Unique Circumstances: Constraints That Do Not Travel
A unique circumstance is a constraint that comes from this holder and no other: a commitment already made, a concentration held elsewhere, an exclusion the holder will not trade away. A unique circumstance is written into the same document as every other constraint and tested the same way. Its distinguishing property is that copying it to another holder produces a limit nobody chose.
Legal and regulatory limits, liquidity requirements and tax constraints all have a clean source, and each is covered separately. A fourth kind is the residue, every line in the document that none of those three explanations reaches. The residue has a shape, it has a test, and it fails in a way the other three do not.
The running example is the Anantara Multi-Asset Portfolio, an invented discretionary mandate of Rs 500 crore run by Faiz Ahmad Ansari for an invented charitable endowment whose committee Rukmini Deshpande chairs. Its stated shape is equity 60.0 per cent at Rs 300 crore, fixed income 30.0 per cent at Rs 150 crore and cash 10.0 per cent at Rs 50 crore, summing to Rs 500 crore exactly. The three shares are policy weightsThe share the holder has decided each asset class should carry. A decision taken in advance, not an observation: actual weights move away from it as prices move.. Every figure that follows belongs to that mandate and to one stated twelve month period.
The record locks the mandate's five stated lines and the year's numbers, and it carries almost nothing about the holder behind them. Wherever a fact about the endowment would settle a question, the record is silent, and the blank is marked NOT SUPPLIED rather than guessed at. Reading a portfolio record from outside puts anybody in that position.
What actually makes a constraint a unique circumstance?
Most documents sort constraints by subject matter. The same subject turns up in every kind, so sorting by subject fails. A limit on unlisted holdings can be imposed by an outside rule for one holder, chosen as a liquidity precaution by a second, and written for a third because of something that holder promised somebody in 2019. Three identical sentences, three kinds of constraint.
Use a different test. Copy the line word for word into a different holder's document, and ask whether a reason for it still exists over there. A constraint is a unique circumstanceA constraint that exists because of something true about this particular holder, so that the reason for it does not come along when the words are copied to anybody else. when the words survive the move and the reason does not. The property has a name: transferabilityWhether the reason for a rule travels with the words when the words are copied somewhere else. A rule with high transferability makes sense wherever it lands; a rule with none makes sense in exactly one place.. A unique circumstance is the line with none.
The test does not ask whether the line is sensible, whether it is common, or whether the manager likes it. The difference lives entirely in the reason, and the reason lives entirely outside the portfolio.
An everyday version. Two households each write down that they will not put savings into anything connected with a particular town. For the first that is a general caution about a place they read about. For the second, their late father's shop is still there, half of it held by a cousin, so further exposure there doubles a risk they already carry and cannot sell. Copying the second sentence into the first notebook imports a rule with no reason behind it.
A mandate excludes unlisted holdings. Is that line automatically a unique circumstance?
Where do these constraints actually come from?
Four places: a commitment the holder has already made to somebody else, a concentration the holder already carries outside the mandate, an exclusionA written line saying the portfolio will not hold a stated thing at all, rather than limiting how much of it may be held. An exclusion is an absolute line, not a size limit. the holder will not trade away at any return, and a restriction that arrived attached to the money.
All four are facts about the holder's whole position rather than about the portfolio. A manager reading only the portfolio can see none of them, and no amount of careful reading of the holdings will make them appear.
Applied to the mandate at hand, all four come out blank. The record locks the portfolio in detail and locks almost nothing about the endowment, and drawing the blanks is more useful than filling them with something plausible.
The second row costs money most often. A holder carries a very large position in one thing, outside the mandate entirely. The manager builds a sensible portfolio inside the window, respecting every stated limit, and the two together are badly concentrated. Nobody has broken anything. The concentration exists only in the sum, and only the holder can see the sum. If the holder does not say so, the constraint that should have existed never gets written, and no compliance report will notice its absence.
The holder carries a large concentration in one thing, entirely outside this mandate. Reading the portfolio record carefully, can the manager see it?
How is this different from a legal, a liquidity or a tax line?
