The cost stays. The work has gone.

What stranded costs actually are.

A stranded cost is a cost that survives the event that gave it purpose. In a carve-out, the event is the separation of the business unit from its parent. The parent retains people, licences, contracts, leases, and overheads that supported the divested business but cannot be quickly reallocated. The divested business often inherits costs sized for the parent's operating footprint, not for Newco's. Both sides feel the drag, and both sides usually underestimate the volume.

Stranded costs are different from one time separation costs. Separation costs are the project spend to execute the carve-out. Stranded costs are the recurring run rate that remains after the project is done. A poorly run carve-out can leave 5 to 15 percent of the divested business cost base stranded for 12 to 24 months. For Newco that translates directly into EBITDA drag during the value creation window. For the seller it shows up as parent dilution.

The investment thesis usually counts on a clean operating cost base. The diligence model often assumes that the carve-out lands at the run rate implied by the carve-out income statement. That assumption is wrong unless the carve-out is actively managed for stranded cost removal. The gap between the assumption and the operating reality is where value creation plans miss their numbers.

The TSA itself contributes. Every service the seller continues to provide under TSA has a price that includes the seller's allocations, overheads, and mark-up. Where the divested business previously absorbed those costs through shared services at parent rates, Newco now sees the cost on an invoice with a different number. Programs that fail to compare the TSA cost to a standalone alternative cannot identify which TSA services are stranded.

Where the costs hide on the Newco side.

The Newco stranded cost map starts with people. Carve-outs that take a percentage of a corporate function rather than a clean unit often inherit roles that were sized for the parent's scale. A two person tax team for a business with one third the entity count, a five person treasury team handling two cash accounts, a procurement team supporting categories the divested business does not buy. Each of these is a stranded cost until the role mix is right sized.

Software licences are the second category. Volume agreements at the parent often produced low unit prices that Newco cannot replicate at standalone scale. The licence count assumed at separation may exceed the actual user base. Modules that the divested business never used remain on the contract. Programs that inherit licences without rationalising them carry stranded software cost into year two. The IT context is in the carve-out IT separation playbook.

Real estate and facilities follow the same pattern. Square footage sized for parent occupancy, leases that match parent terms, facilities services contracts that fit the parent footprint. The divested business often needs a fraction of that space. Surrendering, subletting, or renegotiating leases takes 6 to 18 months. Programs that delay the property review carry full lease cost while the space sits underused.

Third-party contracts are the quiet leak. Insurance policies, audit fees, professional services retainers, telecommunications, software maintenance, and many vendor contracts have minimum commitments tied to parent scale. Each contract has a renewal or notice date. Programs that build the contract calendar at signing and renegotiate at each renewal date capture the stranded cost. Programs that wait for the contracts to surface lose the optimisation window.

Where the costs hide on the seller side.

The seller carries its own stranded costs after a carve-out. Corporate functions sized for the larger group continue at the old run rate while the revenue base shrinks. Allocated costs that the divested business used to absorb now sit in the parent unallocated bucket. Shared service centres that ran at scale for the combined group run at a higher unit cost for the smaller group. The seller stranded cost number is often disclosed in the deal financials but rarely tracked through remediation.

The TSA can absorb seller stranded cost temporarily. While the seller provides services to Newco, the cost of running those services is partially covered by the TSA fee. That cover ends when the TSA exits. Sellers who plan only for the day the TSA ends find their stranded cost surfaces as a step change in operating cost. Programs that sequence stranded cost removal alongside the TSA exit plan land softer.

From the buyer perspective, the seller's stranded cost matters when the TSA is on cost-plus pricing. The seller has an incentive to load TSA pricing with stranded allocations that should not be charged to Newco. The TSA exit roadmap is the lever to control that exposure. Slow exits keep the cost base elevated; fast exits force the seller to take ownership of its remediation. The exit context is in the TSA exit timeline explained.

Reverse TSAs invert the dynamic. Where Newco provides services back to the seller, the seller's stranded cost effectively becomes the seller's choice to keep buying from Newco. The reverse TSA fee can subsidise Newco's run cost during the transition. The structure works only if the reverse TSA is properly priced and properly bounded.

The Day One stranded cost plan that actually works.

The plan starts with a baseline. Within 30 days of close, Newco needs a documented stranded cost baseline that compares actual run rate to the target operating model. The baseline covers people, software, real estate, third-party contracts, and TSA services. Every line has a target run rate, a remediation date, and an owner. Programs without a baseline cannot prove progress and cannot defend the value creation plan.

The plan needs a single accountable owner. Stranded cost removal is a programme, not a project, and the work cuts across IT, HR, finance, procurement, facilities, and operations. A single owner at the CFO or COO level keeps the workstreams synchronised, escalates blockers, and reports run rate progress monthly to the operating committee. Programs that distribute ownership across functions miss the cross category trade offs.

Sequencing matters. Some stranded costs can be removed only after a system migration. Some can be removed only after a notice period. Some can be removed only after the TSA exits. The dependency graph for stranded cost removal needs to map onto the TSA exit calendar and the technology roadmap. Programs that try to remove cost in the wrong order create service gaps that force temporary cost back in. The TSA exit context is in TSA exit cost benchmarks.

The reporting cadence keeps the plan honest. A monthly stranded cost report that shows baseline, actual, target, and forecast keeps the operating committee focused. The report needs to call out new stranded costs as they emerge, since the operating model evolves through the first year and new stranded patterns appear in every quarter. The discipline is in the reporting, not the analysis.

When stranded costs become permanent.

Stranded costs harden when the remediation window closes. Renewal dates pass. Lease terms reset. Headcount becomes culturally embedded. Software contracts auto extend. Programs that miss the first 12 months of remediation typically carry the stranded cost for the contract life, which can extend three to five years. The cost of inaction compounds and the budget benefit becomes uncatchable.

Cultural factors are the silent driver. Functions that feel busy in the first quarter rarely volunteer the case for headcount reduction. Vendors with long relationships rarely volunteer price reductions. Property managers rarely volunteer to renegotiate. The remediation work needs an external pressure source to move at the required pace. The pressure source is usually the operating partner or the CFO.

Measurement traps make the problem worse. Many finance teams measure stranded cost only by comparing year over year run rate. That measure misses the inflation effect, the new system cost, and the new compliance cost that all pile in. The honest measure is run rate against a clean standalone benchmark, refreshed quarterly. Programs that hold to that benchmark see the gap and close it. Programs that drift to softer measures gradually accept the stranded cost as baseline.

The buyer pays twice if the stranded costs harden. Once in the year one EBITDA drag. Again in the exit multiple, since acquirers discount businesses with elevated cost bases relative to peers. The cost of failing to remediate stranded costs is therefore double counted in the holding period return. The Day One readiness context is in carve-out Day One readiness.

More on carve-out execution.