How to map Private Credit Seniority: A Reading Order

The starting point is what a claim can reach, not what it is called or what it pays. A name records what somebody wrote on the document. The order of payment governs what happens when there is not enough money for everybody, and that order is set by the security and the contract rather than by the label. The worked case below is one invented fund's own record.

Somebody slides a sheet across the table. On it are three loans, three rates and three amounts, and the question is which of them an analyst would rather be holding. Almost everybody answers by looking at the rates. Reaching for the rate is the natural thing to do. The rate is the only number on that sheet that looks as though it describes the loan itself. It is not. A rate is a price attached to a position, and the position is decided somewhere else entirely, in a security document and a contract that the sheet probably did not show. The order in which to ask is set out below. By the time the rate arrives, what it is sitting on top of is already clear.

The six steps below are a sequence, not a checklist. The distinction matters more than it sounds. A checklist is a list of things to look at, and it can be worked down in any order and ticked off entirely. A sequence is different: each answer changes what the next question is worth asking, and one of the six steps below is worth almost nothing until the five before it are settled. Worked in order, the last step establishes something. Worked backwards, as most readers work it, the last step establishes nothing at all and feels as though it established everything.

Why is the rate the last thing to look at, and not the first?

Consider a shop on an ordinary street closing down. The owner sells the fittings, the stock and the delivery van, and Rs 4,00,000 comes in. Standing outside are the bank that lent against the van, the wholesaler who supplied goods on credit, two staff who are owed a month of wages, and the landlord with three months of arrears. Everybody in that queue knows what they are owed. Not one of them knows what they will get until somebody works out the order. And the order is not decided by who is owed most, who has waited longest, or who is angriest. The order is decided by what each of them has attached their claimSomebody's right to be paid out of that money. to, and by what each of them signed.

Private credit is that queue at a much larger scale, and the list of who stands where is what people mean by a borrower's capital structureThe full list of claims on a business, in the order they are paid.. The arrangement is written down years before anybody needs it. The lender who has taken security has attached its claim to named things. The lender who has not is standing behind, waiting for what is left. Every one of the six steps below is asking a question about one single event: money arrives once, and a list of claims is paid out of it until it runs out. Everything else about a loan is downstream of that one sum. The order below therefore begins there and not at the coupon.

Reading a stressed position from the claims downwards rather than from the coupon inwards is the shape that Stephen Moyer sets out in Distressed Debt Analysis, 2005, and the order below follows it. A charge, what makes one claim rank behind another, a covenant, what payment in kind does to a lender's income, what a unitranche combines: every one of those is a mechanism covered separately, in full, and each is named below and then left alone. The subject is where to look and in what order, and what each answer has just ruled out.

Step 1. What is there to be paid from?

Before anybody is owed anything, there has to be something to be owed out of. So the first question is not about the loans at all. The first question is about the single event that turns a business into a sum of money. Somebody buys the enterprise, or the assets are sold off piece by piece, or a court-supervised process reaches an end and a figure is struck. Whichever of those happens, the answer is one number, and that number arrives once.

Nilgiri Direct Lending Fund I, an invented private credit fund managed by Nilgiri Alternatives Advisors Private Limited, holds eight positions in its loan book, and every figure worked from here on belongs to that fund's own record. One of them, a subordinated position of Rs 20,00,00,000 with no security of its own, stopped paying: the borrower missed a coupon in that fund's Year 3 Q1 and did not cure it. The enterprise behind that borrower was afterwards sold in a distressed sale for Rs 68,00,00,000. The sale is the realisationThe single event that turns a business into a sum of money., and Rs 68,00,00,000 is the whole of what there is.

A claim only means something against a pot. Fixing that number first is what makes every later step answerable. Settling step 1 has already ruled something out: any answer that needs a second event to arrive. Nobody may say that the position will do better once the business recovers, or once a better buyer appears, or once next year's earnings come through. There is no next year in this arithmetic. There is one sale, one figure, and a list of people waiting outside the room. An analyst reaching for a future event to make the numbers work has not finished step 1, and nothing after it will hold.

