NAV Financing vs Preferred Equity: Two Different Levels

One is borrowing by the fund itself, secured on everything it still holds and sized against the net asset value (NAV) of those holdings, repaid with interest out of them. The other is a claim inside the capital structure of one company, ranking behind every creditor of that company and ahead of its ordinary shares. Both are described as raising money against something already held, and they sit at two entirely different levels.

Put those two sentences side by side and the trouble is already visible. Both arrangements are described in almost identical words: money raised against something a fund already has. The shared description is why the two names end up in one question, and it is why the comparison so often ends in a belief that is simply not true. The difference between these two is not what they cost, how they are documented or how clever they are: it is where the claim sits and what that claim is entitled to. Fix the level first and everything after it is arithmetic. Leave the level unfixed and every number attaches to the wrong thing.

Why does this comparison go wrong before it even starts?

Start away from funds entirely. Imagine a household that has a house it lives in and a small stake in a cousin's printing business. Two very different arrangements are open to it. The household could go to a bank and borrow against the house: the loan is the household's own obligation, the bank has a claim over the house, and the household pays it back with interest out of whatever it earns. Or the cousin could issue the household a special class of shares in the printing business, ranking ahead of the cousin's own ordinary shares but behind the paper supplier and the bank that lent to the business.

The loan and the special shares are not two flavours of the same thing. One is a debt of the household, sitting above everything the household has. The other is a position inside somebody else's business, and it can never reach the house, the household's salary or anything else the household holds. If the printing business fails, the special shares are worth nothing and the house is untouched. If the household stops paying the bank, the bank comes for the house and the printing business carries on without noticing. Same phrase, two completely different levels, and the level is what decides who can come for what.

The two arrangements compared here are exactly that pair, in a private fund's clothing. A loan against a fund's holdings is a fund-level arrangementOne entered into by the fund itself, above all its holdings.: the fund borrows, the fund owes, and the lender's claim sits above the whole portfolio. Preferred equityA claim ranking behind creditors and ahead of the ordinary shares. is a position-level arrangementOne inside the capital structure of a single company.: it lives inside one company's capital structure, and a fund meets it by holding it as one position among many.

The word preferred is used at both levels, and the overlap is worth saying out loud rather than passing over. Arrangements described as preferred are written at fund level too. No fund-level preferred instrument is fixed anywhere in this record for a Nilgiri vehicle, so no rate, no size and no priority return can be quoted for one. The record does fix the position-level case, twice over, as a labelled counterfactual on a realisation that actually happened, and that is where the mechanism is worked out below.

What is each one actually a claim on, and what is it entitled to?

Now the precise version. A loan against a fund's holdings is debt, and the mechanism that follows is usually taken on trust when it should not be. The loan carries a principal, a rate of interest and a date, and those three things together are the whole of what the lender is entitled to. The loan is a fixed obligation of the fund, it ranks ahead of the investors in being paid, and it does not care in the slightest what the portfolio turns out to be worth. If the five remaining holdings of Nilgiri Growth Partners Fund II, invented, eventually produce far more than anyone expected at its Year 9 Q2 record date, the lender still receives its principal and its contracted interest and not one rupee beyond. A lender's entitlement does not rise with the portfolio, and that is the single most useful sentence anybody can carry away about debt.

Preferred equity is not debt, and calling it debt with a nicer name is how readers go wrong. There is no repayment schedule that the company breaches by missing. A preferred claim sits inside one company's capital structure, behind every creditor of that company and ahead of that company's ordinary shares, and what it is entitled to is written into the document that created it rather than into a schedule of payments. The document might fix a priority return that accrues whether or not it is paid. The same document might allow the holder to share in what is left over once that priority return is satisfied, or it might not. The document might fix no date at all. A preferred claim's entitlement is a question about a document, not a question about an instrument type.

The dependence on the document is why no rate and no size attaches to preferred equity here. The record fixes preferred equity nowhere as a live position: Nilgiri Direct Lending Fund I, invented, holds eight positions and every one of them is a loan of some description. Preferred equity appears in this record only as a labelled counterfactual, where a claim of Rs 20,00,00,000 is put in place of a subordinated loan of exactly that size so the two can be compared on identical facts. A rate or a priority return supplied here would be a number that nothing else in the record could reconcile against. An honest blank is the better outcome.

Read the two columns of that figure downwards rather than across and the shape of each arrangement becomes obvious. A fixed obligation is fixed in both directions, so the lender's column barely changes between the good state and the bad one. The obligation does not grow when things go well, and it does not shrink when they do not. The lender stands in front of the investors, so the only thing that changes in the bad state is who absorbs the shortfall, and the answer is the investors. The preferred column changes completely between the two states, and in both of them the answer begins with the same six words: whatever is left after the creditors.