Each of the other three constraint types carries one clean answer. A legal or regulatory constraint is imposed, so nobody in the room can waive it. A liquidity requirement is a schedule of dates and amounts. A tax constraint follows the holder's position, changing what a realised gain is worth without changing what may be held. Tax arithmetic is covered separately.
A unique circumstance is different on two counts at once. A unique circumstance is chosen, so whoever holds the pen can waive it. And it is specific, so nobody outside this holder can reconstruct why it is there, unlike a liquidity schedule that explains itself the moment its dates are read. Chosen and specific together make it the line most often deleted by a new committee. A new committee can waive it and cannot see the reason for it, and those two facts arrive in the same meeting.
How is one written so that it can be checked?
Harder here than anywhere else in the document. A unique circumstance usually starts life as a sentiment. Somebody says the endowment should not be involved with a certain kind of thing, a drafter writes that down almost word for word, and what lands in the document is a sentence that reads like a constraint and cannot be tested.
The same test applies here as to every other line in the document: an outsider holding only the portfolio record must be able to say today whether the line is satisfied, and a unique circumstance that fails that test is an instruction to the manager rather than a constraint on the portfolio. An instruction is not worthless, but calling it a constraint means nobody will ever check it.
The other standing habit of this sequence: read every constraint twice, once for what it stops and once for what it still permits. A line that stops one thing absolutely may still permit most of the exposure the holder thought they were removing, and a committee that has only done the first reading believes it has more protection than it has.
A line in a policy statement reads that the portfolio will respect the holder's stated values. Can that line govern the portfolio?
Which of this mandate's lines can even be tested?
The mandate carries five stated things, and the first job is to notice that one is not a constraint at all. Equity between 50 and 70 per cent. No single holding above 5 per cent of the portfolio. No unlisted holdings. A minimum credit standing on the fixed income sleeve, stated as a policy rather than a rating symbol. And the equity policy weight of 60.0 per cent, a chosen point inside the range rather than a limit.
A portfolio sitting away from its policy weight has not breached a line, and a scorecard that includes the policy weight will report breaches that never happened. So the policy weight has to be excluded from the count before anything is scored. Excluding it leaves four lines to test, and the tests are not all of the same kind.
Hold on to the middle column. The equity range is written in per cent, and its rupee edges of Rs 250 crore and Rs 350 crore hold only at the stated total; the cap is the same, 5 per cent being Rs 25 crore here and something else anywhere else. Every line written as a percentage carries a hidden rupee figure that depends entirely on the base, and every line written in rupees carries a hidden percentage that does the same.
Did the equity range ever bind over the stated year?
The answer needs the worst thing that happened to the portfolio inside the stated twelve months. The record carries a drawdownThe fall from a portfolio's highest point to its lowest point inside a stated window, measured peak to trough. A different window gives a different figure. The window is always quoted with the figure. of 9.7 per cent, against 8.1 per cent for the composite benchmark over the same window. On Rs 500 crore that fall is Rs 48.5 crore.
Now the deliberately unfair test. Charge every rupee of that fall to the equity sleeve, as though fixed income and cash had not moved. The record does not say what actually happened, but this arrangement pushes the equity weight down hardest, so if the range survives it, the range survived the year. Equity goes from Rs 300 crore to Rs 251.5 crore and the total from Rs 500 crore to Rs 451.5 crore, and Rs 251.5 crore over Rs 451.5 crore is 55.70 per cent.
The denominator falls too, so a 9.7 per cent fall does not move a 60 per cent weight by anything like 9.7 per cent. Both the sleeve and the total lose the same Rs 48.5 crore, and the weight slides only 4.30 percentage points. The base moving does most of the work here.
How far was the floor? If equity and the total both fall by the same amount, the weight hits 50 per cent when Rs 300 crore less that amount equals half of Rs 500 crore less that amount. Half of Rs 500 crore less that amount is Rs 250 crore less half the amount. So half the amount is Rs 50 crore and the amount is Rs 100 crore. The floor sat Rs 100 crore of equity fall away, and the year's worst move of Rs 48.5 crore covered 48.5 per cent of that distance, so the equity range never bound.
Did the single holding cap ever bind?
