The Special Purpose Vehicle: Ring-Fencing One Asset

A special purpose vehicle is a company formed to hold one asset and do nothing else. The cash it produces cannot be mixed with anything, and the claims against it cannot reach anything. Tapti Crossing Infrastructure Private Limited is one: it holds a single crossing, and the Rs 1,260 crore lent to it stops at the vehicle's own boundary. The boundary works in both directions.

The picture almost every reader carries is half right, and the missing half is the interesting one. Most people, asked what a special purpose vehicle is for, will say it protects the parent from the project. The money goes somewhere separate, and if the thing fails, the failure stays in the box. The protection is real. The protection is also the smaller half of the arrangement, and a reader who stops there cannot explain the one fact that matters most: the lenders wanted the box too, and insisted on it. If the box existed only to shield the people who put the equity in, no lender would have signed a line of it.

Here is the everyday version. Think of a household where one member takes a loan to open a small tailoring unit in a rented shed. If that person signs the loan in their own name, the lender can eventually reach the household's savings, the scooter, the fixed deposit set aside for a wedding. Now imagine the shed and its machines are put into a separate registered firm, and the lender agrees to look only at the shed. Two things happened at once, and they are not the same thing. The household's savings became unreachable from the shed. The shed's takings became unreachable from the household too, and that second change is what the lender was buying. The till is now the lender's cover, so nobody in the house can dip into it on a bad month.

The second change is the one that everything about the arrangement turns on. A ring-fence is not a shield with a favoured side. A ring-fence is a wall, and a wall inconveniences everyone standing near it. Tapti Crossing Infrastructure Private Limited, an invented single-asset toll road company, was formed to build and operate one crossing. The project cost Rs 1,800 crore, funded 70 to 30 with Rs 1,260 crore of debt and Rs 540 crore of equity. In the modelled year it collects Rs 310 crore, spends Rs 62 crore running the road, and so produces Rs 248 crore of earnings before interest, tax, depreciation and amortisation (EBITDA). Its debt service in that year is Rs 182.70 crore. Cost, debt, equity, EBITDA and debt service do all the work below.

What is a special purpose vehicle, and what does single purpose actually restrict?

A special purpose vehicle is an ordinary company in every mechanical respect. The vehicle is incorporated, it has shareholders, it has a board, and it files what companies file. Nothing exotic has been contrived. A deliberate narrowing is what makes the vehicle special: its objects clauseThe part of a company's constitution that sets out what business it is permitted to carry on. Anything outside it is beyond what the company may do. points at one activity, its contracts all attach to that activity, and its borrowings are raised for that activity and no other. Tapti Crossing Infrastructure Private Limited exists to build and operate one crossing. The company has no second line of business, no legacy division, no trading history, and no other source of cash.

Now the point readers most often get backwards. Single purpose is a restriction on what the vehicle may do, not a restriction on who may hold its shares. The sponsors hold the equity and put in Rs 540 crore of it. The sponsors appoint directors, they receive whatever the road distributes, and in every ordinary sense the shares are theirs. The arrangement never claimed the sponsors were strangers to the vehicle, so none of that undoes it. The arrangement claimed the vehicle could not be pointed at anything else, and that its obligations were its own.

Ask yourself why a lender would care about the difference. A lender lending against a business is lending against a portfolio: many customers, many products, many years of history, and the reasonable expectation that if one part struggles another carries it. A lender lending to Tapti Crossing Infrastructure is lending against exactly one stream of collections from exactly one crossing. One stream sounds worse than a portfolio, and in one sense it is. But it is knowable. There is no second division that might swallow the cash, no unrelated acquisition, no sudden decision to enter a new market with money the crossing generated. The narrowness that makes the vehicle fragile is the same narrowness that makes it legible.