Whether the borrower could have paid, whether the business was sound, and whether the sale price was a fair one are separate questions. Judging whether a borrower can pay is credit assessment, a subject of its own covered separately, and this order simply takes the realisation as given and works from it. The two questions feel similar and they are not. One asks whether the money will arrive. The reading order asks who gets it when it does.

Step 2. Who has a claim on it, and has the list captured all of them?

The next move is to list everybody standing in the queue. Not the loans that happen to have been shown. Everybody. On this borrower the list has three entries: a bank's senior secured term loan of Rs 60,00,00,000, Nilgiri Direct Lending Fund I's own subordinated position of Rs 20,00,00,000, and the ordinary shares underneath both of them. Two of the three are debts and one is not, and all three are claims on the same Rs 68,00,00,000.

The word claim is doing real work there, and it is deliberately wider than the word loan. A claim is anybody's right to be paid out of that money, and a great many of them are never written on a facility agreement at all. Wages owed to staff. A supplier who delivered goods on thirty days of credit and has not been paid for them. Tax. Rent arrears. A guarantee this business gave for somebody else's borrowing which has now been called. The amount somebody has to be paid because a court said so. None of those has a rate, none of them appears in a lender's own file, and every one of them is standing in the same queue as the loans. Step 2 fails far more often by omission than by arithmetic.

Watch what one missed name does. Suppose, purely as a counterfactual on this same invented borrower, that Rs 20,00,00,000 of trade creditors had also been in the queue, alongside the fund's own claim and with the same absence of security. The bank still takes its Rs 60,00,00,000 first. Rs 8,00,00,000 is left. Two claims of Rs 20,00,00,000 each now share that Rs 8,00,00,000 pro rataShared in proportion to the size of each claim., so each takes Rs 4,00,00,000, and the check is that 60 plus 4 plus 4 makes 68. The fund's recovery falls from Rs 8,00,00,000 to Rs 4,00,00,000. Nothing about the fund's own document changed. One name was added to a list, and the answer halved.

One missed name is why step 2 is a sweep of the whole business rather than a sweep of the lending file, and why it comes before any question about who ranks where. A list that is not finished cannot be put in order. Settling step 2 rules out reading the two loans on the sheet by themselves. A reader whose only documents are those two loans does exactly that.

Step 3. What can each claim on that list actually reach?

Take the finished list from step 2 and go down it one entry at a time, asking a single question of each. Out of the Rs 68,00,00,000, which part can this claim get its hands on? Not how much it is owed. The amount owed is already written down, and the amount owed is not the question. Ask what the claim can reach.

On this invented borrower the bank's term loan has a first chargeA right over named assets that puts one claim ahead of others. over the fixed assets and a first charge over the receivables, so its claim is fastened to named things and is paid out of them before anybody without such a fastening sees a rupee. Nilgiri Direct Lending Fund I's own Rs 20,00,00,000 has no charge over anything at all. The fund's position is unsecuredA claim with no charge, paid only from what is left.. An unsecured claim can reach only the part of the Rs 68,00,00,000 that is still there once the charged claims have been satisfied. The ordinary shares can reach nothing until every creditor in the queue has been paid in full.

The gap between what a claim is owed and what a claim can reach is the single most useful thing this whole order produces. On the fund's position the two numbers are Rs 20,00,00,000 and Rs 8,00,00,000, and the second is a fact about where the claim stands rather than a fact about the borrower. Both numbers are true at once and a reader who only carries the first one around will be surprised every time.

A charge, how one is created, what it attaches to, and the effect of a second charge over the same asset are all mechanism, and every one of them is explained in full in the treatment of seniority within private credit, covered separately. The order works the same whatever the security turns out to be, so a reader who finds any of those words unfamiliar can turn to that treatment and come straight back. Settling step 3 rules out ranking claims by how large each one is. Size decides how much a claim asks for. Size never decides what a claim can reach.

One practical note. In India, charges created by a company are matters of company record with the Ministry of Corporate Affairs at mca.gov.in. The rules on what must be registered, by whom and in what manner are set there, and they change. A lender's file records only what that lender took. The company record is where the whole list of charges sits, so step 3 in real life means checking the company record rather than the lender's own file.

Step 4. In what order do they get paid, and who set that order?

