How to assess Fixed-Income Liquidity Context

Assessing liquidity context means turning a loose sentence about selling into a claim with its conditions written beside it, then asking which of those conditions anybody could observe and where the observation would sit. Six questions do that. Where no observation exists the step stops and the gap is written down, and whatever survives is labelled as an assumption.

Those six questions are a whole method, and the method can be run on a train with nothing in hand. The method needs no screen, no data feed and no figure of any kind. A pen is enough, together with the willingness to write a longer sentence than the one that first suggests itself.

Why can this be done with no data at all?

A sentence about selling sounds like an observation. Such a sentence almost never is one. Somebody writes that the bonds are liquid, or that they trade regularly, or that a position can be got out of in size, and every one of those sentences has the grammar of a report from the world. Underneath, in the great majority of cases, sits a supposition: a view formed over years of watching, held honestly, and never once checked by anybody else.

The words are identical, so the difference between the two is not in the words. It is in whether anything behind the sentence could be checked by a second person. The question of whether a check is possible is answerable without data. The observation itself is not needed in order to ask whether an observation exists, and a quantity need not be known in order to know that a quantity is the sort of thing a venue either publishes or does not.

A household selling a scooter knows this instinctively. The sentence it will sell is a thing anybody will say. The sentence it will sell this week, at something close to what the last one in the lane went for, is a completely different one, and only the second can be shown to be wrong by Sunday. Nobody needs a market study to tell those two apart. The household needs only to notice that the first sentence has no week in it and no price in it, and so nothing in it can fail.

The whole method below is that noticing, made systematic and applied to a debt instrument instead of a scooter. Six questions, asked in a fixed order, each one handing its output to the next.

What are the six questions, and in what order do they run?

The sequence follows, whole, before any of it is opened up. Taken in once as a list, its shape carries through the rest of this guide.

1. What is actually being claimed?The one sentence separates into the several claims it contains, and the claim actually meant is the one written down.
   Hands on: a written claim naming a size, a period and a price.
2. What would have to be true for it?List the conditions underneath that claim, in plain words, before looking for any evidence at all.
   Hands on: the claim, with its conditions travelling beside it.
3. Which conditions could be observed, and where?Sort the conditions into the ones with an observable counterpart and the ones without, then name where each observable one would be found.
   Hands on: a route for each observable condition, and a category refusal for the rest.
4. What happens where the observation is missing?The step ends. Write the gap in the place the figure would have gone, with one line saying what would fill it.
   Hands on: a gap against every condition the platform cannot reach.
5. What can still be said?Collect what survives the gaps, which is more than it looks: the terms, the amount any cost of selling would come out of, and the absence itself.
   Hands on: the material for a paragraph that is entirely defensible.
6. How is what is left written down?Two sentences, carrying the assumption, the reason for choosing it, and what changes if it is wrong.
   Hands on: the only output this sequence produces.

Notice what is not in that list. There is no step that scores anything, no step that ranks anything against anything else, and no step that ends in a verdict about whether an instrument can be sold. The output of all six questions together is a paragraph, not a number. That can look like a thin return for six steps of work; the last block of this guide is about exactly why the thicker-looking alternatives are worse.

Step one: what exactly is being claimed?

Take the sentence the bonds can be sold and lay it out flat. The sentence is at least four claims wearing one set of words.

The first is that somebody would take the other side at all. The second is that they would take it at the size actually held. The third is that they would do it inside a stated period. The fourth is that they would do it near the last price anybody paid. The four claims have different evidence sitting behind them, they fail in different ways, and they cost the holder different amounts when they fail. Writing the sentence without saying which one was meant leaves the reader to pick, and readers pick the strongest one.

Step one produces one written claim, and its job is to name the size, the period and the price out loud. Something like this: the whole holding could be sold inside one week, at a price near the last one paid. Longer than the sentence that first suggests itself. Also testable, and the shorter version was not.

The scooter comes back here. The sentence it will sell contains the same four claims and hides them the same way. At the size held becomes the whole scooter rather than the mirrors and the seat. Within a stated time becomes this week. Near the last price paid becomes near what the neighbour got. The household that says all three out loud has not become more pessimistic; it has become checkable by Sunday.

Step two: what would have to be true for that claim to hold?

The claim written in step one gets its conditions listed underneath it, in plain words, before the search for evidence begins. Evidence found first bends the conditions to fit it, so the order matters.

