Project Finance vs Corporate Finance: What Differs

Corporate finance lends against a whole business, its history and everything it holds. Project finance lends against one asset's cash and stops there. Tapti Crossing Infrastructure Private Limited carries gross debt of 5.08 times earnings before interest, tax, depreciation and amortisation (EBITDA) where a packaging maker on the same measure carries 1.55 times, and that gap follows from what is being lent against rather than from carelessness.

What is each of these two arrangements actually lending against?

Two people ask a lender for money on the same afternoon. The first runs a household that has been in the same trade for twenty years: three earners, a shop, some gold, a spare room let out to a tenant. If one income stops, four other things carry on. The second has no history at all. She has a written arrangement to run the tea counter inside one office building, and the only money she will ever see is what that counter takes.

A lender would probably lend to both. The lender would not read the same things before deciding, would not ask for the same protections, and would not lend the same amount against the same annual earnings. The difference between those two borrowers is the difference between corporate finance and project finance, and everything below is the finance version of that afternoon.

The company: what a corporate lender is looking at

Harivansh Packaging Limited, an invented manufacturer, makes rigid and flexible packaging for food and personal care customers. The company borrows the way an ordinary manufacturer borrows: no transaction, no purchase, no combination. Its revenue is Rs 3,180 crore and its EBITDA is Rs 477 crore, a margin of 15.0 per cent. Depreciation and amortisation of Rs 138 crore takes it to earnings before interest and tax (EBIT) of Rs 339 crore. A finance cost of Rs 60 crore and other income of Rs 21 crore give profit before tax of Rs 300 crore, tax of Rs 75 crore at an effective 25.0 per cent, and profit after tax of Rs 225 crore.

Underneath that sits a net worth of Rs 1,650 crore, borrowings of Rs 740 crore and cash of Rs 140 crore. The shape matters more than the amounts: many customers, many products, a management team that can decide next year to sell something different, and a balance sheet full of things a lender could look at if the packaging business stopped paying.

The project: what a project lender is looking at

Tapti Crossing Infrastructure Private Limited, an invented project company, was formed to build and operate one crossing. The vehicle has no other business, no second stream of cash, and no past. The crossing cost Rs 1,800 crore to build, funded with Rs 1,260 crore of debt and Rs 540 crore of equity, a 70 to 30 structure. In the year modelled it collects Rs 310 crore of revenue and spends Rs 62 crore running the crossing, so its EBITDA is Rs 248 crore, an 80.0 per cent margin that is ordinary for a road and would be remarkable almost anywhere else.

Notice what the second record does not contain. There is no revenue history. There was no revenue at all until the crossing opened. There is no product range. There is one product. There is no management decision to model. The granting authority fixed what the vehicle may charge, what it must maintain and for how long it may operate, all of it in writing before the first load of concrete arrived.

Corporate finance lends against a business that can change what it does, and project finance lends against one asset producing one stream under arrangements settled before construction began. Both halves of that sentence carry weight. The project is narrower, a weakness, and more predictable, a strength, and it is both of those things for exactly the same reason.

What can each lender reach if the borrower cannot pay?

Think about the shopkeeper again. If she borrows and signs a personal guarantee, she has put the house behind the shop: when the till comes up short the lender can move past the shop to everything else she has. If instead the lender agrees to look only at the shop's takings, the house is untouched, and the lender knows it before lending a rupee. The interest rate will differ, the paperwork will differ, and the amount will differ, all because of that one clause.

The word for this is recourseThe set of people and assets a lender may pursue for repayment if the borrower does not pay. Wide recourse reaches past the borrowing entity; narrow recourse does not.. A lender to Harivansh Packaging Limited has recourse to the whole company. The company is one legal person holding the plants, the stock, the receivables and the Rs 140 crore of cash, and a claim against that company is a claim against all of it. If one product line collapses, the lender is still standing in front of everything else.

A lender to Tapti Crossing Infrastructure Private Limited is in a different position, and the structure is what puts it there. The arrangement is non-recourseLending where the claim is limited to one borrowing entity and the assets charged to it, with no right to pursue the shareholders standing behind that entity., so the claim runs to the vehicle and to what is charged to it, and no further. The sponsors who put in the Rs 540 crore of equity are behind a wall. The wall has a name, the ring-fenceThe legal separation that keeps one project's assets, cash and borrowings inside a single company, so that neither its lenders nor its shareholders can reach across the boundary., set out separately under its own name.

