Infrastructure Funds: Long-Dated Assets and Contracted Cash Flow
An infrastructure fund buys assets whose cash flow is fixed by a long contract rather than reset by a market each year. Nilgiri Real Assets Fund I, an invented fund, holds an operating solar asset on a twenty-five year output contract with fifteen years still to run, producing Rs 9,90,00,000 a year on that fund's own cost, and a road under construction producing nothing. The contract outlives the fund's own ten-year term.
Two payments can look identical on a bank statement and be completely different objects. Rs 1,00,000 lands on the first of the month from a shop occupying a floor of a building. Rs 1,00,000 lands on the first of the month from somebody buying electricity. Same figure, same day, same account, and no way to tell them apart by looking at the credit. The difference is not in the payment. The difference is in the document sitting behind the payment, and specifically in how many years that document has left before somebody has to negotiate it again. A rent is reset when its lease ends, usually a matter of a few years. A contracted output payment is fixed for a period that can run longer than the entire life of the fund holding the asset. That single asymmetry produces everything else, including the collision between a fifteen-year contract and a ten-year fund.
What does an infrastructure fund actually hold?
Start with somebody who has bought a tempo. In the first version of the story, the driver takes it to the transport nagar each morning and finds a load. Some mornings there are four people bidding for the vehicle and the rate is good. Some mornings there is nobody and the rate is whatever gets the diesel back. The vehicle is the same vehicle every day. The income is decided fresh, every day, by whoever happens to be standing there.
Now the second version. The same driver signs an agreement with a school: run this route, these many trips, for this many years, at this monthly figure. Nothing about the tempo has changed. The engine is the same, the diesel is the same, the driver is the same. The change is that a piece of paper now decides the income instead of a queue of strangers deciding it every morning. And notice what that paper did and did not do. The paper fixed the amount. The paper did not make the school able to pay, it did not stop the gearbox failing, and it said nothing whatsoever about what the tempo will earn in the year after the agreement runs out.
An infrastructure fund is that second arrangement, at a very large scale, run by a pooled private vehicle on behalf of investors who committed money to it. The fund buys or builds long-lived physical assets, and the defining feature of what it buys is that somebody has already agreed, in writing and in advance, to take the output and pay for it. A payment fixed that way is a contracted cash flowA payment whose price and quantity are fixed in advance by a signed agreement., and the agreement producing it is an output contractAn agreement to buy what an asset produces, at a set price, for a set number of years.. The party on the other side of it, the one who actually has to make the payment month after month, is the counterpartyThe party on the other side of a contract, who has to actually make the payment., and naming that party separately matters a great deal, for reasons that emerge below.
Nilgiri Real Assets Fund I is a closed-end property and infrastructure fund, registered as a Category II Alternative Investment Fund, managed by Nilgiri Alternatives Advisors Private Limited, with Nilgiri Trusteeship Services Private Limited as trustee and Nilgiri Financial Holdings Private Limited as sponsor. As with every private vehicle of this shape in India, what the investors sign is a trust deed and a contribution agreement, and the role the global vocabulary calls the general partner is discharged by the manager and the trustee between them. At the record date this fund has Rs 3,75,00,00,000 deployed across five real assets. Two of those five are infrastructure, and they could not be more different from each other: one is producing contracted income right now and the other is producing nothing at all.
How does money travel from a promise into a contracted payment?
How Infrastructure Funds Work
The whole circuit runs in five steps, and the last three carry the weight. Step one, investors commit money to the fund, a promise rather than a transfer. Step two, the manager calls that capital when it is needed and the investors pay it in. Step three, the money is spent, and this is the step with two doors in it. Step four, the asset either produces or does not yet produce. Step five, the contracted payments arrive, and because the fund does not live forever, the asset eventually has to be sold to somebody else.
Steps one and two belong to the machinery of the vehicle rather than to infrastructure, and they are settled elsewhere: how a commitment works, how a capital call works, how the management fee is charged and how money is eventually paid back to investors are all covered separately and are used here without being explained again. Step three is the one that matters, and it opens onto a choice. An asset can enter an infrastructure fund through either of two doors, and which door it came through decides everything about how it looks in the fund's schedule for years afterwards. Through the first door comes an asset that was already built and already running when the fund bought it. Through the second comes an asset that has not been built yet and has to be built with the fund's money before it produces a single rupee.
Nilgiri Real Assets Fund I holds one of each. Asset 3 came through the first door: an operating solar generation asset, already running when it was bought, cost Rs 90,00,00,000. Asset 4 came through the second: a road asset under construction, with Rs 70,00,00,000 committed to it and Rs 45,00,00,000 drawn at the record date. Which is why the fund's asset schedule, read carelessly, looks like a mistake. One line has a full year of income against it and the next line has a blank.
