The Issuer: Who Borrows and What That Says About the Bond

The issuer is the party that has agreed to make a bond's payments, and it is the one term of the promise that cannot be checked by adding anything up. Two borrowers can write identical schedules for identical lengths of time and still be lent to at different rates. A lender asks for more wherever being paid in full and on time is less certain.

A schedule of payments is worth exactly what the party behind it can deliver. Four terms fix a bond, and three of them sit on the paper in a form open to verification: the amount owed, the dates it moves on, and the rate written into the contract. The fourth is who owes it, and there is nothing to add up. The gap where that fourth term sits is why credit is a subject at all, and every later method that measures, prices or ranks credit exists because a schedule alone cannot close the gap.

Who is the party that actually owes the money?

Take the ten year 8.50 per cent bond once more. The document sets its face amount at Rs 1,000/-. The document lays out ten dates, one a year. And it fixes 8.50 per cent as the contracted rate. Rs 85.00/- then lands on each of those ten dates. Every rate in this material runs on a once-a-year compounding clock, and that matters: the identical figures applied twice a year would not price the identical schedule.

Each of those four terms costs something different to verify. The face amount is printed and read off. The dates are printed and counted. The contracted rate is printed, and multiplying it by the face amount confirms the coupon amount: 8.50 per cent of Rs 1,000/- is Rs 85.00/-, and if the document said anything else, one line of arithmetic would catch it. Three of the four terms are self-checking, and the fourth is not checkable from any figure on the document at all. Who has agreed to make those payments is a name, and no amount of adding will establish whether the name is good for it.

The plain version is worth sitting with first, without any of the vocabulary. Two neighbours each hand over a signed slip promising Rs 85.00/- every year for ten years and Rs 1,000/- at the end. The slips are word for word identical. One neighbour draws a steady salary and has never missed a payment on anything. The other has just put every rupee into a new venture that may or may not work. Nobody would pay the same for the two slips, and there is nothing written on either slip that says which is which. The slip describes the obligation completely. The slip describes the person behind the obligation not at all.

So the issuerThe party that signed up to make the payments on a bond, and the party a holder has a claim against if the payments stop. is a term of the promise rather than a fact about the surroundings. The issuer is written into the document alongside the face amount and the rate, and the name is as much a part of the contract as either figure. The issuer behaves differently for one reason only: the other terms are quantities, and the issuer is a party.

What does knowing the borrower not change?

The half that surprises people comes first. Getting it the wrong way round quietly wrecks everything that follows. A reader who has just learnt that the borrower matters will often conclude that a weaker borrower is handled somewhere inside the calculation, as though there were a step in the sum that allowed for the party. There is no such step. The identity of the borrower changes none of the arithmetic, and a schedule discounted at a given rate on a given compounding convention comes out at the same figure whoever signed it.

Run the ten year 8.50 per cent bond twice. Same Rs 1,000/- of face amount both times. Same ten annual dates. Same Rs 85.00/- landing on each of them, with the face amount alongside the last one. Discount every one of those amounts at 8.50 per cent a year, compounding once a year throughout, and this record puts the result at Rs 1,000.00/-. Now run it again with the document signed by somebody else entirely. Every input is the same, so every step is the same, so the figure is the same: Rs 1,000.00/-, and the difference between the two runs is Rs 0/-. How that sum is actually taken is covered separately; what matters here is that nothing in it has a slot for a name.

The insistence is what makes the rest of the subject tractable. If the party had to be handled inside the sum, every price would need a private adjustment nobody else could reproduce, and two people looking at the same bond would disagree without being able to say where. Keeping the party out of the arithmetic and inside the rate gives the disagreement somewhere to live. Two analysts can agree on every step and still put in different rates. The argument is then about the borrower, and the borrower is what it was always about.

Everything else the borrower does not change is worth listing plainly. Each item on the list is a place a reader might expect the borrower to intrude, and it does not. The borrower does not change the face amount, a figure the document states. The borrower does not change the dates. The borrower does not change the contracted rate, written down before anybody had an opinion about anything. The borrower does not change what the coupon amount comes to. The borrower does not change the compounding convention. None of the five is a matter of opinion, and the borrower is the only term of the promise that is.

What does knowing the borrower change?

One thing, and it is the thing everything downstream is built on. The borrower changes the rate at which a promise of that length is offered at all. Not the sum, not the dates, not the face amount. The rate.

The reason is worth stating as a reason rather than as a rule to be memorised. A memorised rule gets applied backwards within a week. Lending is parting with money now for a schedule of payments later. Wherever the chance of that schedule arriving in full and on time is lower, a lender parts with the money less willingly, and the only lever available for making it willing again is the rate. So a borrower whose repayment is less certain does not get lent to at the same rate and then quietly compensated somewhere else. Such a borrower has to offer more, or it does not get lent to.

