The Offering Documents Compared, and What Each Adds

The three offering documents of a mutual fund scheme differ on five things: who each is written for, whether it covers one scheme or the whole range a manager runs, how much of the scheme it is complete about, what triggers a change in it, and where in a reader's journey it is reached. Compared on those, the set has no gap and no useful overlap, and the comparison ranks nothing.

Here is what sits underneath that. An offering documentA publication describing what a scheme may do, issued before a person enters it. describes what a scheme may do before anybody enters it. An offering document is not a report on what the scheme did. So a comparison of offering documents is a comparison of descriptions, and the honest output of one is a grid of what each description carries, never a verdict about the schemes being described. The single move that keeps a document comparison honest is that the columns stay descriptive: the moment a column turns into a tick, a score or a rank, the work has quietly changed subject from the publications to the schemes.

Two schemes supply every worked instance that follows. Girnar Asset Management Limited, an invented manager, operates the Girnar Large Cap Equity Fund, an open ended equity scheme, and the Girnar Broad Market Index Fund, a tracker following an unnamed index. Kalyani Bhagat manages the equity scheme and Sohail Merchant heads operations at Girnar Asset Management. Every other party around these schemes is pointed at by what it does rather than by any name: the party holding the scheme in trust, the party holding its securities, the party keeping the register of holders, the auditor and the distributor. A role is what an offering document describes, so a name adds nothing.

The three documents themselves, and why a scheme has three of them rather than one, are covered separately. The comparison below assumes all three and works only on the differences between them.

What criteria separate one offering document from another?

The criteria are chosen and written down before the first document is opened. A comparison that invents its questions while reading ends up describing each document in the terms that flattered it. Three descriptions wearing one heading is not a comparison. A criteria setThe fixed list of questions a comparison asks of every item, in the same order, decided before any item is examined. fixes the questions and the order, and then every document has to answer the same thing whether or not it has anything interesting to say there.

Think of a household comparing three schools. If they visit the first and are impressed by the playground, the second and are impressed by the library, and the third and are impressed by the teaching staff, they have not compared three schools. The household has collected three compliments. A criteria set is the boring discipline that stops that: playground, library, teaching staff, distance and cost, asked of all three in that order, including the ones where a school has nothing to show. The rows where an item has nothing to say are the rows that do the separating. Fixing the criteria before the reading starts is what keeps those rows in.

Five criteria carry the comparison, chosen as a working set rather than drawn from any published framework. First, who the document is written for. Second, whether its coverageWhether a publication is about one scheme or about the whole range of schemes a manager runs. is one scheme or the whole range a manager runs. Third, the subject the document is complete about, where completenessThe subject a document is exhaustive about, which is a different question from how long the document runs. means the subject it is exhaustive on rather than how long it runs. Fourth, what triggers a change in it. Fifth, where in a person's journey it is actually reached.

A criterion earns its place by separating. A criterion that every document scores the same way has been wasted. The commonest wasted criterion is length. Length feels like the most obvious difference between a full scheme document and a short form, and it separates least, because length is a consequence of the other four rather than a property of its own. A document written for somebody at the moment of signing is short because of that moment, not because somebody decided to be brief. Putting length in the set spends a row restating an answer the journey row has already given.

What happens when the same five criteria are run across all three?

Run them in the same order and the grid fills itself. The scheme information document is written for the person deciding whether to enter, covers one scheme, is complete about that one scheme, changes on a revision or through an addendumA notice issued to change something in a document that has already been published. that alters something already published, and is reached before applying. The key information memorandum is written for the person signing, covers one scheme, is complete about nothing while staying current on the essentials, changes whenever the full document changes, and is reached at the moment of applying. The statement of additional information is written for the person asking about the arrangement rather than about the scheme, covers the whole range a manager runs at once, is complete about the manager and the shared arrangements, changes on its own cycle, and is reached whenever a question arises rather than at any fixed point in a journey.

Read the grid down the columns rather than across one row and the finding appears. No two of the three match on all five criteria, and that is what stops any one of them being a subset of another in practice. That has to be stated carefully, because the loose version of it is wrong. Two of them do match on one criterion. On coverage, the scheme information document and the key information memorandum both answer one scheme, and only the statement of additional information answers the whole range. A grid with a single criterion in it would therefore have shown two of these documents as the same object. Five criteria show they are not. The argument for a criteria set fits in one sentence: any single question can make two different things look identical, and the separation lives in the combination.