What makes a cash flow contracted rather than market-set?
What makes a payment contracted rather than market-set?
The opposite of a contracted cash flow is a market-set cash flowA payment that is renegotiated at intervals against whatever the market will bear., and the honest way to describe the difference is not that one is better than the other. The difference is that the two payments are held up by different documents, and the documents have different lengths. Document length is the whole of it. Everything a reader thinks is different about the two payments follows from that one structural fact, and nothing follows from the words people attach to them.
Look at the two assets in Nilgiri Real Assets Fund I that make the contrast cleanly. Asset 3, the solar generation asset, sells its output under a single output contract running for twenty-five years, with fifteen still to run at the record date, and it produced net operating income of Rs 9,90,00,000 in the year, being 11.0 per cent on that fund's own cost of Rs 90,00,00,000 for that asset. Asset 2, the warehousing park, produced Rs 7,20,00,000, being 9.0 per cent on that fund's own cost of Rs 80,00,00,000 for that one, and it produced that out of leases which will be renegotiated several times inside the same period the solar contract still has to run.
Both figures are one fund's own income divided by that same fund's own cost for that one asset, in one year, so neither says anything about what an asset of that kind yields anywhere. A different cost would have produced a different percentage from exactly the same income. The same holds for every percentage below. A percentage of that kind is a division that already happened, not a rate anybody is offering.
The warehousing park is paid under leases and the solar asset under a twenty-five year output contract. Which of the two needs a schedule of when its documents expire?
What does long-dated mean, and long compared with what?
The word long-datedDescribing an asset whose contracted income runs for a long period, often longer than the fund holding it. gets used as though it means old, or large, or serious. The word means none of those. Long-dated means the document behind the income has a lot of years left on it, and the only useful way to use the word is to finish the sentence: long compared with what? A reader who blurs the three answers makes the mistake set out below, so the three are worth separating.
Long compared with a lease, first. A lease on a floor of a building is typically renegotiated several times inside fifteen years, so a single document running fifteen more years is long in the way a fifteen-year loan is long next to a thirty-day one. Long compared with the fund, second, and this is the one that bites: fifteen contracted years sit inside a vehicle whose contracted life is ten. Long compared with anybody's ability to see ahead, third. Nobody drafting a twenty-five year document knows what the fifteenth year looks like, and the contract does not make that year knowable, it simply fixes a price for it and moves on.
With the two income patterns drawn next to each other, the difference stops being a word. A contracted payment is a flat line running to a date. A market-set payment is a set of steps, one step for each time a document ran out and the price was made again. And there is a discipline here that costs nothing and saves a reader from a bad habit: the steps do not have a known direction. Drawing them all going up would be a claim nobody is in a position to make.
The market-set line is drawn as steps. Which way do the steps go?
What are the two infrastructure assets in this invented fund?
Everything below is arithmetic on four figures, so here they are plainly. Asset 3 of Nilgiri Real Assets Fund I is an operating solar generation asset that was already running when the fund bought it. The solar asset cost Rs 90,00,00,000. The asset sells its output under a twenty-five year contract with fifteen years still to run at the record date, and it produced net operating income of Rs 9,90,00,000 in the year. The income is 11.0 per cent on the fund's own cost of Rs 90,00,00,000 for that asset and is a statement about nothing else at all.
Asset 4 is a road asset under construction. Rs 70,00,00,000 is committed to it, of which Rs 45,00,00,000 is drawn at the record date, leaving Rs 25,00,00,000 not yet drawn. The road produces no income. Not a small amount, not an uncertain amount: nothing, and it will produce nothing until construction finishes. Its own risks, and there are four of them, are covered under greenfield and brownfield assets, where asset 4 is worked against asset 3 in exactly that way.
Getting the pairing wrong is easy, so state it carefully: a producing asset sitting beside a silent one does something specific to the fund's headline figure. The fund's net operating income across all five of its real assets is Rs 34,90,00,000 a year. Divided by the Rs 3,75,00,00,000 the fund has deployed, that is 9.3 per cent. Divided by the Rs 3,30,00,00,000 that is actually producing income, it is 10.6 per cent. Both figures are correct and they differ by more than a point. Asset 4 is the entire reason, and the mechanism is plain: Rs 45,00,00,000 of drawn capital sits in the first denominator and not in the second. The Rs 3,30,00,00,000 producing income is 88.0 per cent of the Rs 3,75,00,00,000 deployed. Quoting either percentage without naming which denominator produced it is not a rounding difference, it is a false statement, and that split is worked in full under the net operating income line.
This fund's net operating income is Rs 34,90,00,000 a year. One person quotes it as 9.3 per cent and another as 10.6 per cent. Who is wrong?