Turn that around and the direction stays honest in both readings. A borrower whose repayment looks more certain can borrow at less. Neither statement says anything about what will actually happen: certainty here is a judgement being made by lenders at the moment of borrowing, not a fact about the future, and blurring the two turns a definition into a forecast.

Percentage points and basis points, which are not the same word twice

Units get mangled all over this subject at exactly this point, so one paragraph of slowing down earns its place. Subtracting one rate from another gives a distance, and a distance between rates is measured in percentage pointsThe unit that results when one rate is subtracted from another. The unit keeps that distance separate from a per cent of some amount, and a per cent of an amount is a different quantity entirely.. One percentage point is 100 basis points. The two words are never swapped and never used to mean each other. "Per cent" invites the reader to ask "per cent of what?" and there is no answer. A gap between two rates is not a proportion of anything.

Take this record's own pair. The five year government spot rate is entered at 6.90 per cent a year. The five year issue of Palash Cements Limited, an invented cement maker, is entered at 9.10 per cent a year. Subtract: 9.10 less 6.90 leaves 2.20 percentage points. Multiply by 100 and the same distance reads 220 basis points. Run it the other way as a check. The reverse costs nothing and catches a slipped decimal every time: 6.90 plus 2.20 points gives 9.10, so the two ends and the gap agree.

Notice the claim the picture makes, and the claim it refuses. The drawing puts three rates on one horizontal scale because all three are rates per year on the same compounding convention, so they lie against a common ruler and their distances read off. The drawing does not say the three are the same kind of object, and they are not. The difference between the three kinds is the next thing to settle.

Which kinds of borrower issue bonds at all?

The useful way to sort borrowers is not by the names a reader happens to recognise. Names are the least durable thing about this subject. The durable sort is by what the obligation actually rests on. Four kinds cover almost everything.

A sovereignA national government borrowing in its own currency, as distinct from a state government, a city body or a company. government borrowing in its own currency is the first, and it sits apart from the other three for a structural reason rather than a patriotic one: it is the party that issues the currency the promise is written in. Second come other public bodies. Such a body looks similar to a sovereign from a distance and is not the same thing. How far the government behind it has actually committed itself varies, and the commitment has to be read rather than assumed. Third are companies, where the promise rests on the business and on whatever has been pledged behind it. Fourth are banks and other financial firms. Such a firm is a company too, but it sits under a supervisor, and the supervisor's requirements shape what the firm may do while the bond is outstanding.

The four kinds are separated by what stands behind the promise and by what a holder is dealing with if the promise is not kept, and neither of those is legible from how well known the name is. A large borrower everybody has heard of and a small one nobody has are not thereby on different footings. The footing is the structure, not the recognition.

Two warnings sit alongside that structure. The first is that the four kinds are not a ranking: no kind is safer, better or preferable to another by virtue of its position, and a reader who takes the order of the rows as a league table has read something into the drawing that is not in it. The second is that the differences between the kinds are structural descriptions and not guarantees. Saying that a sovereign government borrowing in its own currency issues the currency the promise is written in describes a structural feature of that promise. The description does not say what any particular government will do.

The abstraction is doing a lot of work, so here is the everyday version. Two questions matter: who is asking for the money, and what is left if they stop paying. A cousin borrowing against a salary, a neighbourhood association borrowing against subscriptions it collects, a shopkeeper borrowing against a shop, and a lender down the street borrowing to lend it on again are four different propositions. Every one of them can write an identical slip of paper. The slip is identical and the four propositions are not, and the difference lives entirely in what is behind the slip.

Can the borrower be identified from the coupon?

No, and there is an unusually clean way of showing it. The demonstration was already sitting in the record before anybody went looking for one.

The ten year 8.50 per cent bond that this whole sequence works on has no issuer. Not a hidden one, not one held back for later: the record that supplies the instrument names no party behind it at all. Two reasons stand against filling that gap with a plausible invented name. The first is that an invented name would read exactly like a recorded one, and a reader would carry it forward as though it had been established. The second reason is better. The gap itself is the lesson. Here is an instrument specified completely enough to price, to schedule and to check, and it establishes nothing at all about who is behind it, because a coupon rate, a face amount and a set of dates simply do not carry that information.

Now the opposite experiment. The point is only half made until something changes when a borrower does get named. Palash Cements Limited's five year issue carries a contracted coupon rate of 9.10 per cent a year on Rs 1,000/- of face amount. The rate puts Rs 91.00/- on each of its five dates, against the Rs 85.00/- on each of the ten year bond's. Naming the borrower did not change any of that arithmetic. Naming the borrower did change one thing: there is now something to ask questions about. Before the name, "is 9.10 per cent a lot?" had no possible answer. After it, the question at least has a subject, even where the answer lies elsewhere.

