Exchange Traded and Over the Counter: Who Is Opposite You

A standardised contract written once by an exchange, taken by anybody who wants it and settled against a clearing corporation that steps in as the side opposite each holder, is one route. Two named businesses sitting down, drafting their own terms and depending on each other for every payment due is the other. The privately agreed route is the one the worked figures below sit on.

Two questions separate the routes, and only one of them earns the attention it gets. The first asks who wrote the terms. The answer decides how much of the document a business is allowed to change. The second asks who is standing opposite when a payment falls due. The answer names who the business is actually relying on. Almost every practical difference between the two routes falls out of the second question. Most treatments lead with the first question, and that is why readers finish them still unable to say who owes them money.

Putting a worth on the agreement part way through takes a forward path for the floating benchmark, dated period by period. No such path is assumed anywhere below, so no worth is put on the agreement at any point, on either of the two routes. The opening period needs no such path, and it is worked in full.

One agreement sits underneath every figure in this guide, and its notional reads Rs 1,000 crore. One row of it carries a rate that was written down at the start and never moves again, 7.20 per cent for each year the agreement runs, and handing that row over is the obligation Chitrakoot Cements Limited took on. The opposite row is tied to the floating benchmark, whose reading is looked up rather than agreed, and Saranga Capital Limited is the side that pays it. Whatever one side sends out, the other side takes in.

What exactly is an exchange traded derivative?

An exchange traded derivative is a contract whose terms were written once, by an exchange, and are then used unaltered by everybody who takes one. The quantity it stands over is the same for every holder. The dates are the same. The thing it references is the same. The way it settles at the end is the same. The document existed before any particular business turned up, so nothing on it was drafted with one in mind.

Somebody taking one of these makes exactly two decisions: which side of the contract to be on, and how many. Everything else was settled in advance by parties they will never meet. Settling everything in advance is not a flaw in the route. The uniformity is the whole design, and the design is what makes the next part possible: because every contract of that description is identical to every other, one of them can be matched against another without anybody reading either.

The trade goes through the exchange. Then something happens that most descriptions skip over in half a clause. A clearing corporation places itself between the two sides. The holder's arrangement is then with that institution rather than with whoever happened to be selling. The word for that substitution is novationThe original pair of obligations is torn up and rewritten so that an institution appears on both sides. The trade survives; the pairing does not., and the mechanics of how the institution manages what it has taken on are worked through on their own further on in these notes.

A ticket window is selling seats on the morning train. Every ticket for that train carries the same route, the same conditions, the same refund arrangement, and none of it was drafted for the person at the counter. The person at the counter decides two things only: which train, and how many seats. If they want a different route or a different set of conditions, there is no negotiation available to them at that window; there is only a different train, or no train. And when the train is cancelled it is the railway they are dealing with, not the person who happened to be in the queue behind them.

What exactly is a privately agreed derivative?

Two named businesses sit down and settle the whole thing between themselves. The two of them choose the size. The two of them choose the dates the payments fall on. The rate that gets written into the fixed row is theirs to choose, and so is the benchmark the floating row is tied to. The counting convention that turns a rate into a number of rupees is theirs. So is what happens if either of them fails to pay. Then they write it all down, and that document governs their arrangement and no other arrangement anywhere.

The two facts that matter most about this route are the two that get assumed rather than stated. Give it the same attention as the route above. Nothing has been bought from anybody, and no third party has taken any part of it on. There is no seller. There is no institution that stepped into the middle after the fact. The arrangement is bilateralExactly two sides, and no third party with any part in it. The word says how many, not how big. from the first day to the last, and the promise a business is holding is the promise of the specific name printed on the front of its own copy.

Somebody has a hall. Somebody else wants it for a wedding. Between them they settle the date, the hours, how much goes down as a deposit, who pays if the generator fails and what happens if the guests run over by two hours. Nothing about that arrangement was printed in advance, and nothing about it can be handed to a stranger. There is no ticket office anywhere in the transaction, and on the day, if the hall is locked, the person to ring is whoever holds the keys and signed the paper.

