Clawback: Returning Carry That Was Paid Too Early

A clawback is what happens when a manager has been paid a share of profit and the fund's whole life turns out not to have produced it. Money already received has to come back. The idea only exists because some arrangements pay carried interest deal by deal, before anybody can know what the fund as a whole will return.

Consider something that has nothing to do with funds. A householder hires a painter to do four rooms and agrees to pay as each room is finished. Room one is done, the painter is paid. Room two is done, the painter is paid. Then room three turns out to have plaster underneath that was never going to hold paint, and fixing it swallows more than the whole job was worth. At the end of it the painter has been paid for two rooms out of a job that, taken as one job, lost money. Nothing dishonest happened, nobody was careless, and yet somebody is holding money that the completed job did not produce. The only question left is whether the agreement with the painter said anything about giving it back.

A clawback is that situation with larger numbers and a written contract behind it. A private fund pays its manager two things: a management fee for running the fund, and a share of the profit for producing one. The share of profit has a name, carried interestThe manager's share of a fund's profit, paid out of what the fund distributes to investors., and a clawback only ever reaches that second payment. The fee is not clawed back.

How can somebody be paid a share of profit that turns out not to exist?

Because profit at a fund is a whole-life number and the cash arrives a bit at a time, years apart. A private fund buys companies one at a time and sells them one at a time. The first sale might land in year six and the last one in year eleven, and there is no moment in between where anybody can look at the fund and say what it finally made. The manager, meanwhile, is a business with salaries to pay in every one of those years.

So the contract has to decide when the manager gets its share. There are two ways to write it, and the entire subject falls out of the difference between them.

The first way is whole-of-fundAn arrangement paying carried interest only after all the capital the fund has drawn has been returned to investors.. Every rupee of capital that was ever drawn from investors, for investments, for the fee and for expenses alike, has to go back to them before the manager receives one rupee of carried interest. The manager waits until the arithmetic is finished, or nearly finished, and only then is paid.

The second way is deal by dealAn arrangement paying carried interest on each realisation as it happens, rather than waiting for the fund's whole life to finish.. Each time one holding is sold, the profit on that one holding is worked out and the manager takes its share of it there and then. A sale that turns one holding into cash is called a realisationThe event that turns one holding into cash, usually a sale., and under this arrangement a realisation is a payday.

Only the second way pays before the answer is known, so only the second way can ever pay too much. Under the first, the manager is paid out of a number that is already complete, so there is nothing that can later shrink underneath it. Under the second, the manager is paid out of the best available number so far, and the best available number so far can get worse. A clawback is the clause that says what happens when it does.

Which of the two arrangements can overpay, and which one structurally cannot?

Put the question as a single test and the answer stops being a matter of opinion. Ask: at the moment the manager is paid, is every rupee of drawn capital already back with the investors? If the answer has to be yes before any payment can happen, then no payment can ever exceed what the fund produced. The fund has already produced it. If the answer can be no, then a payment can be made out of an early win that a later loss undoes.

A clawback is therefore not a penalty and not a sign that anything went wrong. A clawback is a correction term. An arrangement that deliberately pays early needs a mechanism for unwinding an early payment, and an arrangement that pays only at the end does not need one at all. Which of the two a fund carries is a term, negotiated and written down before a single rupee was drawn, and it is one of the first things worth finding in a fund's documents.

What would this fund have paid its manager under the other arrangement?

The worked case runs on one fund. Nilgiri Growth Partners Fund II, an invented closed-end growth and buyout fund, is managed by Nilgiri Alternatives Advisors Private Limited, with Nilgiri Trusteeship Services Private Limited as trustee and Nilgiri Financial Holdings Private Limited as sponsor. Its investors and the manager between them committed Rs 5,00,00,00,000. Every figure below is measured at the same single date, the end of Fund II's Year 9 Quarter 2.

Fund II carries a whole-of-fund arrangement. By that date it had drawn Rs 4,80,00,00,000 from its investors and distributed Rs 4,38,00,00,000 back to them, so it was Rs 42,00,00,000 short of returning what it had called. Its manager has therefore received no carried interest at all, its clawback exposure is nil, and the account that would hold money back against a clawback is empty. The nil exposure is Fund II's real position, and it is the number to keep hold of against the arrangement the fund does not carry.

Here is the counterfactualA worked case of what would have happened under a term the fund does not actually carry. It did not happen.. Suppose the same fund, with the same nine holdings, the same purchases and the same sale prices, had been written deal by deal instead. Then each realisation would have paid the manager 20.0 per cent of the profit on that one holding, in the quarter after the cash arrived, and the three realisations that produced a profit by Fund II's Year 8 Q2 would have run like this.

The arithmetic in that table is worth doing yourself: 3.60 plus 26.60 plus 18.00 crore is Rs 48,20,00,000. All three of those payments belong to a fund that had still not returned its investors' capital. Under the arrangement Fund II actually carries, the manager has been paid nothing. Under the one it does not carry, the manager would be holding Rs 48,20,00,000 while investors were still waiting to get back what they put in. The gap between the nil actually paid and the counterfactual Rs 48,20,00,000, on the same fund on the same day, is the entire reason a clawback clause exists.