Two bonds carrying the identical contracted coupon rate can therefore sit at opposite ends of any judgement anybody makes about being repaid, and the rate will not separate them by so much as a basis point. The rate was a term written into a document at one moment. Whether the party behind it keeps to the document is a different question, and the document cannot settle that question about itself. Somebody eventually started publishing opinions about borrowers for exactly that reason.

What is a rating, and where does it come from?

A rating is an opinion. A firm that publishes ratings builds a scale, defines what each grade on that scale is supposed to mean, sets out in a methodology documentThe publication in which a firm that grades borrowers sets out how its own scale is built and what it takes into account. The document is issued by that firm rather than by any authority. how it decides which grade a borrower gets, and then publishes an opinion about how likely a particular borrower is to meet its obligations. Every part of that sentence is doing work. The grade rests on a firm's scale, a firm's definitions and a firm's opinion, and it carries a date.

Three boundaries follow, stated in order rather than left to be inferred.

First, no scale is written out here. Building a scale, and deciding what its steps stand for, is work each publisher does for itself and writes up itself. The Securities and Exchange Board of India (SEBI) at sebi.gov.in holds the separate ground of how a published opinion has to be disclosed and kept under review. Both of those get revised, so a copied scale reads correctly on the day it is copied and goes quietly stale afterwards, with nothing on the face of it to say which day a reader is on.

Second, Palash Cements Limited is given no grade, here or anywhere on this platform. The record that supplies it holds none. An invented grade would look precisely like a recorded one, in the same typeface, in the same row, and nothing would separate the two. So the row is drawn and left empty. Drawing an empty row is a different statement from leaving the row out. An empty row says a specific thing: here is a question, and here is where it is answered.

Third, a rating is not a fact about the future. It is an opinion formed at a moment by a firm applying its own published method, and opinions are revised. Treating one as a settled property of a borrower, the way the face amount is a settled property of a bond, mistakes the kind of object it is. How a rating is actually built, and what it does and does not claim, are covered separately.

What kind of rate is each figure in that comparison?

Here is the whole comparison this record supports, worked once and then fenced. The five year government spot rate is entered at 6.90 per cent a year. Palash Cements Limited comes to the market for five years at 9.10 per cent a year, and 9.10 per cent is that issue's own yield at issueThe rate attached to one particular bond at the moment it is first sold, rather than a point read off any curve.. Subtract, keeping the unit attached on the way: 9.10 less 6.90 leaves 2.20 percentage points, and since one percentage point is 100 basis points, the same distance reads 220 basis points. Both rates are assumed rather than observed, both compound once a year, and neither one records a level at which anything was ever bought or sold.

Now label them. Two rates sitting on one horizontal scale are not thereby the same kind of object, and two kinds are in play in a single sentence here. The 6.90 per cent figure is a spot rate, attached to one future date and read off a curve; the 9.10 per cent figure is a yield on one issue at the moment it was sold, attached to that instrument and to nothing else. They can be subtracted because both are rates per year on the same compounding convention. The two rates cannot be substituted for one another, and a reader who lets them merge will stay quietly confused about which curve anything sits on.

A third object is worth naming. A forward rate is constructed out of two rates for two different dates rather than observed anywhere at all, and it answers a different question from either figure above. The forward rate is named here and carries no number. The two parent spot rates are not standing in view, and a forward quoted alone teaches nothing. The whole point of the object is that it was already inside the curve rather than a separate opinion about anything.

The four things missing behind that gap

The subtraction is finished, and the honest part of the worked instance starts now. The gap is 220 basis points and the account stops there, and it is worth being precise about why stopping is a finding rather than a shortfall. Four things are missing, and each of them would be needed to say anything further.

So the distance is 2.20 percentage points, the distance is 220 basis points, and it exists because a lender asked for more from one borrower than from the other. The distance's composition, what it is compensation for, and whether it is enough are all covered separately, and not one of them can be answered from anything above. A spreadThe distance between two rates quoted for the same length of time. Where a spread comes from, and how it splits up, is worked through separately. being wide or narrow is a judgement that needs the four missing rows above, and offered without them it would be invention in a confident voice.

What a single recorded pair of rates will support

A control here would have to move the certainty of being repaid and show the rate travelling with it, and moving anything takes more than one reading of it. The record holds one pair of borrowers and one distance between them, full stop. Every position of such a control other than the single recorded one would be an invented rate, and an invented rate that slides smoothly under a finger is a great deal more persuasive than the same figure sitting still in a sentence.

Where does the 220 basis points actually sit?

The question is worth a part of its own. Separating two things that a table of figures shoves together turns up everywhere in this subject once it has been done once.