The dashed row in that grid is worth pausing on. The dashed row marks the coverage criterion, the one place where two of the three give the same answer. A comparison that quietly drops its non separating rows has started editing its own evidence, so that row is drawn rather than hidden. A criteria set is judged on the whole set and not on its best row, so the row where two items agree stays in the grid and does its job of showing that this question alone would not have been enough.

The journey criterion explains the row everybody reaches for first, so it deserves its own picture. The scheme information document is reached while a person is still deciding, so it has room to be complete. The key information memorandum arrives at the moment of signing, attached to the thing being signed, so it has to be readable standing at a counter. The statement of additional information is reached whenever a question about the arrangement comes up, and a moment that can arrive at any time is not a point in the journey but a permanent availability. Length is downstream of the moment, and once the three moments are visible the difference in bulk stops looking like a choice anybody made.

Does the offering set overlap itself, or leave a gap inside it?

Neither, and the two halves of that answer are worth taking one at a time. Start with overlap. The key information memorandum and the scheme information document plainly cover the same ground, and a reader meeting that for the first time usually calls it duplication. The overlap is repetition rather than duplication, and the difference is that the short form draws on the full document and stops at the same edge. The short form does not reach past that edge. Measured as new ground covered, the key information memorandum adds exactly nothing to the offering set, and that is a property of the set rather than a criticism of the document.

Now the gap. Between one scheme described in full and the shared arrangements described in full, there is no strip of the offering left uncovered. The scheme information document handles what this particular scheme may do. The statement of additional information handles the manager and everything common to every scheme it runs. The two regions meet at a seam, and the seam is clean: nothing that belongs to a description of the offering falls between them.

A reader does still walk away with gaps, and it is important to be precise about what kind. The holdings of the Girnar Large Cap Equity Fund on any particular date are not in its offering documents, and neither is what one of its units is worth today. Neither absence is an omission. A description of what a scheme may do was never built to report what it did, so the missing answers are gaps in the purpose of the set rather than holes in the documents. Where those reports live is covered separately.

How to compare fund documents without producing a recommendation?

One discipline carries the whole of it: compare the documents on what each carries, never on the schemes they describe. The distinction sounds small and it is the entire boundary. Reading the Girnar Large Cap Equity Fund's offering documents beside the Girnar Broad Market Index Fund's answers what each scheme may do, how each states its charges, and what each says about the risks it carries. Reading them does not answer which scheme is better, and it cannot. An offering document describes the boundaries a scheme is permitted to work inside, so a comparison of two of them is a comparison of permissions, and permissions are not results.

Here is the everyday version. Two neighbours hand over the rental agreements for the two flats they let out. The agreements can be compared in enormous detail: what each permits, what each charges, what notice each requires, what each says about repairs. At the end of it both agreements are known thoroughly. Nothing about the plumbing, the neighbours or the morning light was ever in either document, so which flat is the better place to live is still unknown. A document states what may happen, and a reader who wants to know what did happen has to go somewhere else entirely.

There is a tell that a comparison has drifted, and it is visible in the grid rather than in the reasoning. The moment a column heading turns into a score, a tick or a rank, the work has stopped being about the publications. A tick means a bar has been set. A bar means a standard has been chosen. A standard chosen by the reader, applied to two publications, and reported as a difference between two schemes is a verdict resting on evidence the publications could not have supplied. Nobody sets out to do this. The drift happens because a grid full of descriptive cells looks unfinished, and a tick makes it look decided.

So say what the honest output looks like. Naming it is what stops the drift. The output is a grid of what each document carries, with the rows where the documents differ marked, and the reason for each difference named beside it. A difference because the two schemes do different things is one kind of note. A difference because one manager writes at greater length is a different kind of note, and saying which is which is most of the value. A structural comparison stops at structure, and stopping there is a finished output rather than an abandoned one.

The discipline only becomes real when it costs something, so one row of that grid is worth working properly. The charges row on the Girnar Large Cap Equity Fund. Its scheme information document sets out two plansOne version of the same scheme, carrying its own expense ratio, with the holdings identical across versions. of the same scheme, one carrying an expense ratio of 1.65 per cent a year measured against the scheme's assets, and one carrying 0.85 per cent measured on that same base. The gap is 0.80 percentage pointsThe unit for the difference between two percentages, used so that a gap is not confused with a proportional change. a year, and it closes both ways: 0.85 plus 0.80 is 1.65, and 1.65 less 0.80 is 0.85. The key information memorandum for the same scheme carries the plan names and sends the reader to the full document for the terms.