Set that beside an ordinary corporate borrower for contrast. Harivansh Packaging Limited makes rigid and flexible packaging. The company shows revenue of Rs 3,180 crore, EBITDA of Rs 477 crore and borrowings of Rs 740 crore. The Rs 740 crore is not attached to a machine or a plant in the way the crossing's debt is attached to the crossing. The borrowings sit against the whole business, and the whole business stands behind them. If one customer leaves, another pays. If one plant burns, the others run. The flexibility is real, and it costs the lender the ability to say precisely what is producing the cash it is being repaid from.

Which way does the ring-fence run, and why does the direction matter?

Two directions, and both were bargained for. Outward: a claim against Tapti Crossing Infrastructure Private Limited cannot cross the boundary to reach the sponsors. If the crossing collects less than the schedule requires and the vehicle cannot pay, the project lenders take what the vehicle holds and stop there. Inward: a claim against the sponsors cannot cross the boundary to reach the crossing's cash. If a sponsor's other business fails, or a sponsor's other lender enforces against it, the Rs 248 crore the crossing produces and the Rs 91.35 crore sitting in its reserve are not part of that argument.

Both directions were negotiated by somebody who wanted them, and a reader who remembers only the outward one cannot explain why the project lenders demanded the structure at all. Say it plainly: the outward block is the sponsors' half of the deal, and the inward block is the lenders' half. The sponsors get a cap on what a failure can cost them. The lenders get a pool of cash that nobody else is allowed to touch, including the people who put in the equity and appointed the directors.

The two-sided bargain is why the arrangement holds together rather than being a one-sided favour that lenders would refuse. In an ordinary corporate borrowing, a lender ranks against a business alongside every other creditor of that business, and the cash it is repaid from can be spent on a new factory or an acquisition before it ever reaches the lender. In a ring-fenced structure, the collections arrive in accounts the lenders can see, are applied in an agreed order, and cannot be diverted into anything the sponsors would rather do with the money. The vehicle's narrowness is the lender's security in a way that no covenant on a diversified borrower can replicate.

Go back to the tailoring shed for a second. The household is protected because the lender cannot reach the wedding deposit. But the household also cannot reach the shed's till, and that is not an accident or an oversight. The second restriction is the exact condition on which the lender agreed to stop at the shed. Take away the second restriction and the first one becomes indefensible. The people holding the shares would be free to empty the till, and nobody agrees to look at one till on those terms.

What does the vehicle protect the project lenders from?

From everything the sponsors are and everything the sponsors do. The list is longer than it first looks, so unpack the answer slowly. The project lenders to Tapti Crossing Infrastructure Private Limited are protected from the sponsors' other businesses, whatever those may be. From the sponsors' other borrowings, and from the lenders behind those borrowings. From a decision by the sponsors to expand somewhere else using cash the crossing generated. From a dispute that has nothing to do with roads. From, in the plainest terms, being one creditor among many in a group whose obligations they never underwrote.

The Rs 248 crore the crossing produces sits inside a company that cannot be asked to help with anything else, so it is available for debt service by construction rather than by anybody's good behaviour. Notice the phrase by construction. Availability by construction does not depend on a promise, a covenant tested once a year, or a management team choosing to prioritise the loan. The cash cannot go anywhere else because there is nowhere else for it to go. The company has one asset and one set of obligations.

Put that against the ordinary case. A lender to Harivansh Packaging Limited is looking at Rs 477 crore of EBITDA produced by a business that could, tomorrow, buy something, build something, lose a customer, or take on another Rs 500 crore of borrowing. Every one of those events changes what stands behind the loan, and none of them requires the lender's blessing unless a document says so. The lender is not underwriting a road. The lender is underwriting a set of future decisions by people it does not control.

There is a second protection, quieter and just as important. The debt service reserveCash set aside inside the borrower to meet a stated number of future debt service payments if collections fall short in a period. of Rs 91.35 crore, being two quarters of the Rs 182.70 crore annual debt service, is cash the sponsors cannot take out. In a corporate borrower, cash on the balance sheet is available to management for whatever they judge best. Here it is not. The Rs 91.35 crore is parked, visible and reserved for one job. The ring-fence is what makes that possible: cash cannot be fenced inside a company that has a dozen other things to spend it on.