No single holding may sit above 5 per cent of the portfolio. At the stated total, 5 per cent is Rs 25 crore. The largest holding in the Rs 300 crore equity sleeve is Rs 23 crore: 4.6 per cent of the portfolio, 7.7 per cent of the sleeve. Both are correct, and the cap is written against the portfolio, so 4.6 against 5 per cent governs.
How far from binding? The naive answer divides Rs 25 crore by Rs 23 crore and reports a required rise of 8.696 per cent. The naive answer is wrong, for the same reason the drawdown arithmetic is surprising: when one holding rises and nothing else moves, the portfolio total rises too, so the rupee value of a 5 per cent cap rises with it and the holding chases a target moving away from it.
Set it out properly. If the holding rises by some proportion, its value is Rs 23 crore times one plus that proportion, and the portfolio becomes Rs 500 crore plus Rs 23 crore times it. Setting the first to five per cent of the second gives Rs 23 crore plus 23 times the proportion equals Rs 25 crore plus 1.15 times it, so 21.85 times the proportion is 2 and the proportion is 9.153 per cent. Check it: Rs 25.105 crore inside a portfolio of Rs 502.105 crore is exactly 5 per cent.
Nothing like a rise of 9.2 per cent in one holding, on its own, with everything else still, happened inside the stated year. The single holding cap did not bind either. Two lines scored, two lines that shaped nothing.
The largest holding is Rs 23 crore and the cap is 5 per cent of the portfolio. How far must that holding rise, by itself, to reach the cap?
What did the no unlisted line actually remove?
This line looks most like a unique circumstance and, on this record, cannot be shown to be one. The register carries 28 listed names in the equity sleeve and no unlisted holding at all, so the line removed nothing that was held. Its observed costThe part of a constraint's cost that can actually be counted from the record: how many times it bound over a stated period and how many rupees it moved when it did. over the stated twelve months is zero holdings and Rs 0/-.
Was it written because of something about this holder, or as a plain preference for daily prices? The record does not say why any of the mandate's lines exists, so whether the no unlisted line is a unique circumstance or an ordinary risk line is NOT SUPPLIED. That missing sentence is the whole difficulty with unique circumstances, and it is missing from most policy documents too.
How is a line with no price behind it tested?
The fourth constraint is different in kind. A minimum credit standing sits on the Rs 150 crore fixed income sleeve, 30.0 per cent of the portfolio. The standing is stated as a policy rather than a rating symbol, and no breach is recorded over the stated twelve months. But its test cannot be computed from a price the way an equity weight can: somebody has to apply a judgement to each holding and record that they did.
A line whose test is a judgement rather than an arithmetic needs the date the test was last applied written beside it. Without that date a clean report means only that nobody has reported a breach, and nobody reporting one is a different statement from nobody having found one. The record carries no such date, so the date of the last credit standing test is another NOT SUPPLIED.
So how many of this mandate's lines actually bound?
The equity range never boundA constraint binds when it actually stops something: the portfolio reached the limit and a decision had to change because of it. A constraint can be perfectly real and never bind. and was never within half its distance, the cap needed a 9.2 per cent solo rise, the no unlisted line removed nothing held, and the credit standing line recorded no breach and cannot be scored from prices. Not one of the Anantara Multi-Asset Portfolio's measurable lines bound during the stated twelve months.
A count of zero does not say the constraints were useless: a line that never binds may be the reason nobody ever proposed the thing it forbids, and that effect leaves no mark. Nor does it say everything is fine. A committee reading a clean compliance report has learned that nothing was breached, and has learned nothing about whether its constraints are doing any work, and those two questions have different answers.
Over the stated twelve months, how many of this mandate's four measurable lines actually bound?
Is a constraint that never bound a constraint that cost nothing?
No. Zero holdings and Rs 0/- is a computable and checkable number, but a constraint does not only remove things that were proposed and rejected. A constraint also removes things never proposed at all. Everybody in the room knew the line was there.
The second part is the unobserved costThe part of a constraint's cost that comes from options nobody put forward, because everybody knew the line would refuse them. The cost is real and it leaves no trace in any record.. The unobserved cost is real, and it is uncomputable from any record a portfolio produces: no register anywhere contains the proposals that were never made. Estimating it would mean inventing alternatives nobody offered and then pricing them. A number built that way is about the estimator rather than about the portfolio.
