it makes the venture's failure survivable, and a survivable failure can be admitted. A venture funded invisibly from the core cannot fail cleanly, because its failure would take capability the core depends on. The organisation therefore will not let it fail — not through cowardice, but through correct assessment of the consequences. The stop decision is deferred past the point where stopping would have taught anyone anything.
The third runs the other way: that a separate vehicle signals half-heartedness. The reverse is closer to true. A vehicle with a defined capital base is a firmer commitment than an open-ended internal budget line, because the amount is fixed, visible and reported. Internal funding is the softer commitment precisely because nobody ever has to say the number out loud — which is also why the fourth misreading survives, the assumption that of course the enterprise would stop the venture if the evidence turned. Ask who would have to say so, to whom, and what that person would be giving up. Usually no one is positioned to make the call.
Reframing the Issue
The useful question is not whether to incorporate a company. It is: which failures does this enterprise want to survive, and which does it want to be able to declare?
Structure decides which future decisions remain available. Fund the bet from the core and its failure becomes an enterprise event. Fund it in a vehicle of its own, with a stated maximum exposure and a named authority to refuse more, and its failure becomes a portfolio event — expensive, instructive and bounded. Both are legitimate. What is not legitimate is choosing by inertia and discovering the consequences at the moment you most need alternatives.
What Separation Actually Buys
It buys reversibility. A defined capital base creates a real boundary, and stopping becomes a discrete act with a known cost rather than an argument about how much more the core can absorb. Setting that boundary requires knowing how much loss the core can carry before its economics invert [Related article: The Number Below Which the Enterprise Is Structurally Loss-Making].
It buys honest accounting. Inside a large operating business a venture can be subsidised invisibly through absorbed overhead, borrowed staff time and facilities nobody charges for. A separate reporting entity forces those transfers into view, and the venture's economics stop being a matter of opinion.
It buys a priced risk for anyone joining you. An investor admitted to the whole enterprise buys proven cash flows they did not underwrite, at a valuation distorted by a bet nobody can yet value. An investor admitted only to the vehicle buys the risk they actually assessed — a better deal for both sides and a shorter negotiation. It also lets you bring in a participant who holds something you need, a channel or a presence in a market you do not know, without giving them a claim on your core. What such a participant wants beyond a return deserves its own scrutiny [Related article: Who Is Funding Your Growth, and What Did They Claim Beyond Money?].
And it answers a documented organisational failure. Clayton Christensen argued that established organisations systematically starve opportunities unable to clear the resource-allocation tests built for the existing business: the new activity loses every internal contest to a mature line with better near-term numbers, and loses on merit each time. A separately capitalised unit answers that mechanism directly.
What Separation Costs, and Who Pays It
The costs are usually understated in the proposal. There is duplicated overhead — directors, reporting, audit, insurances, banking, systems — and a governance load falling on people already fully committed. There is the discipline of pricing every service the core provides, because a shared service never charged for is a subsidy, and an unpriced subsidy is the thing separation was meant to prevent. There is a talent cost: people must choose a side, the operators you most want are the ones the core can least spare, and secondment terms decide whether any of them move.
There is a more serious cost. If the venture's entire thesis rests on the core's customers, plant, brand or accumulated technical capability, separating it may remove the advantage it was relying on. The honest structure may then be an internal ring-fence — a hard budget, a distinct reporting line, an explicit stop authority — rather than a new entity that recreates the dependency through contracts.
Finally, the tax, accounting-consolidation and directors' duty consequences of entity separation differ by jurisdiction and by structure, and for Australian entities require verification with qualified legal and tax advisers before any structure is settled [FACT CHECK REQUIRED]. ERANORTH is not a law firm or a financial adviser, and nothing here is structuring advice.
When the Separation Is Decorative
A structure that is legally distinct can be economically inseparable, and boards are slow to notice the difference. The tell is recourse in fact rather than in form: parent guarantees, letters of comfort, the parent's name on the vehicle's proposals, its customers used as proof points, key staff on its payroll whose reputations are publicly staked on the outcome. Each can be justified individually. Together they mean the parent will rescue the vehicle — and a vehicle that will be rescued is not ring-fenced. It is an internal project wearing a company name, and you have paid for governance you do not have.
Two tests are worth applying before the first dollar moves. State in advance the maximum the parent will contribute, and name the individual authorised to refuse anything beyond it; if the answer is that the board would consider it at the time, there is no boundary. Then ask whether the vehicle could fail without the core's customers, lenders and best people being materially affected. If not, the operational separation is incomplete however clean the legal one looks.