What does the vehicle protect the sponsors from?

From an unlimited loss. If the crossing never carries the traffic anyone hoped for, if collections come in far below the modelled Rs 310 crore year after year, if the whole thing has to be handed over, the sponsors lose the Rs 540 crore of equity they put in and nothing more. Their other businesses continue. Their other lenders are not affected. No one arrives with a claim for the unpaid balance of Rs 1,260 crore.

The cap on the equity is what makes a project of this size financeable by a sponsor who could never have carried it on their own balance sheet. Work the counterfactual and the point becomes obvious. Suppose the sponsors had borrowed Rs 1,260 crore on their own account and built the crossing as a division. Their borrowings rise by Rs 1,260 crore, their leverage rises with it, every existing lender to their business is now standing behind a road they never assessed, and a failure of the crossing is a failure of the whole group. Very few sponsors have the balance sheet to absorb that. The ring-fence lets a business of moderate size stand behind a project several times its own scale. The sponsor stands behind the equity cheque and not the debt.

Now the price, and it is a real one. The sponsors cannot reach into the cash either. In the modelled year the crossing produces Rs 248 crore of EBITDA and pays Rs 182.70 crore of debt service, leaving Rs 65.30 crore for the sponsors, which on Rs 540 crore of equity is a 12.09 per cent cash return in that one year. The 12.09 per cent is one year's cash return and not a return over the life of the concession. And even that Rs 65.30 crore is not simply available: it comes out only after the obligations above it are met and the reserve is where it is meant to be. In a bad year the sponsors get nothing, and they cannot vote themselves a payment out of the reserve to fix it.

The household version again, and it is exact. The tailoring unit has a good month and the till is full. The household would like to use some of it for a nephew's fees. The shed's takings are the lender's cover, and the agreement says what happens to them first, so the household cannot. The protection and the restriction are the same clause read from two sides.

Why is the ring-fence a dial rather than a switch?

Because in practice sponsors very often agree to support the vehicle in some limited way, and a reader holding a two-position picture in their head cannot place those arrangements at all. Such a reader sees support and concludes the whole thing was theatre. The conclusion is wrong.

The honest description is a range. At one end sits a vehicle with no sponsor support of any kind, where the project lenders look only at the project and stop there. At the other end sits a borrowing where the sponsors stand behind the whole obligation, a corporate borrowing wearing a project's clothes. Everything in between is limited recourseA lending arrangement where the lender's claim reaches the project and a stated amount of sponsor support, and no further.: support for a defined period, most often during construction and until the asset is producing cash; support capped at a stated amount; support triggered by a defined event and by nothing else.

A limited support undertaking is a position on a range, not a violation of the structure. It is information rather than a verdict. A support undertaking shows which risks the lenders were unwilling to take on the project alone, and by elimination which ones they were. Support that falls away once construction is complete says the lenders were comfortable with a finished road collecting tolls and uncomfortable with a half-built one. Falling-away support is a precise statement about where the risk was thought to sit, and the statement disappears entirely once the whole arrangement is filed under either fully protected or not really protected.

Tapti Crossing Infrastructure Private Limited sits at the far end with no sponsor support at all. The zero end isolates the mechanism and leaves nothing to attribute to a guarantee. Real arrangements commonly sit somewhere short of that end, and where any particular one sits is a fact about that arrangement, recorded in its documents, not something a reader can infer from the words project finance appearing in a headline.

What does a bad year cost, and who does the cost land on?

Everything about the crossing is held steady and one year's revenue is reduced by 25 per cent. Nobody can forecast a year's traffic on a new crossing, so the size of the fall is chosen rather than predicted. Revenue falls from Rs 310 crore to Rs 232.50 crore. The road still has to be lit, patrolled and maintained whether or not the traffic came, so operating cost stays at Rs 62 crore. So EBITDA is Rs 170.50 crore. Debt service is unchanged at Rs 182.70 crore, being Rs 119.70 crore of interest at the project's own contracted 9.5 per cent on Rs 1,260 crore plus Rs 63 crore of scheduled principal.