Where do the terms come from on each route?

On the standardised route the terms arrive already written, and a business can take them or leave them. Taking the terms as written sounds like a limitation, and it buys two real things. There is nothing to argue about, so the first purchase is speed. The second is sameness: a contract identical to everybody else's can be compared, matched and offset without a single line being read.

On the privately agreed route every line is open, and every open line has to be worked out. The negotiation buys fit: the dates can be made to land where a business actually receives money, the counting convention can be made to match its own books, and the failure clauses can be argued to a place both sides can live with. The bill for that fit arrives as time spent, and as a document that resembles nobody else's.

Standardisation buys speed and sameness at the cost of fit, and negotiation buys fit at the cost of speed and sameness. Stated in both directions like that, something becomes obvious that a one sided description hides: neither property is a virtue on its own. Sameness is worthless to a business whose position is peculiar. Fit is worthless to a business that needs to be in and out inside a week. Neither of those properties makes either route right for anybody.

Who is actually standing opposite?

Who is standing opposite is the question the comparison turns on, and that question is worth slowing down for. On the standardised route the institution in the middle takes the place of whoever the business originally dealt with. The original party walks out of the picture entirely. The expectation of payment now runs to the institution, and it runs there for every day the position stays open. The business never chose that institution in any meaningful sense; it came attached to the venue, the way the railway comes attached to the ticket.

On the privately agreed route the side opposite is the name written into the document, and it stays that name until the last payment date, with nobody else in the chain at any point. If that name is late, there is no institution to go to instead. If that name is gone, the arrangement is a claim against whatever is left of it.

On one route the business chose the party it is relying on, and on the other route it did not. That sentence is the one the whole comparison exists to deliver, and it is not a complaint about either route. A business that has looked hard at a name, checked it, and decided it is content to depend on that name has done something an institution in the middle cannot do for it. A business that wants no view on any particular name at all is better served by not having to form one. Both of those are sensible positions.

What does that change about what is being carried?

A business in one of these arrangements is carrying two things at once, and a single word like risk covers both of them so smoothly that most people never separate them. The first is how the referenced thing moves: whether the floating benchmark comes in above or below the fixed rate, and by how much. The second is whether the side opposite is actually there on the day a payment falls due.

Take them one at a time. On the privately agreed route both are live. The floating benchmark does whatever it does, and separately, one specific named business has to be in a position to pay when the calendar says so. The movement and the payment are two different questions with two different answers, and nothing about the first says anything about the second.

On the standardised route the second question is not the same question. The institution in the middle is built to be the side that is always there, and how it is built to manage that is handled on its own further on. The point to carry forward is only that the question changes shape.

The first of the two does not change at all. This is the part that gets forgotten, and it is worth stating flatly: the referenced thing moves exactly the way it moves, whoever is standing opposite. Move an arrangement from one route to the other and the movement in the benchmark is untouched. A business that reads the change of route as a reduction in what it is carrying has quietly credited itself with something it did not get.

How does a party get out of one, and why is that different?

Here is where the two routes stop feeling similar, and it is the part almost nothing covers properly. On the standardised route a business leaves by taking the opposite side in the same contract. The business holds one contract and then the other, the two cancel, and the position is gone. Cancelling works only because every contract of that kind is identical to every other. The word for that property is fungibleSwap one for another of the same kind and nothing at all changes for whoever holds it. Interchangeability is what the word points at., and it is doing quiet, enormous work: nobody has to find a specific person willing to release a specific business from a specific promise, because there is no specific promise to be released from.

On the privately agreed route nothing cancels by itself, ever. The first document is not identical to anything, so a second document cannot annihilate it. Three routes out exist, and each one has a cost the standardised route does not have. The business can negotiate an ending with the same side, on whatever terms that side is willing to accept. Finding somebody prepared to step into its place is an assignmentSomebody else takes over one side of an arrangement, which the party opposite has to agree to before it can happen., and that needs the agreement of the side opposite before it can happen at all. Or it can enter a second arrangement pointing the other way, leave the first one standing, and carry both to the end.