Why is there a road that pays nothing in a chapter about contracted income?
Because that is what the schedule of a real infrastructure fund looks like, and pretending otherwise would teach the reader to expect a tidier document than the one they will actually be handed. An asset that comes through the second door has a period, sometimes a long one, in which the fund has put real money in and gets nothing back, and during that period the asset sits in the schedule with a cost against it and a blank in the income column. The four risks specific to that period, and the contrast between an asset being built and an asset already running, are covered separately under greenfield and brownfield assets.
The blank is not an error, it is a stage, and the fund's headline percentage moves depending on whether that stage is left in the denominator or taken out. The 9.3 and the 10.6 both exist for that one reason. The solar asset is where the central collision appears, so the road can now be set down.
The contract has fifteen years to run. The fund's contracted life is ten years. What has to happen to the asset?
What happens when a fifteen year contract sits inside a ten year fund?
Nilgiri Real Assets Fund I is a closed-end fund. Its termThe fixed number of years a closed-end fund runs for before it must wind up. is ten years from its own final close, on the same arrangement the manager's other funds carry, with two possible extensions of one year each. At the end of it, the fund has to give the money back. The wind-up is not a preference or a plan, it is what the documents say, and everybody who committed capital signed on that basis.
Asset 3's output contract has fifteen years still to run at the record date. Fifteen is more than ten, the fund's base term. Fifteen is also more than twelve, the longest the fund could possibly run if both of its one-year extensions were taken. So there is no version of this fund's own life in which it holds asset 3 until the contract ends, and that is true whatever year of its life the fund happens to be in on the record date. Now put a number on the gap, and here the number is useless without its base attached. Against the twelve year maximum the overhang is at least three contracted years. Against the ten year base term with neither extension taken the overhang is at least five. The extensions exist and a floor of five does not survive them, so three is the figure to quote when only one is wanted. The fund is not sitting at its final close on the record date, so both figures grow the further into its own life the fund already is. Some of its years have already gone.
Notice also what those already-elapsed ten years of the contract are not. The contract runs twenty-five years and fifteen remain, so ten have run, but the fund did not collect all ten. Asset 3 was already running when the fund bought it, so an unknown part of that decade was somebody else's income. The point is small, and a schedule hides it quietly: the age of a contract is not the length of anybody's holding period.
Where does this fund's money from the solar asset actually come from, given the collision?
What must the fund do with the asset, and what does that mean for the buyer?
The fund has to sell the asset, with the remaining contract attached. A sale is the only exit available to a closed-end vehicle that reaches the end of its life while an asset it holds is still under a live agreement. And once that sentence is said out loud, the arithmetic of where the fund's money comes from changes shape completely.
Think of somebody who has taken a shop on a nine-year agreement and sublet it on a fifteen-year one. In year nine they cannot hand over fifteen years of income to themselves. The subletter can hand over the shop, with six years of somebody else's agreement attached to it, and whatever a buyer will pay for those six years is what those six years are worth. Not what the six years will pay out, but what somebody will pay today for the right to receive them. The two numbers are different and they are decided by different things.
So the fund's money from asset 3 is a stream of contracted payments for part of the period, plus one price at the end, and the price is a negotiation rather than a contractual entitlement. The remaining contracted years turn up inside that price, discounted by whatever the buyer thinks they are worth and adjusted for whatever the buyer thinks of the counterparty. And a buyer looking at a contract with only a few years left has a smaller cushion in front of the same open question that has been sitting there the whole time: what happens on the day the contract ends. The nearer the end of a contract comes, the more of the asset's value depends on the answer to that question, and nobody on either side of the table knows it.
The reader who thinks the fund is long-dated
Here is the error, and it is made almost entirely by readers who have just understood the mechanism. A reader sees that the fund holds a fifteen-year contracted stream, and carries the word long-dated across from the asset to the vehicle. The move is natural and it is wrong. The fund is not long-dated. The fund has a ten year term with two possible one-year extensions, exactly like the manager's other funds, and at the end of it the money goes back to the people who put it in.
The misreading does not cost a rounding difference. The cost is a wrong picture of where the money actually comes from. A reader holding the wrong picture expects a decade and a half of contracted payments to arrive on their behalf. Some of those payments arrive, the asset is then sold with the rest of the contract on it, and the remaining contracted years become somebody else's income entirely. The sale price is doing an enormous amount of the work in the outcome, and a reader who has not noticed the collision has not noticed the sale either.
The tell is easy to spot once it is known. Any sentence that treats a contract length as though it were a holding period has made this mistake. Fifteen years of contract, in a fund that will not exist for fifteen more years, is not fifteen years of income to anybody in that fund.
Somebody calls this a long-dated fund. Is that right?