For selling at the size held, the condition is that somebody on the other side wants that quantity, or that several somebodies add up to it. For selling inside a stated period, the condition is that those somebodies are present within the period rather than eventually. For selling near the last price, the condition is that there are enough of them that the first one does not get to set the price alone.

Three conditions from four claims, and the arithmetic there is worth a sentence. The bare existence claim, the first of the four, is contained inside the size condition, so it never gets a condition of its own: somebody wanting the quantity is somebody, and a condition that only asks for a body in the room is satisfied by any body at all, including one who wants a fraction of the holding. Four claims collapse into three conditions because the weakest claim is a special case of the second, and folding it in is the only compression this sequence permits anywhere.

From here the claim and its conditions travel together. Not as a courtesy: a reader handed the claim with the conditions stripped off has not been handed a claim, they have been handed a conclusion. There is nothing in it for them to weigh, so there is nothing for them to disagree with, and a reader who cannot disagree is not reading, they are receiving.

Step three: which of those conditions could be observed, and where?

Every condition from step two either has an observable counterpart or it does not, and deciding which is a completely separate job from having the observation to hand. Step three does the sorting. Step three does not do the finding.

Some things are the sort of thing that can be observed. A quantity dealt on a named day happened, and if anybody wrote it down it can be looked up. A two-way priceA price shown in both directions at once, one level to buy at and another to sell at, with the gap between them part of what is being shown. quoted at a stated time happened too, and so did the gap inside it. Days with any dealing on them can be counted, and counting is the plainest kind of observation there is. So is turnoverThe total amount of something dealt across a stretch of time, arrived at by counting rather than by estimating. across a month.

Other things are not the sort of thing that can be observed at all. A statement about what somebody would do tomorrow is not an observation in any form, however experienced the person making it and however often they have been right before. Such a statement is a supposition. The distinction is about the category, not about the quality of the person. Right here is where an experienced writer is most tempted to promote their own judgement into the evidence column, and that is why the distinction is drawn at this step.

Step three sorts the conditions, and it also has to say where an observable one would be found. Naming the place is where the sorting stops. What a trading venueThe place where dealing happens and gets recorded, as against the place an instrument was first sold to its original holders. has to put out about dealing in a debt instrument, and in what form, is set by the Securities and Exchange Board of India (SEBI) for corporate debt and by the Reserve Bank of India for government securities. Both of them revise it. So the table further down carries a list of headings with nothing typed under any of them.

One more distinction decides whether a quoted gap is an observation, so it belongs here. A price that somebody is committed to dealing at is a different object from an indicative priceA level put on a screen for information only, with nobody committed to dealing at it when somebody turns up.. Both appear on screens. Only one of them is evidence that a condition was met, and a note silent about which kind it read has not really said anything at all.

Step four: what happens where the observation is missing?

The step ends. Three words are the whole instruction, and the shortness of them is the point.

The step does not continue with a substitute. The step does not continue with a proxyA figure taken from something similar and dropped into the slot where the real one is missing, usually without a line saying that it was borrowed. from a comparable instrument. The step does not continue with a figure recalled from a period when somebody was watching this market closely. And the step does not continue with a range. A range is the most seductive of the four, because it sounds like caution while doing exactly what a point estimate does: putting a number where there is no observation.

Instead the gap itself goes down where a figure would otherwise have stood, and a line beside it names the observation that is missing. Not a footnote and not an apology. The gap goes in the cell, in the sentence, at the point in the paragraph where a reader is expecting to be told something. A gap named that precisely is information a later reader can use, and a gap papered over is a defect nobody can ever find.

The last sentence inverts the instinct, and it is the one worth sitting with. Filling the gap feels like doing the job and leaving it open feels like failing to. The matter is settled by what a later reader can do with each. Handed a borrowed figure, they cannot tell it apart from an observed one, cannot trace it, and will build on it. Handed a written gap, a reader learns the exact boundary of the note, and can go and find the missing observation if they need it. The second reader is better off, and so is the writer, whose reputation now rests on nothing they cannot support.

Exactly one place in this whole territory answers a missing observation with a rule instead of a gap, and flagging it keeps it from being taken for a general licence: where a holding has to be written into a set of books and nothing has changed hands, a carrying priceThe value a holding is written into a set of books at, which somebody has to settle even in a stretch where nothing has changed hands. still has to be settled somehow. The carrying price is a valuation question with a keeper and a live norm, named in the table below and covered separately. The norm answers what goes in the books. The norm does not answer whether the instrument could have been sold.

