What is a sale-leaseback?

A sale-leaseback is a single transaction in two parts, executed at the same moment. The owner of an asset sells it to a buyer, and the buyer immediately leases it back to the seller for a defined term.

What changes: legal ownership, the end-of-term position, and whatever control rights the lease gives away. What does not change: possession, and the seller's ability to keep using the asset exactly as before. From the outside — inside the building, on the rack, at the turbine hall — nothing looks different the day after closing.

The reason to do it is that capital sitting inside an owned asset is doing one job. A sale-leaseback releases that capital for another one, at the cost of a periodic rent obligation and the loss of whatever the asset is worth at the end. It is most useful where the owner has more productive uses for capital than the asset itself represents — which describes most parties building at speed in a market where the constraint is time.

The transaction is not a loan, and the difference is not cosmetic. There is no borrower and no principal balance. There is a sale, and a lease with a term, a rent, and a set of end-of-term rights. Whether a particular transaction is characterised as a true sale or, in substance, as a financing is a question for counsel and the auditors, and the answer drives tax and accounting treatment that should be settled before the structure is agreed rather than after.

The transaction is easier to price once the three things being exchanged are separated.

Capital now, against rent later. The seller receives proceeds and takes on a fixed periodic obligation. That obligation behaves like debt service: it is owed whether or not the asset is productive in a given period, and it consumes the same cash flow that any other financing would.

The residual. This is the substance of the deal. The buyer now owns whatever the asset is worth at the end of the lease, and has priced the transaction on a view of that number. Where the buyer's view is optimistic the rent is lower; where the asset's end value is genuinely unknowable, the buyer either prices conservatively or declines. Everything in [residual value](residual-value) applies directly here, because a sale-leaseback is fundamentally a trade in the residual.

Control at the end of term. What happens when the lease expires — renewal rights, a purchase option at a fixed or fair-market price, an obligation to return the asset in a specified condition — determines whether the seller has released capital or merely deferred a problem. An operator that must have the asset after the term, with no right to keep it, has sold its own negotiating position along with the asset.

A fourth item is traded quietly and is worth naming: covenant and reporting obligations. The lease will contain undertakings about how the asset is used, maintained, insured and reported on. Those are real constraints on operations, and they are usually discovered after signing rather than negotiated before it.

Every layer of a data center holds capital that can, in principle, be released this way. What differs is how comfortably a buyer can price the end-of-term position.

A sale-leaseback is only available where the asset can genuinely be sold. That sounds trivial and is where most proposed transactions fail.

• Clean title, free of competing claims. Equipment already subject to a security interest, a vendor's retention of title, or an unpaid supplier's claim cannot be sold without dealing with that first. On equipment ordered against staged payments this is a live issue more often than sponsors expect.
• The asset has to be identifiable. A buyer is acquiring specific property and needs it described at a level that distinguishes it from everything else in the building. On a fleet of near-identical units this means serial-level schedules, maintained rather than captured once — the identification problem set out in [collateralizing compute](/compute/collateralizing-compute) applies to ownership just as it does to security.
• The buyer has to be able to reach it. An asset that sits in a facility controlled by a third party is only worth what the buyer can access, inspect and, if it ever comes to it, remove. Consents and access rights from the facility owner are part of the transaction, not an afterthought to it.
• The underlying rights have to survive the transfer. Selling a facility whose power rights or site control do not follow it, or whose lease prohibits the transfer, produces a buyer holding less than it paid for. The [site financeability gates](/sites/what-makes-a-site-financeable) are the same questions asked from the other side.
• Term has to match use. A lease shorter than the period the seller needs the asset creates an end-of-term problem; one longer than the asset's economic life creates rent payable on something no longer worth running. This is the [tenor](tenor) question, and it is the most common structural error in the transaction.

Where those conditions hold, a sale-leaseback is one of the cleaner ways to release capital from a built asset — which is why it appears at every layer of the stack. Continuum structures these transactions layer by layer: establishing what can actually be sold, what has to travel with it, and how the lease terms sit against the contracts already running through the asset.