Which questions does a contracted payment leave completely open?
Three, and they are why the comforting words used about contracted income do not survive contact with the document. A contract fixes a price and a quantity. The contract is silent on everything else, and the three silences sit at three different points in time.
Before the contracted period, the open question is whether the asset produces at all. A signed agreement to buy output is not an agreement that output will exist. If the solar asset generates less than it is expected to, for any reason at all, the contract does not manufacture the missing units. There is no asset yet, so for asset 4 this question is at its sharpest.
During the contracted period, the open question is whether the counterparty actually pays. Payment is the question most often skipped, and it is a credit question rather than an infrastructure question. Somebody has to be good for fifteen more years of payments. A signed document is an obligation, not a payment, and a fifteen-year obligation is a fifteen-year credit assessment of whoever wrote it, renewed silently every single year. A contract removes one uncertainty entirely, the price, and replaces it with two others: what the paying party is good for over fifteen years, and what is left on the day the contract ends.
After the contracted period, the open question is the value of the asset once the document behind its income has run out. The name for it is residual valueWhatever an asset is worth once the contract behind its income has run out., and the region is left blank: this record does not say and neither does anybody else. One thing about the value is structural. The closer an asset gets to the end of its contract, the larger the share of its value that rests on an unanswered question, and a buyer paying for a contract with a few years left is buying a much bigger share of that question than a buyer paying for one with fifteen.
What is one thing a twenty-five year output contract does not say?
How is this different from a listed infrastructure vehicle?
A listed infrastructure investment trust also holds infrastructure assets. Its units are quoted on an exchange, so a holder can sell the unit without waiting for anybody to sell an asset. The listed trust is covered separately, and its conditions are set by the Securities and Exchange Board of India at sebi.gov.in, a regulator that changes them.
The conditions attaching to a listed vehicle are set by a regulator, and they move. The structural contrast is what carries: in a private vehicle of this shape, the investor's exit is the fund's exit, and the collision between a fifteen-year contract and a ten-year fund is a real constraint on the fund rather than something a holder can step around.
Why is a listed infrastructure vehicle treated only as a single sentence of contrast?
Given an hour with an infrastructure fund's asset schedule, what should be asked?
Reading a schedule is the practical end of the matter, and far more people read one than ever commit money to a fund: an analyst covering a manager, somebody at an institution reviewing a quarterly pack, a student handed a case, a person sitting on an advisory committee. Four questions get most of the way, and every one of them can be asked of a single sheet of figures.
The first question is which denominator the headline percentage used. The fund's own Rs 34,90,00,000 is 9.3 per cent of the Rs 3,75,00,00,000 deployed and 10.6 per cent of the Rs 3,30,00,00,000 producing income, and a schedule quoting one figure without naming its denominator has said less than it appears to. The second question is how many years each contract has left, not how many years it was written for. A twenty-five year contract with fifteen to run and one with two to run are entirely different objects and both are described by the same phrase.
Third, ask who the paying party is on every contracted line, and what happens to that line if they stop paying. The words used for these assets sound like the words used for physical certainty, so the vocabulary of infrastructure is worst at prompting that question. The payment is a promise by an entity rather than a property of the concrete. Fourth, ask when the fund has to sell, and put that date next to the contract end date. If the sale date comes first, part of what the schedule shows as contracted income is really a future sale price wearing a contract's clothes. Every one of those four questions is answerable from a single schedule and none of them requires a forecast.
Where the vehicle in this worked case sits
The mechanism is not specific to any country: a contracted output payment, a fund with a fixed life and an asset that outlasts it behave the same way in any market. Nilgiri Real Assets Fund I is registered as a Category II Alternative Investment Fund. The categories, the registration and the conduct rules attaching to them are set by the Securities and Exchange Board of India at sebi.gov.in, and the regulator changes them. The same holds for the listed infrastructure vehicle named above.
Sources
| Source | Document | Site |
|---|---|---|
| Securities and Exchange Board of India | The published framework for Alternative Investment Funds, covering categories, registration, reporting and conduct. Listed infrastructure investment trusts are governed under a separate framework of the same regulator | sebi.gov.in |
| Indian Venture and Alternate Capital Association | The industry body publishing material on private capital in India, including how private vehicles in this market are described | ivca.in |
| International Organization of Securities Commissions | Cross-border conduct principles for collective investment vehicles, relevant because the economic vocabulary used for these funds was imported rather than drafted in India | iosco.org |
Nilgiri Real Assets Fund I, Nilgiri Alternatives Advisors Private Limited, Nilgiri Trusteeship Services Private Limited and Nilgiri Financial Holdings Private Limited are invented.
Educational material. Not advice on any investment, tax, budget or market position.