The coverage ratioCash available in a year divided by what has to be paid on the borrowing in that year. Above one means the year's cash covered the year's obligation. is therefore Rs 170.50 crore over Rs 182.70 crore, which is 0.93 times. Cover below one means the year's cash did not meet the year's obligation. The gap is Rs 12.20 crore.

Every number in that right-hand column is a fact about the crossing, and not one of them changes depending on what the sponsors agreed to. The road collected what it collected. The road cost what it cost. The lenders are owed what they are owed. The shortfall is Rs 12.20 crore whatever anybody signed. The documents decide a different question entirely: who produces the Rs 12.20 crore.

Take the three settings in turn. At this project's actual setting, with no sponsor support at all, the Rs 12.20 crore comes out of the reserve. The reserve holds Rs 91.35 crore, so it meets the shortfall 7.49 times over and Rs 79.15 crore remains in it. At a hypothetical setting where the sponsors had agreed to meet half of any shortfall, they would put in Rs 6.10 crore, the reserve would meet the other Rs 6.10 crore and would still hold Rs 85.25 crore, or 14.98 times the amount it was asked for. At full support the sponsors put in the whole Rs 12.20 crore and the reserve is untouched at Rs 91.35 crore.

The project's cash, its cover and its shortfall are identical in all three rows, and the only thing that moved is who the shortfall lands on. That is the whole of what a ring-fence decides. A ring-fence is not a device that makes a bad year smaller. A ring-fence is a device that settles, in advance and in writing, whose problem a bad year is.

What makes a vehicle actually separate in practice?

Not the word separate appearing in a document. Separateness is a stack of ordinary administrative facts, and it is worth listing them because the list is far less glamorous than the concept and far more decisive.

The vehicle is incorporated on its own, as its own company, with its own constitution. The company holds its own contracts in its own name, the concession and the construction and operating arrangements included, rather than being a line item inside somebody else's agreements. The vehicle keeps its own accounts, and its collections and payments run through bank accounts in its own name rather than through a sponsor's treasury. And the security taken over it is registered. The chargeA security interest recorded over an asset or a stream of cash, which places the holder ahead of others in a claim against that asset. over its assets and its receivables is a matter of public record rather than a private assertion.

Separateness is administrative before it is legal, and the paperwork is not a formality: a vehicle whose cash is mixed with a sponsor's is a vehicle whose separateness somebody will eventually have to argue about. Notice how the failure happens. Nobody sets out to destroy a ring-fence. Somebody in a treasury function decides that opening a new account for a small collection is a nuisance, or that a shared payment run is more efficient, or that an invoice can be settled from whichever account has cash today and squared up later. Each decision is trivial. Together they produce records where whose money was whose has to be reconstructed rather than read.

The household test is the same one every accountant applies to a small business. If the shop's cash and the house's cash go into one steel box, the shop has no accounts, whatever anybody says about the shop being a separate concern. When a dispute comes, the box is the evidence, and the box says one thing.

What breaks the ring-fence entirely?

Three things, and they are worth naming precisely because two of them look like housekeeping and the third looks like generosity.

The first is a full guarantee from the sponsors. If the sponsors stand behind the whole of the Rs 1,260 crore, then a lender who is not paid by Tapti Crossing Infrastructure Private Limited can be paid by somebody else, and the vehicle has become a place where a borrowing is recorded rather than a boundary that stops anything. The second is a vehicle that acquires a second business. The moment the company holds two assets, its cash is mixed by construction, its lenders are exposed to something they never assessed, and the single-purpose restriction that made the whole arrangement legible has been abandoned. The third is cash that moves in and out without an agreed order. If money can leave the vehicle whenever the shareholders find a use for it, the reserve is a number in a spreadsheet rather than a funded buffer, and the lenders' position is exactly what it would have been lending to anybody.