Sit with what has just been established and what has not. The comparison has established where the charge information lives and in what detail it lives there, a real and useful finding about two publications. It has established nothing about whether the wider charge is worth paying. The 0.80 point gap between the two plans of the Girnar Large Cap Equity Fund pays for distribution service, the holdings behind both plans are identical, and what that service is worth to any particular household sits nowhere in this record. The arithmetic stops at the gap. Whether the wider charge is worth paying to any particular household is a judgement, and no arithmetic settles a judgement. How the two plans differ in themselves is covered separately.

How to compare collective-vehicle documents at a structural level?

Mutual fund schemes are not the only arrangement that pools money from many entrants and publishes documents about it. Every other kind of collective vehicleAny pooled arrangement that holds assets on behalf of many entrants at once. publishes its own, and the documents have different names, different lengths, different readers and different rules behind them. The instinct on meeting a new one is to line the names up beside the three described above and look for the equivalent of each. The instinct fails. The names are not equivalent, and matching them produces a table that is confidently wrong.

A structural comparisonA comparison that asks what each document is built to carry, rather than comparing what the documents happen to say. refuses to line the names up and asks the same four questions of each set of documents instead. First, what kind of vehicle it is and who is accountable for it. Second, who may enter it and on what terms. Third, what is disclosed and to whom. Fourth, who checks the disclosure. The four questions ask what the documents are built to carry rather than what any of them happens to say, so they work on any pooled arrangement's documents, including an entirely unfamiliar one.

The questions and the answers behave differently. Every answer is fixed by rule, the rules differ by vehicle, and the rules move. The four questions travel and not one of the four answers does. Who is accountable for a vehicle, who may enter it, what must be disclosed and who checks it are all set by the Securities and Exchange Board of India (SEBI) for each kind of arrangement, separately, and revised. The four questions therefore travel in full while each answer is read from SEBI at sebi.gov.in in the place where that answer belongs. A printed answer would stop being merely dated on the day it changed and would start being wrong.

The practical gain is the reason to memorise four questions rather than three document names, and it is worth stating plainly. A reader carrying the four questions can walk into any pooled arrangement's documents, including one nobody has described to them, and know exactly what to look for and in what order. Somebody carrying only the names of the three mutual fund documents can do that for mutual fund schemes and nothing else.

The refusal to fill those rows in has to be drawn rather than described, and this is the shape it takes: a card with all four rows named and every one of them left as a blank rowA labelled field left empty on purpose, with the source that fills it named inside the empty space., carrying SEBI at sebi.gov.in inside the blank. A card with those rows filled in from recollection is a liability: it still looks authoritative on the day the rule behind it changes. A blank card with the source named in each row keeps working for as long as the source exists. The reader leaves with an instrument rather than a snapshot.

Why do three of these questions come back with no document at all?

Sending a question to a publication rather than trying to recall the answer is covered separately, and the short version of the routing with it. The criteria set has now made something else visible. A question that lands on no offering document can fail to land for two completely different reasons, and the two need different responses from the reader. Eight questions about the Girnar Large Cap Equity Fund follow, written out in full. Three of the eight come back with no offering document, and they split into a question whose answer sits with the rule maker and two whose answers sit in publications of an entirely different kind.

1. What the scheme says it is trying to do. The scheme information document, in full, with a summary in the key information memorandum.
2. What the plans and the charges are. The scheme information document, in full, with the plan names in the key information memorandum.
3. What the load terms are. The scheme information document, in full, with a summary in the key information memorandum.
4. Who holds the securities. The statement of additional information, which describes the custodian role.
5. Who keeps the register of holders. The statement of additional information, which describes the registrar and transfer agent role.
6. How often the document is revised. No offering document answers this. The revision cycle is set by SEBI at sebi.gov.in, so the row stays empty with the source named inside it.
7. What the scheme held at a stated date. No offering document answers this. It is answered by a report on holdings, which is a different publication and is covered separately.
8. What a unit is worth today. No offering document answers this either. It is a value struck for the day rather than a description of anything.