A household decided years ago never to lend money to relatives. The times somebody asked and was refused can be counted. The relatives who never asked cannot be counted. Everyone knew the answer, and no household ledger has ever contained that second number.
The no unlisted line removed no holding that the mandate actually held. Did the line therefore cost nothing?
Why does a line without a written reason disappear?
Because the reason was only ever in somebody's head, and heads leave the room. The decay works slowly enough that nobody watches it happen.
Follow one line through time. The line is written in a meeting where four people know exactly why. Two years later two have left and the remaining two remember roughly why. Two years after that the last is gone, nobody at the next review can say why the line was written, and it goes as clutter in the minute that approves everything else. A legacy lineA line still in a document after everybody who wrote it has gone, so that no one present can say what it is for. survives exactly as long as the people who remember it, so a document holding four such lines and no reasons loses about one per committee turnover.
The fix costs one sentence, recorded beside the line. The written reasonOne recorded sentence saying why a particular line exists, with the name of whoever confirmed it and the date. converts a line that depends on memory into one that depends on the file, so removing it takes an argument about whether the recorded reason still holds rather than a deletion.
A new committee finds a line nobody in the room remembers writing. What should the document already have carried beside that line?
A Rs 25 crore single holding cap is copied out of a Rs 500 crore mandate and into a Rs 50 crore one. What has that line become?
What happens when the line is copied into another document?
The Anantara mandate's cap can be written two ways that are identical at the stated total: no single holding above 5 per cent of the portfolio, or no single holding above Rs 25 crore. On Rs 500 crore both sentences mean the same thing, and nobody drafting could tell which one they had written.
Now copy each version into a Rs 50 crore mandate. The percentage version becomes Rs 2.5 crore, still 5 per cent, and the portfolio still cannot be built from fewer than twenty holdings, so the rule survived intact. The rupee version stays at Rs 25 crore. On a Rs 50 crore portfolio that is 50 per cent of everything, so a limit written to prevent concentration now permits one holding at half the portfolio. Same words, same intention, opposite effect, and the only difference is which form somebody happened to copy.
Across six target sizes, the percentage form's share and holdings count never change. The rupee form's share runs from 100.0 per cent down to 2.5 per cent.
| Target portfolio | 5 per cent form permits | Rs 25 crore form permits | Which is that of the target | Fewest holdings, rupee form |
|---|---|---|---|---|
| Rs 25 crore | Rs 1.25 crore | Rs 25 crore | 100.0 per cent | 1 |
| Rs 50 crore | Rs 2.5 crore | Rs 25 crore | 50.0 per cent | 2 |
| Rs 100 crore | Rs 5 crore | Rs 25 crore | 25.0 per cent | 4 |
| Rs 250 crore | Rs 12.5 crore | Rs 25 crore | 10.0 per cent | 10 |
| Rs 500 crore, the source | Rs 25 crore | Rs 25 crore | 5.0 per cent | 20 |
| Rs 1,000 crore | Rs 50 crore | Rs 25 crore | 2.5 per cent | 40 |
The equity range has two forms too, and it breaks harder. Carry the percentages into a Rs 200 crore mandate and 50 to 70 per cent becomes Rs 100 crore to Rs 140 crore, a range the portfolio can actually sit inside. Carry the rupee edges instead and the floor of Rs 250 crore is 125.0 per cent of that portfolio while the ceiling of Rs 350 crore is 175.0 per cent. The copied line demands more equity than the whole portfolio contains, so the mandate sits in permanent breach of a floor it can never reach and can never breach a ceiling it can never approach.
Carry the same cap into a portfolio of any size
One control moves the size of the portfolio the cap is being copied into. The source is always the same line on the same Rs 500 crore mandate: no single holding above 5 per cent, or Rs 25 crore at that total. The top pair of bars shows each form as a share of the target portfolio. The bottom pair shows the same two permitted amounts in rupees on one common scale. Each form holds steady in exactly one of those two views and moves in the other, and at Rs 500 crore they coincide and the whole problem disappears.