Notice something the table quietly does. Holding 5, Palar Foods Private Limited, invented, cost Rs 35,00,00,000 and was written off in full in Fund II's Year 6 Q4, returning nothing. A write-off produces no cash and therefore no realisation to pay carried interest on, so holding 5 never appears in the counterfactual at all. The counterfactual takes a share of each profitable exit and nets nothing against it. A whole-of-fund arrangement nets everything by construction. Such an arrangement cannot pay until all the capital is back, and all the capital coming back means every loss has already been absorbed.

What exactly comes back, and who writes the cheque?

Under the counterfactual, the exposure is not one payment. The exposure is a running total that grows every time a profitable realisation happens early and does not fall at all until capital finally goes back to investors. Fund II's counterfactual would have added Rs 3,60,00,000 in Year 6 Q4, Rs 26,60,00,000 more in Year 7 Q3 and Rs 18,00,00,000 more in Year 8 Q2, and at no point in that stretch did the fund finish returning what it had called.

So what would actually come back, and from whom? The obligation sits on whoever received the carried interest: the manager, Nilgiri Alternatives Advisors Private Limited, invented, and behind it the sponsor, Nilgiri Financial Holdings Private Limited, invented. The obligation is not one of the fund, not one of the trustee and not something the remaining holdings can be sold to cover. Somebody who has already banked money has to send money back. Sending banked money back is the whole of the mechanism, and that is why the clause is short to read and hard to enforce.

The amount is easier to see by following the cash rather than the clause. Under the counterfactual, the fund's four cash distributions would have arrived the same way but a slice of three of them would have gone to the manager before the rest reached investors. Rs 63,00,00,000 becomes Rs 59,40,00,000 to investors. Rs 2,03,00,00,000 becomes Rs 1,76,40,00,000. Rs 1,50,00,00,000 becomes Rs 1,32,00,00,000. The counterfactual stops at Year 8 Q2, so the fourth distribution of Rs 22,00,00,000 in Year 8 Q4 is untouched. Add the four and investors would have received Rs 3,89,80,00,000 rather than the Rs 4,38,00,00,000 they actually received, leaving them Rs 90,20,00,000 short of the Rs 4,80,00,00,000 drawn rather than Rs 42,00,00,000 short.

Now run the clawback on that. Hand the whole counterfactual Rs 48,20,00,000 back and investors go from Rs 3,89,80,00,000 to Rs 4,38,00,00,000, exactly where the real whole-of-fund arrangement already stands. A full clawback does not fix the fund; it returns the two arrangements to the same place, and the Rs 42,00,00,000 that is still outstanding afterwards was never the manager's to give. That distinction matters more than anything else in the subject. A clawback corrects a payment. A clawback does not manufacture investment profit that the fund did not make.

What is the escrow holding, and what is it not holding?

An obligation to send money back is only worth what stands behind it. So the contract usually reaches for something stronger than a promise, and the thing it reaches for is an escrowMoney held back from a payment and kept in a separate account against a possible future obligation.. Think of the retention a household holds back from a builder: the last slice of every stage payment is not handed over on the day, it sits with a third party, and it comes out when the job is signed off. The retention is not extra money. The retention is the same money, held somewhere the builder cannot spend it.

Fund II's own documents set that retention at 30.0 per cent of every carried interest payment, and the clawback itself is stated to fall due at the end of the fund's term and to be computed net of taxes actually borne. Both are Fund II's own contracted terms, written before any capital was drawn, and neither is a market convention.

Apply that 30.0 per cent to the counterfactual and the arithmetic is quick. Of the Rs 48,20,00,000 the manager would have received by Year 8 Q2, Rs 14,46,00,000 would sit in escrow and Rs 33,74,00,000 would already have been released. Seventy paise in every rupee of a counterfactual carried interest payment is gone before anybody knows whether it will be owed back. Held money can be reached. Released money can only be asked for.

Two divisions are being run there and a careless reader will merge them, so name the denominator in both. Against the Rs 48,20,00,000 of counterfactual carried interest, the escrow is 30.0 per cent, simply the contracted rate read back. Against the Rs 42,00,00,000 of capital the fund has not yet returned, the same Rs 14,46,00,000 is 34.4 per cent, and the gap between the two is Rs 27,54,00,000, being 65.6 per cent of that second figure. A number stated without its denominator says nothing, and the two denominators here are the counterfactual Rs 48,20,00,000 and the real Rs 42,00,00,000.

Why is the computation net of taxes actually borne?

Because money that was never kept cannot be returned. If a manager received a rupee of carried interest and part of that rupee went straight to the tax authority, the manager is holding less than a rupee. A clause demanding the gross figure back is demanding something the recipient cannot produce, and a clause that cannot be performed is worth about as much as no clause at all.