Lumping those last three together is what wastes a reader's afternoon, so separate them. The revision question has an answer, and the answer sits with the rule maker rather than in any publication Girnar Asset Management issues. A reader who keeps opening documents looking for it will keep finding nothing, correctly. The other two do have answers in publications, just not in offering documents, and a reader who knows that walks straight to a different kind of publication instead of concluding that nobody discloses it. One refusal sends the reader to a regulator and the other sends the reader to a report, and telling them apart is the whole practical value of separating them. A routing tool that can only ever point at a document is worse than useless on the questions where the honest answer is that no document carries it.

What does the same grid look like run against a second scheme?

Run the five criteria against the Girnar Broad Market Index Fund's three documents and something quietly useful happens: all five answers come back the same as they did for the Girnar Large Cap Equity Fund. The scheme information document is still for the person deciding, still about one scheme, still complete about that scheme. The key information memorandum is still the short form reached at the moment of applying. The statement of additional information still covers the whole range Girnar Asset Management runs. One document therefore serves both schemes. The criteria set describes the document types rather than the schemes, so running it against a second scheme returns the identical five answers and proves the criteria were about publications all along.

Now take the one row where the two schemes genuinely differ, a content row rather than a criteria row, and watch how far it can honestly be pushed. The Girnar Large Cap Equity Fund's scheme information document sets out two plans of the same scheme at 1.65 per cent and 0.85 per cent of the scheme's assets a year. The record for the Girnar Broad Market Index Fund states one ratio, 0.20 per cent of the scheme's assets a year, and says nothing whatever about how many plans that scheme has. The honest cell for the second scheme's plan structure is not a guess and not a zero: it is a note that this record carries no entry there, and a document comparison that fills such a cell has invented a disclosure.

The two percentages are the kind of pair that looks sloppy and is not, so they are worth one careful sentence. Sixteen thirty thirds of the wider ratio is 48.4848 recurring per cent, rounded down to 48.48. Seventeen thirty thirds is 51.5151 recurring per cent, rounded up to 51.52. The two roundings went in opposite directions by exactly the same amount. Neither part is exact on its own, and the two rounded parts still add to 100.00 exactly. State the fraction as the exact figure and the percentage as its approximation, and the check row then closes in both directions rather than appearing to.

And here is the line the whole worked instance exists to draw. Those two ratios belong to two different schemes doing two different things, and neither number was ever a measure of how well either scheme is run. Setting 1.65 against 0.20 is arithmetically available and analytically empty. The grid has established where each scheme's charge is disclosed and in what detail, a finding about two sets of publications, and nothing whatever about which scheme anybody should hold.

Do the counts that repeat here mean anything?

Two coincidences are worth flagging so that nobody reads a relationship into them. There are three offering documents, and three of the eight routing questions come back without an offering document answer. The two threes are unrelated: the first counts publications and the second counts questions, and no question is paired with a document by that arithmetic. Likewise the Girnar Large Cap Equity Fund has two plans and the comparison runs across two schemes. Both are collisions of counting rather than relationships, and naming them is cheaper than letting a reader spend a minute looking for a pattern that is not there.

What does a document comparison establish, and where does it stop?

A document comparison establishes three things and they are worth having. First, which publication to reach for, given a question. Second, what each publication is complete about, and so the point at which looking can stop. Third, and most valuable of the three, where a reader will be left guessing. Knowing that no offering document reports what a scheme actually held saves an hour of searching for it. A document comparison is a map of where the answers live, and a map that starts ranking the destinations has stopped being a map.

The same comparison establishes nothing about the scheme's merits and nothing about the manager's skill, and the reason is structural rather than cautious. Every one of the five criteria is a property of a publication. Not one of them touches how the Girnar Large Cap Equity Fund is actually run, what Kalyani Bhagat decides on a Tuesday, or how Sohail Merchant's operations team handles a busy day. Decisions and operations leave traces in reports and in records, and none of the traces is in a document written before anybody entered. A comparison can only ever establish something about the evidence it was run on, and the evidence here is a set of descriptions.

Girnar Asset Management Limited, the Girnar Large Cap Equity Fund, the Girnar Broad Market Index Fund, Kalyani Bhagat and Sohail Merchant are invented.
Educational material. Not advice on any investment, tax, budget or market position.