The same rupee of debt is a completely different instrument depending on where that boundary is drawn, and nothing about either structure matters as much as where it sits. There is nobody else to pursue, so a project lender who has read the arrangements badly cannot recover the mistake afterwards by pursuing somebody wealthier.

What does each lender read before the money moves?

A corporate lender opening a file on Harivansh Packaging Limited has years of material to work with. Statements going back, a customer list that has survived some losses, a working capital cycle that behaved a certain way through a bad patch, an opinion from a rating agency, and management who can be asked what they did last time demand fell. The past is the evidence. A business can change month by month, so the lender forms a view about it and then keeps watching.

A lender to Tapti Crossing Infrastructure Private Limited has none of that. The vehicle was formed for this crossing and has no past to examine. So the reading material is the arrangements themselves: what the granting authority has agreed the vehicle may charge and for how long, who has committed to build it and at what price, who will operate it, what happens if construction runs late, and a cash model that turns all of it into a stream of receipts and payments. The evidence is not history; it is a stack of documents and the arithmetic they support.

Almost all of a project lender's work sits before any money moves, and almost all of a corporate lender's work sits after it. The concentration of effort is why a project financing takes so long to reach first drawdown and comparatively little effort afterwards, and why an ordinary borrowing can be agreed quickly and then monitored for years. The order feels backwards until the reason is visible: no amount of monitoring rescues a structure that was drafted wrong, and no amount of drafting rescues a business that stops winning customers.

Neither lender is being more careful than the other. Each lender can still influence a different point in time, so the same care goes into different places.

Why does the crossing carry leverage the manufacturer would never be given?

A leverage multipleBorrowings divided by a year's earnings, most often EBITDA. It answers roughly how many years of that year's earnings the debt represents. is a shorthand for how many years of a given year's earnings the debt represents. Compute both. Tapti Crossing Infrastructure Private Limited has Rs 1,260 crore of debt against Rs 248 crore of EBITDA, a multiple of 5.08 times. Harivansh Packaging Limited has Rs 740 crore of borrowings against Rs 477 crore of EBITDA, a multiple of 1.55 times. The crossing carries more than three times the leverage of the manufacturer on the identical measure.

Before drawing any conclusion from that, settle the base. The base is where the comparison is most often broken. Gross debtThe full amount of borrowings outstanding, before subtracting any cash the borrower is holding. is what is owed. Net debtBorrowings less cash held. It assumes the cash could be used to repay, which is a reasonable assumption for some borrowers and not for others. is what is owed less the cash the borrower is holding against it. Harivansh Packaging holds Rs 140 crore of cash, so its Rs 740 crore of gross borrowings becomes Rs 600 crore net, and 1.55 times gross becomes 1.26 times net. The record for the crossing carries no cash balance at all, so it has a gross figure and no net figure in existence.

The honest pairing is 5.08 times against 1.55 times, gross set against gross, and setting 5.08 against 1.26 would put a gross number beside a net one and read the mismatch as a finding. The size of that error is computable rather than a matter of taste: the mismatched pair shows a gap of 3.82 turns where the honest pair shows 3.53 turns, and the 0.29 turns of difference is exactly Rs 140 crore of cash divided by Rs 477 crore of EBITDA. The 0.29 turns is not a fact about the crossing; it is one company's bank balance wearing a disguise.

So why was the crossing lent five times its earnings? Because the lender is not asking the same question. Lending 1.55 times to Harivansh Packaging is a bet that a packaging business will keep winning customers, keep its margin near 15.0 per cent and keep replacing the machines, for years, through decisions nobody has made yet. Lending 5.08 times to the crossing is a bet that traffic crosses a river and that a written arrangement holds. The second stream is smaller and narrower, and it is also far harder to interrupt. A lender will advance more against a narrow stream it can predict than against a wide one it cannot.

What do the two records look like set side by side on the same measures?

Here they are side by side, on identical measures and identical bases. The crossing's interest is computed rather than quoted: its own contracted rate of 9.5 per cent on Rs 1,260 crore is Rs 119.70 crore. The manufacturer's finance cost is taken as the Rs 60 crore the record carries. The record gives Harivansh Packaging Limited a finance cost as an amount and not as a rate, so no rate is derived for it.