Where does each one sit when the money finally arrives?

Money arriving is the moment every claim stops being a description and becomes a number. Both arrangements have an order attached to them, and the two orders are separate objects that happen to be described with similar words, so they have to be held apart. The order in which a fund pays its own investors is one thing, covered separately. The order in which one company's sale proceeds reach that company's own creditors and shareholders is a completely different thing, and it is the one that decides what a preferred claim receives.

The record fixes that second order exactly, on a real event in the invented book of Nilgiri Direct Lending Fund I. Position 5 of that fund was subordinated unsecured debt of Rs 20,00,00,000 at a contracted 16.0 per cent. The borrower missed a coupon in that fund's Year 3 Q1 and did not cure it. The enterprise behind it was later sold in a distressed sale for Rs 68,00,00,000, and the claims lined up in a numbered order: first a bank's senior secured term loan of Rs 60,00,00,000 with a first charge over fixed assets and receivables, which recovered the whole Rs 60,00,00,000; then this fund's unsecured Rs 20,00,00,000, which recovered the residualWhat is left of a realisation after the claims ahead have been paid. Rs 8,00,00,000, being 40 paise in the rupee; then the ordinary shares, which received nil. Sixty plus eight is sixty eight, and the arithmetic closes.

Notice that a preferred claim does not appear anywhere in that ladder as a matter of fact. The claim appears as a seat, and the seat is the whole of what the label means. A fund holding preferred equity in a company is not holding a promise from the fund's manager, a protection for the fund's investors, or a claim on the fund at all. The fund is holding one line in one company's own order of payment, and if that company never produces money, that line never produces money either.

Does it actually matter whether the fund held debt or preferred equity?

Here is where most comparisons of these two arrangements quietly cheat. Such comparisons assert that ranking matters, wave at the idea that debt sits ahead of equity, and move on. The same Rs 68,00,00,000 can be run four ways on facts that are otherwise identical, so the assertion can be tested rather than repeated.

The identical outcome is the opposite of what the ranking difference seems to promise. The ranking difference between a loan and a preferred claim is real, but a difference in ranking only produces a difference in rupees when there is somebody standing in the gap. With the bank in front and the ordinary shares behind and nothing at all in between, the two instruments are indistinguishable on this realisation. A distinction that does not change a number on the facts at hand has not yet earned any attention.

So give it somebody to stand in the gap.

The pair of counterfactuals is the only place in this record where the debt against preferred distinction becomes a number, and the conditions for it are narrow enough to be worth stating bluntly. The distinction needed a shortfall, deep enough that not everybody could be paid. A third claim was needed as well, to force the residual to be shared. Remove either condition and the two instruments produce identical outcomes on identical facts.

The best teaching in the pair is visible only from a third party's side, and almost nobody looks there. The trade creditors' own recovery doubles from Rs 4,00,00,000 to Rs 8,00,00,000 without them doing anything whatsoever. They did not renegotiate, did not take security, did not lend more and did not even know the substitution happened. Their position improved because a competitor was removed from the creditor queue: converting one claim of Rs 20,00,00,000 from a debt into a share took it out of the line entirely. A ranking is not a property of an instrument. A ranking is a statement about who else is standing in the queue.

What does the fund-level arrangement do to the reported columns?

Now climb back up a level, all the way above the portfolio, and look at what a loan does to a fund's own reporting on the day it lands. The fund-level arrangement moves a number that people actually watch, and that is precisely why it needs more care than the other half of the comparison.

A counterfactual is meaningless without the record it departs from, so the position comes first, stated plainly. Nilgiri Growth Partners Fund II had drawn Rs 4,80,00,00,000 from its investors and distributed Rs 4,38,00,00,000 back to them as at its Year 9 Q2 record date. The fund is therefore Rs 42,00,00,000 short of returning the capital it has called. Four of its nine holdings are gone entirely, three sold and one written off in full; five are still held, being holdings 4, 6, 7, 8 and 9, carried at Rs 1,08,00,00,000, Rs 21,00,00,000, Rs 39,00,00,000, Rs 81,00,00,000 and Rs 33,00,00,000 respectively. The five held positions sum to Rs 2,82,00,00,000, the fund's net asset valueThe reported value of what a fund still holds. at that date. Not one rupee of that Rs 2,82,00,00,000 has ever been sold to anybody.

Watch what happens to the reported columns. Cumulative distributions rise from Rs 4,38,00,00,000 to Rs 4,94,40,00,000, so distributions to paid inCumulative distributions divided by cumulative capital contributed. rises from 0.9125 to 1.0300. The ratio crosses 1.0000, the single number treated as the line between a fund that has returned its investors' money and one that has not, and it crosses without a single holding having been sold to anybody. Meanwhile the holdings are now carried net of what is owed on them, so the estimate column falls from Rs 2,82,00,00,000 to Rs 2,25,60,00,000, taking residual value to paid in from 0.5875 down to 0.4700.