Only now are they ranked, and the ranking falls out of step 3 almost by itself. On this borrower the payment orderThe sequence in which claims are paid until the money runs out. is: first the bank, second Nilgiri Direct Lending Fund I, third the ordinary shares. Write it as a numbered list and keep it. Every later step operates on that list.

The second half of the question is the half readers skip, and it is the more useful half. Who put them in that order? Two things did, and only two. The order of payment is set by the security a claim took and by the contract a claim signed, and by nothing else. Security is why the bank stands first: its claim is fastened to named assets. Contract is the other route. A claim ends up behind another with no security involved at all when somebody has signed a document agreeing to wait. Both routes produce the same result: a claim that gets paid earlier and a claim that gets paid later.

The same test already runs at home without being called that. If a household with a home loan sells the house, the housing lender is paid out of the sale because its claim is fastened to that house, and the card company that lent the same household money for a wedding is not fastened to anything and takes what is left. Nobody in that story asks which lender charged more. The order was fixed years earlier, in what each lender took and what each of them signed, and it is not renegotiated on the day the house is sold.

Step 4 is where the labels come thickest, so a word about vocabulary. Senior, subordinated, second charge, mezzanine, unitranche, and the arrangements lenders sign between themselves to split recoveries behind the scenes: these are all names for particular ways of achieving one of those two routes. Each is set out in full in the private credit treatment of seniority and in the treatments of direct lending, syndicated participations and payment in kind, all covered separately. The work at step 4 is only to establish the sequence and to name for each pair of claims which of the two routes put one ahead of the other. Settling step 4 rules out the size of a loan and the rate it carries. Neither appears anywhere in the two routes.

Step 5. What happens when the money is run down the list?

Step 5 is the arithmetic step and the shortest one. Start with the whole sum, hand it to the first claim until that claim has been paid what it is owed or the money runs out, whichever comes first, then hand what remains to the second, and so on until there is nothing left. The run-down stops on its own. Nobody decides where.

On this invented borrower: Rs 68,00,00,000 arrives. The bank is owed Rs 60,00,00,000 and takes all of it. Rs 8,00,00,000 is left. Nilgiri Direct Lending Fund I is owed Rs 20,00,00,000 and takes the whole of that Rs 8,00,00,000, and no more remains. Nothing is left, so the ordinary shares receive nil. Check the whole thing by adding it back: Rs 60,00,00,000 plus Rs 8,00,00,000 is Rs 68,00,00,000, and the arithmetic closes.

Now put each of those into paise in the rupeeThe share of what was owed that a claim actually received., the only unit that makes two different claims comparable. The bank received Rs 60,00,00,000 on Rs 60,00,00,000 owed, or 100 paise in the rupee. The fund received Rs 8,00,00,000 on Rs 20,00,00,000 owed, or 40 paise. The shares ranked behind both and received nothing at all. Quoting a rupee figure without saying what it is a share of is the commonest way to make one of these positions unreadable. The same Rs 8,00,00,000 is a small number on a Rs 20,00,00,000 claim and would be the whole answer on a Rs 8,00,00,000 one.

The figures here are what one invented realisation produced on one borrower. The figures are not an expected recovery, not a typical recovery, and not figures anybody may carry to another borrower. One realisation on one borrower supports no statement about how much a position of any kind recovers, or about how often borrowers stop paying. Settling step 5 rules out any figure quoted with no base underneath it, including the ones worked above.

Step 6. Now look at the rate. What was it attached to?

Five steps done, and the rate has not appeared once. The rate enters now. Nilgiri Direct Lending Fund I's subordinated position on this borrower was contracted at 16.0 per cent. The 16.0 per cent is the fund's own contracted rate on that one position, and it is not a fact about what borrowing costs anybody in India.

Read against the finished map, that 16.0 per cent stops being a headline. The rate is now a number attached to a specific place: second in the order, with no charge of its own, reaching only what remained after a first charge of Rs 60,00,00,000 had been satisfied out of a Rs 68,00,00,000 sale. The rate did not describe the position. The position is what the rate was attached to, and that only becomes visible once the position has been established by the five steps in front of it.