Read the pair honestly. The crossing carries more than three times the leverage on the same measure and a little over a quarter of the interest coverageA year's earnings divided by a year's interest. It asks how many times over the earnings could pay the interest bill before anything else is paid.. Interest alone takes 48.3 per cent of the crossing's EBITDA and 12.6 per cent of the manufacturer's. On every measure a credit committee is used to, the crossing looks like the worse borrower, and it was funded anyway.

The crossing was funded because a lender looking at one structure asks whether it produces one stream, and a lender looking at the manufacturer asks whether a business will keep winning customers for years. The two questions are different, and a ratio built to answer the second one does not automatically answer the first.

The funding mix says the same thing from the other end. The crossing was built with 70.0 per cent debt and 30.0 per cent equity. Harivansh Packaging Limited sits on capital employed of Rs 2,390 crore, being net worth of Rs 1,650 crore plus borrowings of Rs 740 crore, so its borrowings are 31.0 per cent of the total and its shareholders' funds are 69.0 per cent. The two structures are close to being each other's mirror image, and nobody arranged that deliberately; it falls out of what each lender is willing to advance.

One caution matters here. Neither leverage figure settles whether the crossing can repay. Repayment depends on a schedule and on a period, and this record carries neither: no length for the right to operate, no term for the debt, no traffic forecast and no year by year build.

Who decides where the cash goes once it arrives?

Ask a household this question and the answer is obvious: whoever holds the money decides, within whatever they have promised. A company works the same way. When Harivansh Packaging Limited collects from its customers, the money lands in its own accounts and its treasury decides what to do with it: repay a borrowing early, buy a machine, hold it as part of the Rs 140 crore cash balance, or pay it out to shareholders. The lenders have agreed limits, and inside those limits a person exercises judgement every week.

Tapti Crossing Infrastructure Private Limited does not work that way at all. Toll collections land in accounts that pay things in a fixed order: running costs first, then interest and scheduled principal, then the topping up of a reserve, and only then whatever is left for the sponsors. The structure does not offer that as a choice, so nobody at the vehicle decides to pay the sponsors ahead of the schedule. The order itself, and the reserve inside it, are covered separately under debt service; what matters here is only that the order exists and is not discretionary.

The ring-fence has a practical meaning here, and it runs in both directions: the lenders cannot reach out past the vehicle, and the sponsors cannot reach in ahead of everything the schedule puts before them. The ring-fence is usually described as a protection for sponsors. The same wall is a restraint on them, and a project lender lends five times earnings partly because that restraint is real.

What absorbs a bad year in each structure?

Every business has a bad year eventually, and the interesting question is what takes the weight. Think of a household where three people earn and one loses work: the other two carry the month, someone defers a purchase, and the household survives without anyone outside noticing. Now think of the tea counter. One quiet month and there is nothing else in the arrangement to lean on except whatever was set aside earlier.

Harivansh Packaging Limited is the household. A slow quarter in one product is offset by another, a weak region by a strong one, and a genuine squeeze by drawing a working capital line, holding back a discretionary spend, or simply letting the Rs 140 crore cash balance do its work. None of these are dramatic. Every one of those moves is ordinary, and some combination is always available; that availability is what a corporate lender is really lending against.

Tapti Crossing Infrastructure Private Limited has one line of defence and then a hard stop. The hard stop is precisely why a project financing carries a debt service reserve at all. Debt service in the year modelled is Rs 119.70 crore of interest plus Rs 63 crore of scheduled principal, or Rs 182.70 crore in total, and a reserve equal to two quarters of that is Rs 91.35 crore. The reserve converts a timing problem into a funded buffer. A reserve is not a second business, and once it is spent the vehicle is at the end of what it can absorb.

A reserve is the substitute for having somewhere else to turn, and that is why a project has one and an ordinary company usually does not. Read that way, the reserve stops looking like extra caution and starts looking like what it is: the only shock absorber a single-asset structure can be given. The mechanics of how it is sized and drawn are covered separately under debt service.

Which of the two has a finite life, and what does that change?