On one route leaving is a transaction, and on the other route leaving is a negotiation. The sting is in the second half: a negotiation takes exactly as long as the other side wants it to take, and the other side knows perfectly well which party is the one who wants out. The third route out carries a further meaning. Two arrangements running at once is two sets of payment dates, two sets of paperwork, and two names to depend on rather than one.

Is the headline figure money on either route?

Both documents open with a big number and neither of those big numbers is a sum anybody is holding. A standardised contract states a lot sizeA single standardised contract stands over a fixed quantity, and that quantity is what this phrase names. and a number of contracts. A privately agreed arrangement states a notional. The two documents give the figure two different names, and the different names hide the same fact: in both cases the figure is there to scale a computation, and in neither case does it move.

Worked on the agreement above, it runs like this. The fixed rate is 7.20 per cent, applied across a full year to the notional, so the fixed row lands on Rs 72.00 crore. The benchmark showed 6.00 per cent for the opening period, and across the same notional the floating row arrives at Rs 60.00 crore. One row set against the other leaves Rs 12.00 crore, and it travels from the fixed side to the floating side.

Nothing is being compared across dates or across currencies here. One date, one currency, two rows, so a subtraction is allowed.

There is a second way through the same arithmetic that removes both rows entirely. 1.20 percentage points is what separates the two rates. Multiply the notional by that gap and Rs 12.00 crore drops out directly, with neither row computed at all. Read it in a smaller unit if that helps a reader feel the scale: the difference works out at twelve paise on every hundred rupees of notional, for the year.

Whatever the headline figure is called on a document, find the schedule of movements and see whether that figure appears in it. On this agreement Rs 1,000 crore does not appear in the schedule at all and Rs 12.00 crore does. Checking the schedule is the reading rule, and it never depended on the name, so it survives the change of name. A standardised document would carry a different phrase at the top and the same test would apply to it unaltered.

What does the same agreement look like on the other route?

Now put one arrangement through both routes and watch what survives the crossing. As it stands, the agreement is privately agreed. The notional of Rs 1,000 crore was picked by the two sides because that is the size the position needed to be. The fixed rate of 7.20 per cent for each year was argued to that level and written down. The floating benchmark showed 6.00 per cent for the opening period, the counting fraction across one full year is 1.0000, and Rs 12.00 crore of difference settles from one side to the other. Chitrakoot Cements knows precisely who owes it money in every later period, by name, and the name will be the same name every time.

Now imagine the same shape as a standardised contract. The size would not have been picked; it would be some multiple of whatever unit the contract came in. The dates would not have been picked. There would be no argument available, so the rate would not have been argued to any level. And the party owing the money would be an institution rather than a name anybody selected.

Every field on the standardised version of this arrangement is left blank on purpose, and the reason is worth stating plainly: specifications are set by an exchange and by an authority, and they are revised deliberately and often. A figure copied out today goes on being trusted long after the original has changed. The five rows below therefore compare the two versions on the questions that need no specification at all.

Read across the rows rather than down the columns, and ask which of the five gives the same answer on both sides. The answer is none of them. Not one row is shared. Five differing rows are the reason the two are taught as two separate things rather than as one thing with a settings menu attached, and the reason a reader who has learned one of them has not thereby learned the other.

Which route costs more, and what would it take to say?

No answer to that is available, and the reason is worth setting out rather than asserting. It is a fair question, and readers arrive with it. People want to know whether the same economic arrangement is dearer one way than the other. Wondering about the cost is exactly the right instinct.

Neither route can be priced without an input that no agreement carries on its face. Whether an arrangement was struck across a table or taken through an exchange, saying what it is worth part way through takes a schedule of expected benchmark readings, and no such schedule is assumed anywhere here. The absence applies symmetrically. It is not that the standardised route is unpriceable and the other one is priced. Neither can be.