Step five: what can still be said once the gaps are named?

More than it looks, and this is the step people skip. Having written three gaps, a writer tends to conclude that the liquidity paragraph is dead and to drop the subject. A paragraph that says nothing gets rewritten by somebody who wants it to say something, so the overcorrection is precisely how a gap ends up quietly filled two drafts later.

Three things survive. The first is the terms of the instrument. The terms say what the instrument pays, when it matures, how much of it was issued, and whether it is one of many identical claims or a single negotiated arrangement. All of that is knowable, and none of it is a liquidity claim. The terms matter to the question without answering it.

The second is the absence itself, stated as a finding rather than as an embarrassment. A note can say outright that it holds no quantity dealt, no quoted gap between a buyer and a seller, and no count of days with any dealing at all, and the sentence is both true and useful. The sentence tells a reader the shape of what is unknown, a different and better thing than telling them nothing.

The third is the consequence of the assumption. The consequence can be worked even where the assumption cannot be tested. A cost of selling that has never been observed can still be put through arithmetic at a declared size, to show how much of a return it would eat. Being unable to measure a cost does not prevent measuring what that cost would do. That arithmetic is covered under setting a liquidity assumption against a liquidity fact, because it needs a reader who has already been taught to keep the two apart.

Here the invented issuer earns its place, and the only figures in this guide belong to this paragraph.

The 220 basis points is the amount any cost of selling would have to come out of. Notice what that sentence does and does not do. The subtraction never names the cost. The subtraction never says the 220 basis points is enough to cover the cost, or that it is not. The subtraction names the pocket and leaves the pocket empty, the honest position when nothing about the cost has been observed.

Step six: how is what is left written down?

Three parts, and it fits in two sentences.

The first part is the assumption, with the word assumed inside it rather than implied by tone. The second is the reason for choosing that assumption rather than a different one. Often the reason is nothing better than that the assumption is the conservativeChosen so that being wrong costs less rather than more. Being conservative is a reason for picking an assumption and it is not evidence that the assumption is right. one, and if that is the reason, say so plainly. The third is what changes if it turns out to be wrong, stated as a direction at the very least and as a figure wherever the arithmetic allows one.

Whatever would change the writer's mind belongs beside the assumption itself, inside the same pair of sentences, never in a footnote. A reader who disagrees with the assumption then knows exactly which sentence of the note to stop believing, and exactly which observation would settle it. One kind of note invites an argument. The other kind can only be swallowed or ignored, and that gap is the whole of it.

What does the sequence return when it is run on the invented issuer?

Palash Cements Limited, an invented borrower, sits beside a five year government node that is equally made up, and both rates compound once a year. Nothing about how easily anything of theirs could be sold exists anywhere on this platform, and the starvation is exactly why the sequence is run on them: a method can be watched behaving when it is starved.

Step one returns a written claim, and here it is: the whole holding could be sold inside one week, at a price near the last one paid. Step two returns its three conditions: buyers wanting that quantity, present inside the week, and numerous enough that the first of them does not set the price alone. Step three returns three routes and one refusal: a quantity dealt, a quoted gap between a buyer and a seller, and a count of days with any dealing at all are each observable somewhere and each routed to the keeper who decides what gets published; what a buyer would do next week is observable nowhere and is refused as a category.

Step four is where the run stops handing back method and starts handing back holes. This platform holds no dealt quantity, no quoted gap and no count of days, so step four returns three gaps, one against each condition. Across the whole run, not one of the six steps hands back an observed figure about dealing, and the single step that ends in gaps ends in three of them. That is worth stating precisely, because the loose version of it, that most of the steps end in a gap, is not true and would make the sequence look more broken than it is. Five of the six steps hand back writing. One hands back the shape of what is missing.

Step five returns what survives. The terms of the instrument, all of them known. The 220 basis points of spread, the pocket any cost of selling would have to come out of. And the absence itself, a finding in its own right.

Step six returns the written form, the only output the whole sequence produces. Two sentences, five parts, every one of them pointable at.