Copied into a target portfolio of Rs 500 crore, the 5 per cent form permits Rs 25 crore in one holding and the Rs 25 crore form permits Rs 25 crore, so both read 5.0 per cent of the target and both leave the portfolio needing at least 20 holdings. This is the source mandate itself, and it is the one size at which the two forms cannot be told apart.
Does any of this change for a private holder?
Not at all. Picture a household with one salary. The household's largest single asset outside its savings is a shareholding in the same employer that pays that salary, received over years and never sold. And the home sits in the town built around that employer's plant.
A portfolio well spread across many holdings, none of them the employer, is still stacked on a concentration nothing in the savings record shows. The unique circumstance here is not a sentiment: this household's salary, largest asset and home all depend on one thing, so a line excluding further exposure to it is a constraint about the holder's position rather than about markets. Copy that line into the document of a neighbour who works elsewhere and it is arbitrary.
How does a committee actually use this on a Tuesday?
In three columns and about twenty minutes a year. For each of the mandate's four measurable lines, and any unique circumstances beside them, Rukmini Deshpande's committee could write how many times it bound over the stated period, how many rupees it moved, and the date its test was last applied. Here the first two columns come out zero across the board, and a row of zeros is a finding rather than a formality.
Faiz Ahmad Ansari, running the mandate, gets from it a separation between the lines shaping his decisions and the lines that are simply true. A line that has never bound is not evidence that it should go, and a line that binds constantly is not evidence that it should stay, but a committee that knows which is which is having a different conversation from one that only knows nothing was breached. Anybody reading a portfolio from outside can build the same table.
The error that gets made, and what it costs
A new committee inherits a policy statement carrying an exclusion nobody in the room wrote. The new committee looks for a reason beside it and finds none. The compliance history shows it has never bound: zero holdings removed, Rs 0/- of value moved, across every period on file. The committee removes it as legacy clutter, in the same meeting that approves the year's review, and the minute records it in one line.
The line was a unique circumstance. The line existed because of a commitment the holder had made elsewhere, and that commitment has not gone anywhere. A constraint that never bound leaves no trace when it goes, exactly as it left none while it stood. The removal is therefore invisible in every report that follows. Every figure in the next four quarterly packs is what it would have been either way.
The profile is what matters here. An exclusion with an observed cost of zero holdings and Rs 0/- is precisely the profile a committee reads as harmless, and on the invented Anantara mandate the no unlisted line has exactly that profile. The cost surfaces later, in a position the holder should never have been able to take, and by then the reason is two committees away and nobody left can say what was lost.
The fix is not cleverness, it is one sentence of drafting. Every unique circumstance carries a written reason and the name of the person who confirmed it. Removing it then requires the reason to be addressed rather than the line to be deleted. A committee that cannot answer why a line exists has not established that the line is pointless; it has established that the file is incomplete, and those are very different findings.
What is a unique circumstance not?
A unique circumstance is not a preference dressed up: a preference any holder could hold transfers perfectly well and belongs with the ordinary risk lines. It is not an untestable sentence. A line that cannot be checked from the record is an instruction to the manager. It is not a claim about what the excluded thing is like: a claim of that kind would be a statement about markets. And no template can supply it. The transfer arithmetic above proves that rather than asserting it.
The whole of it reduces to one property: the words travel and the reason does not, so a unique circumstance copied into somebody else's document is a limit nobody chose, governing a holder it was never written for. Failing to travel is why the reason has to be written down, and why the uncomputable half of the cost is better named than estimated.
What is the distinguishing property of a unique circumstance, as against every other kind of constraint in a policy document?
References
| Source | Document | Where |
|---|---|---|
| The Anantara Multi-Asset Portfolio mandate | The five stated lines, the equity range, the single holding cap, the no unlisted line and the minimum credit standing on the fixed income sleeve | invented record |
| The Anantara Multi-Asset Portfolio register | Twenty eight listed equity names, largest holding Rs 23 crore, zero unlisted holdings, and the 9.7 per cent drawdown for one stated twelve month period | invented record |
| Securities and Exchange Board of India | Where the applicable requirements are published | sebi.gov.in |
The Anantara Multi-Asset Portfolio, the charitable endowment that holds it, Rukmini Deshpande and Faiz Ahmad Ansari are invented.
Educational material. Not advice on any investment, tax, budget or market position.