So the fund's own documents say the clawback is computed net of taxes actually borneReduced by tax the recipient really paid and cannot recover, rather than by a rate assumed on paper.. Both halves are load-bearing, so read the phrase slowly. Actually means tax that was really paid, not a rate somebody assumed. Borne means tax the recipient could not get back, so a tax already refunded or credited is not deducted. The phrase is doing real work rather than softening the obligation, and its effect is to size the clawback against what the recipient can still lay hands on. How any of that tax is computed belongs to a different subject, covered separately.

Are the two figures near Rs 48 crore the same number twice?

The two figures are not the same number, and a reader working from memory will merge them, so pin them apart now. One of them happened and one of them did not, and they belong to two different invented funds.

Nilgiri Growth Partners Fund I, invented, is a completed fund. Fund I wound up at the end of its own Year 10 Quarter 4, having drawn Rs 2,40,00,00,000 from investors and distributed Rs 4,80,00,00,000 back, so the profit above return of capital over its whole ten-year life was Rs 2,40,00,00,000. Its manager received Rs 48,00,00,000 of carried interest, and that is exactly 20.0 per cent of that profit: multiplying Rs 48,00,00,000 by five gives Rs 2,40,00,00,000 to the rupee. Fund I ran a whole-of-fund arrangement, its whole-life result was known before the manager was paid, and it therefore never had a clawback exposure of any size. That figure is real, within this invented record, and it is the only one of the two that is.

Rs 48,20,00,000 is the other one, and it never happened. The figure belongs to Fund II, a different fund with a different set of nine holdings, and it is what a deal-by-deal arrangement would have paid by Fund II's Year 8 Q2 had Fund II carried one. Fund II carries whole-of-fund and has paid nothing. The two figures are Rs 20,00,000 apart, which is four tenths of one per cent of the smaller of them. The near match is a coincidence of two unrelated arithmetics and it means nothing at all. One is 20.0 per cent of one fund's completed lifetime profit. The other is the sum of three separate slices of three separate holdings in another fund that has not finished.

Why would anybody write an arrangement that needs a clawback at all?

Because waiting is expensive for the people doing the work, and this invented record shows exactly how long the wait can be. Fund I ran its whole ten years, made four distributions, and the entire Rs 48,00,00,000 its manager received arrived in the last one. Ten years of salaries, rent, diligence on 412 opportunities that mostly went nowhere, and the share of profit lands once, at the end. Fund II is now in its ninth year and its manager has received no carried interest at all.

Set against that, an arrangement that pays as each realisation lands is not a trick. Paying as each realisation lands is a different answer to a real timing problem: it moves money to the manager when the manager has produced something, rather than years after. The clawback is what makes that early payment honest. The clause says the early number is provisional and the whole-life number is the one that settles. Neither arrangement is fairer, better for investors or safer than the other, and which is more common is a question of market practice rather than of contract. The two arrangements are two different contracts with two different timings, and each has a mechanism that fits it.

The trade the reader should be able to name is this. Whole-of-fund makes the manager wait and needs no correction. Deal by deal pays sooner and needs one, and the correction is only as good as what secures it. The trade is the whole subject, and the arithmetic above is the same trade with rupees attached.

How does somebody who reads fund documents for a living use this?

Not by looking for a clawback, and not by treating its presence or absence as a mark for or against anything. The use is narrower and more mechanical, and it runs in four steps that go in one order.

One term decides whether any of the rest matters, so step one is to find which arrangement the document sets. If carried interest is payable only after all drawn capital has been returned, the clawback clause is a tidy-up for edge cases and the reader can move on. If carried interest is payable on each realisation, the clause is load-bearing and steps two to four apply.

Step two is to find when the obligation is tested. Fund II's own wording puts it at the end of the term: once, not at every realisation. A reader who assumes it is tested continuously has invented a protection the document does not contain.

Step three is to find the yardstick the obligation is measured against. Fund II's wording measures the manager's entitlement over the whole life of the fund, so the excess is the gap between what was received and what the whole life justifies. The excess is a subtraction, not a judgement, and a subtraction is why the arrangement is capable of settling.

Step four is to find what secures it, and then to divide. An escrow rate on its own is a number without meaning; it becomes a reading only when placed against something. A person sitting on an investor advisory committee, a treasury analyst reviewing a commitment, or somebody at a fund of funds doing the same work on twenty vehicles will all do the same division: how much is held, against how much could be owed, and against how much capital is still outstanding. On this invented fund that division is Rs 14,46,00,000 against the counterfactual Rs 48,20,00,000, being 30.0 per cent, and Rs 14,46,00,000 against the real Rs 42,00,00,000, being 34.4 per cent. Two divisions, two denominators, two different sentences, and naming which one is meant is the whole of the skill.

One more thing separates somebody who has read a few of these from somebody who has not, and it is who the obligation actually sits on. A clause that names the manager entity and nothing else is a different object from one that reaches the sponsor standing behind it, and different again from one that reaches the individuals who received their share. None of that is visible from a rate. Only the clause shows it. The clause, not the rate, is what a reader looks for.