Read the two sentences whole and notice how little they claim. The sale is assumed possible inside one week at a price near the last one paid; this platform holds nothing that tests the assumption, it was chosen because it is the conservative one, and if it is wrong the cost falls against the 220 basis points of spread rather than anywhere else. Nobody could call that a strong paragraph. Everybody can check every part of it, a different and rarer property.

What can the six questions never produce, and why?

Asking the six questions honestly leaves a want to summarise the result. The want to summarise is worth resisting. The summaries that suggest themselves are all the same summary wearing different clothes, and every one of them undoes the work.

The sequence produces no liquidity score. No ranking of one instrument against another. No rating and no colour code. Each of those compresses several unobserved conditions into a single symbol, and a symbol cannot carry its own conditions. The compression is not a simplification of the answer, it is the exact place the honesty leaves. A reader handed a symbol cannot tell whether it is reporting on the size condition, the time condition or the price condition, so when the assumption fails they cannot tell which part of it failed, and they cannot tell whether the failure was foreseeable.

The sequence also produces no statement that any instrument can or cannot be sold. Not a hedged one, not a qualified one, not one buried in a subordinate clause. Nothing on this platform could support such a statement, and making one anyway would demonstrate the very failure the six steps teach a reader to catch.

The sequence produces a claim with its conditions attached and its gaps named. A claim like that is less satisfying than a score out of five, and it is the only output the evidence allows. The satisfaction the score offers is the satisfaction of not having to think about which condition is being relied on, and the price of that comfort is paid later by whoever is holding the instrument.

Who actually runs this, and what do they do with the output?

A lender deciding how much to advance against a holding of debt instruments runs a version of this without calling it anything. The question they are really asking is what the holding would fetch if they ever had to unwindTo get out of a holding by dealing in the opposite direction to the one that put it on. it in a hurry, and the honest answer almost always contains an assumption. Both lenders make the assumption, so the assumption is not what separates a careful lender from a careless one. The separation is whether the assumption is written into the credit file with a period and a size attached. A year later somebody can then check whether the world moved against it.

An analyst writing a note on a corporate borrower uses the sequence to decide what the liquidity paragraph may say. Very often the honest paragraph is three sentences long and reads as unsatisfying next to a competitor note asserting that the bonds are widely held and actively dealt. The thin paragraph is a real professional cost, and it is worth naming rather than pretending away. The compensation arrives later, when the assertive note has nothing to point at and the careful one has a stated condition somebody can go and check.

A household meets the same structure without any of the vocabulary. Anybody who has tried to get money out of a small savings arrangement in a hurry has discovered that being told it can be withdrawn any time and finding out what any time means in practice are two different experiences. The gap between those two experiences is exactly the gap between a liquidity claim and its conditions, and it is the reason this sequence exists at all. The household that asked, before committing, how many days and at what cost, was running step one.

And an investor reading research written by somebody else can run the sequence backwards as a reading tool. The liquidity sentence in front of the reader is taken, asked which of the four claims it is making, asked what would have to be true, and asked whether the note says where any of that was observed. Three questions, thirty seconds, and it separates the notes worth trusting on this subject from the notes that are simply confident.

Who decides the parts left blank here?

Nothing in the six questions above is set by anybody. Deciding which of four claims was meant, writing the conditions underneath it, and putting the word assumed into a sentence are writing decisions, and no keeper has an opinion about them. The keepers decide something narrower, sitting underneath step three and step four: whether an observation exists to be found at all.

Each row below is keyed to a step of this guide, and each row is left empty. The second column names what would have to stand behind that step once real money moved. The third names who keeps the live wording. Every one of these moves, so the wording is read where the third column points, on the day it is needed. A row typed out from memory is not a stale sentence, it is a false one.

Why is there nothing to drag here?

Six questions have answers, not settings, and an answer is not the sort of thing a slider can move. The subject rules a control out, not the trouble of building one. The one thing worth putting under a slider here would be a cost of selling: a size, a cost, and how much of a return it eats. Such a control exists, correctly labelled, under setting a liquidity assumption against a liquidity fact, where a reader has already been taught to keep those two in different sentences.

Handed to a reader who has not been taught that yet, the control teaches the arithmetic before the caution, and the arithmetic is the easy half. Six questions asked in order are complete as questions; there is no continuous relationship in them for a slider to trace. A ninth question stands below in the place a control would have gone, and it asks for three sentences to be sorted rather than for a number to be moved.

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