Four things would have to be in hand before anybody could answer the cost question, and none of the four is here. Both routes quoted on the same date. The same referenced thing, the same size and the same maturity on both quotes. A provider willing to state where it obtained each of those quotes. And behind each quote, a schedule of what the benchmark is expected to do.

The comparison that can be made is about how each route works. Who wrote the terms, who is opposite, what changes about the load, and what leaving involves. All four are structural facts that hold whatever the pricing turns out to be.

Who sets the conditions on each route?

Three separate conditions bear on the two routes. Every one of them is settled elsewhere, so what a reader can carry away is the routing rather than the answer: each label comes with an address, and each value stays where it lives.

The contract specification is settled elsewhere, in every one of its parts. No quantity that one contract stands over, no expiry, no position limitA ceiling exists on how much of one contract a single participant may hold at once. The phrase names the ceiling., no collateral figure and no percentage. Where the details of a privately agreed arrangement are filed, and in what form, belongs to the same category: the destination is often a trade repositoryDetails of privately settled arrangements get filed in one place so somebody can see the whole picture. The phrase names that filing place., and what goes into it is decided at the address in the row above rather than here.

Where does this go wrong in practice?

A rent agreement gets stamped and registered at the sub-registrar's office. The registration is real and worth having. The register records that the agreement exists, on those terms, from that date. The register does not hand back the deposit when the tenancy ends. The sub-registrar has no part in that. The deposit comes back from the person who took it, and if that person has spent it, registration changes nothing at all about the position. A record that an obligation exists is not a promise that the obligation will be met, and the gap between those two things is where this whole failure lives.

The other half of the damage is the internal record, and it is the half that survives longest. Two arrangements of genuinely different shapes get filed under one heading. Six months later somebody totals that heading and reports a single number upward. The single number folds together positions where an institution stands in the middle and positions where a single name stands opposite, and no reader of it can separate them. Nothing in the arithmetic is wrong. The heading was never built to show it, so the concentrationLeaning heavily on one name rather than spreading the leaning across several is what this word describes. sitting inside the total is simply invisible.

Who actually reads a document for this?

One signed arrangement crosses four desks, and the same sheets get used differently at each. At the first desk it is read for a single line, the name of the side opposite, and that name goes onto a list of names the business is depending on. Nothing else on the document is looked at. The first desk is building a picture of how much of the business's dependence sits on any one name. The arrangement itself never asks that question.

At the second desk only the schedule matters. Somebody copies out the dates and the direction. The day Rs 12.00 crore has to leave is then a day the business already knows about rather than a day it discovers. The second desk is where a mismatch between a payment date and a receipt date turns from a nuisance into an overdraft.

At the third desk the document is read for what would have to happen before it could be ended early, and the answer gets copied into a note about how quickly the position could be unwound if the business needed it gone. On a privately agreed arrangement the honest answer names a negotiation rather than a trade, so the note runs longer than people expect. The standardised answer is the one everybody has heard, so anybody who writes that note before reading the document tends to write it.

At the fourth desk nothing in the document is read at all. The question there is whether the same shape could have been taken as a standardised contract instead. A document describes the route that was taken and says nothing about the route that was not, so the document cannot answer it. The answer comes from going to an exchange and looking at what it lists. A household version of the same desk exists: somebody who has taken a loan from a relative on agreed terms cannot learn from that agreement what a bank would have charged, and has to go and ask a bank.

So which route is the right one?

Which of the two routes is right for a particular reader is decided out of material that lives in the reader's own records and nowhere else: what is already held or already owed, how closely an arrangement has to fit against that, the stretch of time it is expected to run for, the plan for ending it early if ending it early becomes necessary, and the identity of the parties whose promises are acceptable. Not one of those five things is written down here. Enough is set out above to let somebody handed either document say who is standing opposite them.

The two routes do different jobs. A business that needs a position on and off inside a fortnight and a business that needs an arrangement matching an eleven year obligation are not choosing between a better option and a worse one; they are choosing between two tools that were built for different work. Anybody who calls one of them better without asking about the position it has to serve has answered a question nobody asked.