Palash Cements Limited's 9.10 per cent a year is a line in a document. Somebody wrote it, both sides agreed to it, and it binds: it is a term of the obligation in exactly the way the face amount and the dates are terms. The 6.90 per cent a year at the five year government point is nothing of the sort. The government figure is a level that a market carries, sitting in nobody's document, binding nobody, and belonging to no single instrument. And the 220 basis points between them is in neither place. No clause anywhere creates it: the gap is arithmetic a reader performs by subtracting one rate from the other.

Do the same trick with two figures that are both clauses and the result is just as strange. The ten year 8.50 per cent bond's contracted coupon rate and Palash Cements Limited's 9.10 per cent are each a line in a document, one in each. Subtract them: 9.10 less 8.50 leaves 0.60 percentage points, or 60 basis points. The 60 basis points is not a term of either obligation. Neither borrower agreed to it, neither document mentions the other, and the two instruments do not even run for the same number of years. The figure came into existence at the moment of subtraction, and it describes that comparison rather than anybody's contract.

Why labour it? Because tables of rates are the ordinary way this material reaches anybody, and a table quietly turns everything in it into the same kind of thing. A column headed "rate" will happily hold a contracted coupon rate, a level read off a curve and a difference between two of them, in three consecutive rows, in the same typeface. Contract terms, market levels and reader arithmetic are three different kinds of object, and a table is the machine that makes them look like one kind. Telling which of the three a figure belongs to is most of the skill.

How does anybody actually use this?

Three readers, three different uses, and none of them is exotic.

Somebody deciding where to put money

The practical habit is small and worth installing early: whenever a rate turns up, ask what the second column would say. Not "is this rate good", a question with no answer without the second column, but "who has agreed to pay it, and what is left if they stop". A household comparing two places to put savings does exactly this without the vocabulary, and does it well. The two places have names attached, and the names carry a lifetime of impressions. The difficulty on a list of bonds is that the names are unfamiliar, so the impressions are missing, and the rate rushes in to fill the gap. Recognising that the rate cannot fill that gap is the whole content of the habit.

Somebody lending on behalf of an institution

A lender that does this professionally splits the question in two and keeps the halves apart on purpose. One half is what is promised: a document to be read and arithmetic to be checked. The other half is who is promising it: a judgement to be formed and then written down separately, where somebody else can disagree with it. Keeping them apart is what lets two people who disagree find the disagreement. If the party were folded into the arithmetic, two analysts with different views would produce two different prices and could not tell each other where the difference came from. Keeping the party in the rate means they agree on every step, put in different rates, and know exactly what they are arguing about.

Somebody comparing across borrowers

An analyst setting one borrower against another leans on the arithmetic being identical, and the identical arithmetic is the whole use of the point made above. Because the sum does not change with the party, the same schedule can be run for every borrower on a list, and the differences that come out are differences in the rate. A difference in the rate is a difference in what lenders wanted from each of them. The comparison only holds while the schedules genuinely match: same length, same shape of repayment, same compounding convention. Compare a five year promise with a ten year one and the gap between the rates is carrying the length of the loan as well as the borrower, and separating those two effects is a subject of its own that is covered separately.

Where the household version stops

One honest limit on all three. A household comparing two familiar names is using information that a list of unfamiliar names does not carry, and the answer is not to manufacture the missing impression from the rate. The answer is to go and get the information, or to accept that it is missing. Not knowing who is behind a promise is a perfectly respectable state to be in, and much better than the state of having inferred it from a number that cannot carry it.

Who sets the rules about who may issue and who may hold?

Six separate questions crowd in around a borrower. Each belongs to an authority. Each authority keeps changing its mind in writing. Doing so is its job. A requirement copied into a summary goes wrong the day somebody revises it, without any warning at all. The type does not change colour. A reader arriving a year afterwards takes the stale version away with precisely the confidence of somebody who read it the week it was typed.

Here they are. Who is permitted to issue a bond at all, and on what terms. When the sovereign issuer borrows across a year, and how much. The disclosure a borrower has to make about itself when it comes to the market, and what it has to keep disclosing afterwards. How a grading scale must be defined, what each grade on it is taken to mean, and how a published opinion must be reviewed. Which categories of investor are allowed to hold which bonds. And how much of its own money a regulated holder has to carry against a holding in a given borrower's bonds, the capital treatmentHow much of its own money a regulated lender or holder must keep against something it holds. Set by its supervisor rather than by the bond. of that position.

Six requirements get named above and none of the six gets stated, and the omission was designed in rather than forgotten about. The half of this subject that lasts is the shape. A bond has a party standing behind it. The party is one of the terms of the promise. The party moves the rate and leaves the sum alone. Four kinds of borrower differ in what holds up the promise. A grade is an opinion somebody published. Reissue a circular tomorrow and not a word of that changes. The contents of those six rows do change, and going to the source for them means a revision costs one link to re-check instead of a lesson to unlearn.

Palash Cements Limited is invented.
Educational material. Not advice on any investment, tax, budget or market position.