Add the two columns in each state and the point lands on its own. As reported: 0.9125 plus 0.5875 is 1.5000, exactly. In the counterfactual: 1.0300 plus 0.4700 is 1.5000, exactly. The same Rs 7,20,00,00,000 of total value over the same Rs 4,80,00,00,000 paid in, split two different ways, summing to the same figure both times. Every rupee that appeared in the cash column was removed from the estimate column, and no rupee of value came into existence at any point in the transaction. An account that shows the first number rising and does not show the second falling has sold a headline.

What is the mistake this comparison is most likely to produce?

There is one more thing worth noticing in that middle panel, and it is easy to skim past. In between, the fund still has to sell five businesses. The loan did not shorten that list, improve any of those businesses, or find a buyer for any of them. The loan bought time and moved cash forward. Both are real things a manager may want, and neither is the same thing as a result. Whether any manager should do it is a judgement, and a judgement is a different question from the mechanism.

So what does neither of these two arrangements do?

After all that arithmetic, the shortest true sentence about both of them is a negative one. Neither arrangement creates any value at all. What each one changes is when money moves and who is owed what, and those are the only two things either of them changes.

Take the fund-level loan first. Before it, Nilgiri Growth Partners Fund II holds five businesses carried at Rs 2,82,00,00,000 in total at its Year 9 Q2 record date. After it, the fund holds the same five businesses, carried at the same five figures, and a lender holds a claim ahead of the investors. The businesses do not know the loan happened. No customer of holding 7 changed their mind, no plant at holding 8 was built, no order book anywhere moved by a rupee.

Now the position-level claim. Converting a subordinated loan of Rs 20,00,00,000 into a preferred claim of Rs 20,00,00,000 does not put a single extra rupee into the enterprise. Every state above ends at Rs 68,00,00,000, because Rs 68,00,00,000 is what the enterprise sold for. The only thing that moved between the states was who was standing in the queue, and the sums closed at sixty eight every time.

Both figures on that identity check are worth carrying around. Rs 7,20,00,00,000 over Rs 4,80,00,00,000 does not care which column the value is sitting in, so total value to paid in for Nilgiri Growth Partners Fund II stood at 1.50 times at its Year 9 Q2 record date and stands at 1.50 times in the counterfactual too. The only honest thing either arrangement changes about that fund is the timing of cash and the list of people with a claim on it.

How does somebody actually use this distinction at work?

Four different people meet this pair of arrangements and each of them needs a different question answered, so it is worth walking round the table.

A lender looking at the fund-level arrangement asks one question above all others: what am I secured on, and how does it turn into cash? The answer is five unlisted businesses, held by a fund with six quarters left of its contracted ten-year term, none of which has a price until somebody negotiates one. No company is being analysed there. The lender is looking at a portfolio it cannot see through, held by a manager whose own clock is running, and everything it can recover has to come from sales that have not been agreed yet. The framing that a distressed-claims analyst would use here, of the kind Moyer sets out in Distressed Debt Analysis, is the right one: work out what the asset produces, then work out who is standing in front of the claim when it does.

An analyst reading a fund's quarterly reporting asks a different question entirely: has this column moved because something was sold, or because something was borrowed? Those two produce identical-looking movements in distributions to paid in and they mean opposite things. A sale removes a business from the portfolio and settles a number forever. A loan leaves the business exactly where it was and adds an obligation. If a fund's cash column rises and its estimate column falls by the same amount on the same date with no realisation named, the analyst has learned something that no headline multiple will tell them.

An investor in a fund asks the question the fund cannot answer for them: how much of my reported total has actually been sold to somebody? For Nilgiri Growth Partners Fund II at its Year 9 Q2 record date the answer is Rs 4,38,00,00,000 of the Rs 7,20,00,00,000, with Rs 2,82,00,00,000 still an estimate. In the labelled counterfactual it becomes Rs 4,94,40,00,000 of the same Rs 7,20,00,00,000, with Rs 2,25,60,00,000 an estimate and Rs 56,40,00,000 owed to a lender out of it. The reader who tracks both halves is never surprised by either.

And anybody at all, with no fund anywhere near them, meets the position-level version constantly. When a shop closes owing money to the wholesaler, the bank and the landlord, somebody works out that order, and the person at the back gets what is left, which is frequently nothing. The wholesaler who quietly agreed to take shares instead of an invoice has moved from the middle of that queue to the very end of it. Nobody announced the change, the paperwork looked like a courtesy, and the whole of what happened is a change of seat.