The next statement has to be phrased carefully, and for a reason. Whether the 16.0 per cent was enough, or too little, or fair, or whether it compensated anybody for anything, is not something an order of payment can settle. The narrower and more useful statement is the one available: here is a rate, and here is the place in the order it was sitting on. Whether that was a sensible arrangement for either side is a judgement, and a judgement takes more than an order of payment.

There is a second thing the finished map yields, and it is the reason experienced readers do step 6 last rather than skipping it. A rate governs what changes hands while the loan is performing. An order governs what changes hands when it is not. Rate and order are two different questions, answered in two different documents, and a reader who has run this order can hold both without confusing them. Settling step 6 rules out the rate read as a description of rank, and that is the error the whole sequence exists to prevent.

What does the whole order look like when it is run end to end?

Here is the sequence in one place, worked on the one invented realisation used throughout. Taken downwards, each entry in the third column is only available because the entry above it was settled first. The dependence of each entry on the one above it is the whole argument for working in order rather than down a list of things to look at.

Six answers, and the last one is the only one that would have been legible on its own if somebody had simply stated it. Being legible on its own is precisely why the rate is worth so little as an opening. On its own, 16.0 per cent records what somebody agreed to pay. Sitting on top of five settled answers, it records what that payment was in exchange for a place in a queue, and the queue can be pointed at.

What goes wrong when a reader starts at the rate instead?

Hold on to the shape of that. The whole order exists for that reason. The two readings of the fund's own position, 40 paise and 85 paise, differ by a factor of more than two, and every input a rate-first reader would have looked at is identical across them. The rate did not move. The amount lent did not move. The borrower did not move. The only thing that moved was an answer to step 4, and step 4 sits three steps before the one most readers open with.

Two things about that second picture deserve saying out loud. The first is that its right hand column was settled without a single rate, the bank's included, and the record does not fix that one. The order did not need it. The second is that the left column is not wrong as a list; it is simply a list of a different thing. Rates ranked by size are a fact about three contracts. Rates ranked by size are not a fact about where anybody stands, and the moment a reader treats one as the other, everything downstream inherits the mistake.

How does anybody actually use this order in a working week?

Four people use it, and they use it for four different purposes. The order looks like a lender's tool and is not only that.

An analyst inside a lender uses it before the money goes out. Handed a borrower's structure, they run steps 1 to 4 on the arrangement as it would stand after their own facility is signed. The position they are about to take is then a settled fact rather than a hope. The number that comes out of step 3, what this claim would be able to reach, is the one that goes into the paper. Doing that before signing is the same exercise as doing it after a default, run on a realisation that has not happened yet, and it is the only version of the exercise anybody can act on.

Somebody monitoring an existing book uses it when something moves. A borrower breaches a term in its agreement, or asks for a change, or offers more security in exchange for something, and each of those is a proposal to alter an answer somewhere in steps 2 to 4. Restructuring a stressed position, and what a covenant breach actually starts, are covered separately in the treatment of workouts. The order gives that reader the ability to say precisely which step a proposal moves, and therefore what else moves with it.

An investor reading a private credit fund's report uses it in a smaller and more honest way. The report will name what the fund holds and will use the labels covered separately: senior secured, subordinated, unitranche, second charge, a participation in a facility shared with other lenders. Running this order turns each of those from a word into a question about reach, and it guards against what a report cannot guard against: reading the rates down a schedule and inferring an order from them.

And there is a version of this at kitchen-table scale. Anyone who has signed as a guarantor for somebody else's borrowing has taken a place in a queue that was probably never mapped: not a claim on the sale of anything, but an obligation that arrives if a particular thing happens. Somebody at home who has done steps 1 to 4 on their own arrangements, including the loan against the house and the loan against the vehicle, knows something about their position that no interest rate on any statement will tell them.

What does the finished map establish, and what does it not?

Be strict about this. The finished picture is persuasive, and a persuasive picture is exactly the sort of thing a reader over-reads. The map answers a narrow question completely and answers nothing else at all.

One last boundary. Standing earlier in the order is not thereby a place worth being in, security is not thereby worth what it costs, and none of these positions is thereby a thing anybody should hold. A claim paid earlier is paid earlier: that is mechanism, and mechanism is all the order establishes. Turning it into a preference would require knowing what each position was bought or lent at, what else was on offer, and what the person doing the buying was trying to do, and none of those is available here.