The single test is this: if a lender that is not paid by the vehicle can be paid by somebody else, the ring-fence is decorative, and the pricing on the facility will have been struck as though it were. That last clause matters. Lenders are not fooled by a structure chart. A facility priced against a project's own cash and a facility priced against a group's balance sheet are different instruments at different rates, and the rate paid shows which one the lenders thought they were writing. The crossing's own contracted rate is 9.5 per cent, struck for one asset in one year, and no rate travels from one facility to another.

If the debt shows up in the sponsors' group accounts, is there recourse?

No, and this is where most readers lose their grip on the whole idea, so take it slowly. A sponsor holding a controlling stakeA shareholding large enough that the holder directs the company's decisions, and large enough to bring its figures into a set of group accounts. in Tapti Crossing Infrastructure Private Limited will bring the vehicle into its group accounts. ConsolidationAdding a controlled company's figures line by line into the accounts of the company that controls it, so the group is presented as though it were one entity. adds the vehicle's figures line by line, so the Rs 1,260 crore turns up inside the sponsors' group borrowingsThe total borrowings shown in a set of consolidated accounts, added across every company included in them.. A reader opens the group balance sheet, sees the number, and concludes that the ring-fence was a story.

Appearing in a set of accounts is a reporting outcome and creates no claim on anybody. Consolidation answers the question who controls this, and it answers it correctly. Consolidation does not answer the question who can be made to pay. The lending documents settle that question, and nothing else does. Two entirely different questions, and the accounts only ever answered the first one.

The household version, one more time. The confusion vanishes with it. Suppose a parent lists everything the house contains for an insurance schedule, including a scooter the son bought on a loan in his own name. The scooter is on the parent's list because the parent runs the house. The loan is still the son's. Nobody at the lender reads the insurance schedule and concludes the parent has agreed to pay. The list describes control, not obligation.

So a group borrowings figure that includes non-recourse project debt overstates what the group actually has to pay on its own account. Consolidation reports something true, so the correction is not to ignore it but to read it for what it says. And the practical fix, the habit worth building, is to state the leverage twice: once including the vehicle and once excluding it, saying which is which every time.

How does a lender, an analyst or a sponsor's finance team use this?

Three readers, three different first questions.

A project lender reads the structure as a list of leaks to close. Are the collections arriving in accounts it can see. Is the reserve funded, and with whose money. Can anything leave the vehicle before the obligations above it are met. Could the vehicle take on a second business or a second borrowing without asking. Every one of those is a question about whether the ring-fence will still be a fence in year seven, and the answers sit in documents rather than in a model. The lender is not asking whether the crossing is a good asset. The lender has already decided that. The lender is asking whether the cash the crossing produces will still be reachable when it needs it.

An analyst covering a listed sponsor reads it as a labelling problem, and does the two-figure exercise described above before anything else. Three figures get named: what the group reports, what the group has to pay on its own account, and the difference between the two. Then a second question that the accounts will not answer: what did the sponsors actually promise. A support undertaking capped at a stated amount is an exposure. A commitment to fund cost overruns during construction is an exposure. Neither appears as debt anywhere, and neither is zero.

A sponsor's own finance team reads it as a question about what they can and cannot count on. The Rs 65.30 crore the vehicle produces for equity in the modelled year is not cash in their hands until the obligations above it are met and the reserve is where it is supposed to be. Planning group spending around it before that is the same mistake as a household spending a bonus in March that is paid in June. The discipline the ring-fence imposes on the sponsors is the same discipline that made the lenders willing to lend against one road in the first place. The two are never separable.

And an equity investor considering the sponsor reads it as a question about scale. The ring-fenced structure is what allows a business to stand behind a project several times the size its own balance sheet could carry, at a cost of Rs 540 crore and a set of restrictions. Whether that trade was worth making is a judgement about what the money would otherwise have done.