A company is read as continuing. Nobody writing about Harivansh Packaging Limited assumes it stops on a stated date; the whole apparatus of statements and multiples rests on it carrying on. So when a facility reaches maturity with the borrowing still outstanding, the ordinary answer is refinancingReplacing a borrowing that is falling due with a new one, so that the old lender is repaid out of the new lender's money rather than out of earnings.: a new lender advances money, the old lender is repaid, and the business continues. The maturity was never really a deadline for the business, only for one agreement.

Tapti Crossing Infrastructure Private Limited runs on a right that ends. When that right ends, so does the vehicle's ability to collect anything at all, and there is nothing left to lend against. A new lender cannot advance against a crossing the vehicle may no longer operate. So the debt has to be repaid out of the crossing's own cash while the right is still running. Project debt therefore amortises on a schedule rather than sitting as a bullet repaymentA borrowing repaid in one lump at maturity, with only interest paid until then. It relies on the borrower finding new money or a large receipt on the day. to be dealt with on the day.

A finite life changes the repayment question completely. A company can refinance a maturity, and a project has to have repaid before its right runs out. The finite life explains the Rs 63 crore of scheduled principal sitting in the crossing's debt service alongside the interest, and it explains why a project lender cares about the schedule far more than about the multiple.

Now the discipline. The length of this project's right is not in the record. Neither is the term of the debt. The record therefore does not establish how many years of Rs 63 crore there are, whether the schedule steps up, or whether the debt is fully repaid by the end. A missing term cannot be replaced with a plausible one. The repayment answer turns entirely on the figure that is absent.

When is a borrowing called project finance even though it is not?

The term is most often misused at exactly this point, and it is worth being blunt about. A facility can be arranged around a new plant, described as project finance in every email about it, drawn into a company formed for the purpose, and still not be a project financing in any sense that changes a lender's position.

Three arrangements do it. The first is a full guarantee from the sponsors: if the parent stands behind the debt, the lender's claim reaches the parent, and the vehicle is a container rather than a boundary. The second is a vehicle that holds more than one asset: if the same company also runs three other things, the lender is lending against a small, diversified portfolio, and a portfolio is a business rather than a project. The third is any other route by which the lender can reach outside, whether a support undertaking, a shortfall arrangement or a charge over something the sponsors hold.

Ask one question: what can the lender reach when the cash falls short, and the answer names the structure. If the answer stops at the vehicle and what is charged to it, it is a project financing and every difference set out above applies. If the answer includes a sponsor, a parent or another asset, it is corporate lending with a project attached, whatever the cover of the agreement says. The label is a description somebody chose, and the reach is a fact somebody drafted.

The label matters more than a naming quibble. The whole justification for 5.08 times leverage was that the lender had priced a narrow, predictable stream and given up every other claim in exchange. Take away the giving up, and the leverage is being carried by the sponsor rather than by the crossing, and the sponsor's own borrowings are larger than they appear.

How does a credit team actually put this distinction to work?

Three people use this every week, and each of them uses it differently.

A lender's credit team asks the reach question before it asks anything numeric. The answer decides which file the team is writing. If the claim stops at the vehicle, the team commits its effort to the arrangements: who has committed to what, what happens if the crossing opens late, and whether the cash model matches the documents rather than the sales pitch. If the claim runs to a parent, the team goes back to reading a business and the project becomes one line inside it.

A sponsor's own finance team signs the equity cheque and therefore models it separately. The Rs 540 crore that went into Tapti Crossing Infrastructure Private Limited is money that could have gone somewhere else, and the whole point of the ring-fence is that the sponsor's exposure is capped at what it has put in plus whatever support it has actually given. The cap is only real if nobody quietly gave a guarantee to get the financing away.

An analyst reading a listed sponsor faces the hardest version. If a sponsor holds a vehicle carrying Rs 1,260 crore of non-recourse debt, does that debt belong in the sponsor's own leverage or not? The economic answer and the accounting answer can differ, and the accounting treatment is settled in the accounting layer rather than here. The question has to be asked explicitly and answered on one stated basis. An analyst who includes the vehicle's debt in the numerator and excludes the vehicle's EBITDA from the denominator has produced a number that describes nothing at all.

In all three uses the sequence is the same: establish the reach, then choose the basis, then compute the ratio, and never in the other order.