The opposite abuse deserves naming too. Separation must not be used to move an unattractive commitment out of view; consolidation requirements and directors' duties exist partly for that reason, and any structure whose appeal is that it hides an exposure should be abandoned on that ground alone [FACT CHECK REQUIRED].
Decision Framework
Five pathways are usually available, trading capital at risk against control and access to capability.
| Pathway | Capital at risk | Reversibility | Access to core assets | Governance load |
|---|---|---|---|---|
| Fund from the operating business | Core balance sheet, uncapped in practice | Low — stopping is an enterprise event | Full | Low, but no real gate |
| Internal ring-fence, hard budget, named stop authority | Core balance sheet, capped by policy | Moderate — depends on discipline holding | Full | Moderate |
| Wholly owned vehicle, separately capitalised | Stated subscription only | High if recourse is genuinely limited | Contracted and priced | Higher |
| Vehicle with external co-investors | Shared, diluted at the vehicle level | High, but exit terms now bind you | Contracted and negotiated | Highest |
| Joint venture with a holder of a needed asset | Shared, plus contributed assets | Moderate — unwinding is slow | Reciprocal | Highest |
Four questions discriminate between them more reliably than any general preference.
Is the venture's failure mode correlated with the core's? If the same downturn takes both, separation buys less than it appears to, because you will be short of cash in both places at once.
Does the venture need the core's assets to work at all? The more it does, the more attractive a disciplined internal ring-fence looks.
Would we accept this venture's cost of capital if it were priced standalone? If not, the core has been subsidising a return the market would not fund.
Could this be funded by the customers who want it rather than by any capital source at all? Advance commitments are the cheapest capital available and the least understood, with a failure mode of their own [Related article: Can Your Customers Fund the Business — and What Breaks When Volume Falls?].
From Strategy to Execution
The immediate work is not incorporation. It is writing down the maximum exposure, the stop conditions and the person authorised to refuse further funding — before the structure is chosen, because doing it afterwards invites the structure to soften the numbers. Then draft the shared services terms as though the counterparty were a stranger. If they would embarrass you in front of an independent director, the separation is not real.
The medium-term work is capability. Running a separate reporting entity requires finance, legal and governance capacity that most operating businesses have sized for one set of books. Benefits ownership needs a name inside the vehicle, not a committee. And the core needs protection from attention drain, which is a management discipline rather than a structural one.
The long-term work is a repeatable pattern. Organisations that take growth bets regularly benefit from a standing template — standard capital structure, shared services terms, governance and stop conditions — so each venture becomes a decision about the opportunity rather than a structuring exercise. Include the terms on which a successful vehicle is reabsorbed, and agree the valuation mechanism at the start. That negotiation is far easier before anyone knows who has won it.
Signals to Monitor
Watch for secondments that quietly become permanent, for funding requests arriving outside the agreed envelope with an explanation attached, and for guarantees or informal assurances issued after the structure was set. Watch for the vehicle's proposals leaning on the parent's reference customers, which converts a bounded financial risk into an unbounded reputational one.
Watch the stop conditions themselves: if they are being renegotiated rather than triggered, the boundary has already failed. Watch the core's cash conversion and staff retention against the trend that preceded the venture. And watch for developments that change the structural logic — a co-investor's distress, a change in tax or consolidation treatment, or a market shift that makes the core's assets more valuable inside the venture than outside it.
Questions for the Leadership Team
- If this venture fails completely in two years, what does the core lose in cash, capability, customer confidence and borrowing capacity — and have we written that down?
- Who is authorised to refuse further funding, and what will refusing cost that person personally?
- What are we contributing that we have not priced, and what would an independent party charge for it?
- If we admitted an investor to the whole enterprise instead, what would we be selling that they did not ask to buy?
- Which commitments would survive if we withdrew every informal assurance the parent has given?
- If the venture succeeds, on what terms does it come back, and who agreed those terms while the outcome was uncertain?
Closing Perspective
The structural question looks administrative and is not. It is a decision about which failures the enterprise is willing to allow to happen.
A growth vehicle that cannot be permitted to fail is not a growth vehicle. It is a mechanism for converting a proven business into an unproven one, slowly, through individually reasonable funding decisions that nobody experiences as a decision. The separately capitalised entity earns its cost not because it protects assets, though it does, but because it restores to the board a capability internal funding quietly removes: the ability to look at disappointing evidence and act while acting is still cheap.
Most enterprises discover the price of that capability only in the year they needed it and found it had been structured away.
About the author
Kevin Jogin is Founder & Principal Advisor at EraNorth. Meet the Founder.