How to Read a Mutual Fund Account Statement, Line by Line

Work down a mutual fund account statement in a fixed order. Fix whose record it is and what stretch of time it covers, then the asset managers named, then the accounts under each, then each scheme line. Take the closing unit count before any rupee total. Finish by hunting deliberately for a holding that never appeared at all.

A common account statementA single record that gathers what one holder holds across more than one asset manager into one document. has arrived, it runs to several sheets, and the reader has four minutes and no idea which end to start from. Nine steps run top to bottom, in an order fixed on purpose. The whole value of a procedure is that nothing gets decided while it is being run.

One invented record carries every step. Girnar Asset Management Limited, an invented asset manager, operates the Girnar Large Cap Equity Fund. The fund is an open ended equity scheme holding net assets of Rs 4,200 crore against 120.00 crore units in issue, so one unit stands at Rs 35.00 exactly. The division of Rs 4,200 crore by 120.00 crore units is worked out in full where these record documents are set beside each other, and is carried here as settled rather than run again. The Girnar Large Cap Equity Fund is held across 3,80,000 accounts and carries a running charge against its assets of 1.65 per cent of net assets a year. Kalyani Bhagat manages the portfolio of the Girnar Large Cap Equity Fund and Sohail Merchant heads operations at Girnar Asset Management.

Three matters are settled elsewhere and treated here as known: what an account is and how a periodic report differs from it, what a unit is and how a value per unit is struck, and what a subscription, a switch and a redemption actually do. Each object printed on a record is defined under its own heading.

In what order should a mutual fund account statement be read?

In the order printed below, and the order is the whole point. Most people open a statement, let their eye fall on the largest rupee number, feel either fine or worried about it, and shut the document. Letting the eye fall on the largest number is not a reading. A reading has a starting field, a route through the document and a rule that says when to stop, and it produces something that can be acted on rather than a mood.

The nine steps form a single shape, worth seeing before any one of them is used. The first two settle the scope of the document. The next two settle its structure. The fifth and sixth read the figures, in a deliberate order. The seventh looks for what is not printed. The eighth names the fields the record is silent about. Naming them stops a reader interrogating a document that cannot answer. The ninth is a stopping rule. A reader who cannot tell when they are finished either quits early or worries indefinitely, so a procedure without a stopping rule is not a procedure at all.

Consider someone walking into a room to check that the tiffin boxes for a wedding lunch have all arrived. The check does not start by opening one. The check starts with which lunch and which hall. Those two answers decide what a full count even means. Then the caterers are counted, then the crates each caterer brought, then what is inside a crate. Then comes the step nobody takes: what should have arrived and has not. A statement is read exactly that way.

What is the first thing to check, before any figure at all?

Two things, together, and neither of them is a number anyone cares about. Whose record is this, and what stretch of time does it cover. Both are read off the top of the document before the eye reaches anything else.

Why first rather than last? Because both facts are scopes, and a scope decides what every figure underneath it means. A reader who checks the covering dates at the end has already misread everything printed above them. Those figures were read without knowing what set they belonged to. The most common way this goes wrong is entirely undramatic: a holding created after the periodThe stretch of time a record covers, printed at the top of it. Anything created outside that stretch is correctly missing from the record. closed is correctly absent, and a reader who never noted the period reads that absence as a loss and spends a week worried about it.

The other half, whose record it is, sounds too obvious to check, and it is the half that catches people. One household can receive records addressed to different people in it. A reader working from a record drawn on one person, while counting holdings that sit in another person's name, will get every count below wrong. Nothing in the document will say so. From the document's point of view nothing is wrong at all.

Which key was the report drawn together on?

The identifying detail printed near the top is not decoration. Find it and note it. A record that gathers holdings across several asset managers was produced by searching a register with something, and that something is the identity keyThe identifying detail a record was searched on. Whatever the search ran on sets the outer limit of what the record can possibly show..

The key sets the outer edge of everything below it, so knowing what the search ran on establishes in advance what this document could not possibly contain. That single sentence is why step two exists and why it comes before the counting. A search cannot return a holding it never looked at. If an account was opened years ago carrying a different identifying detail, the search does not fail loudly and print a warning; it simply does not find that account, and the record it produces looks complete because it has no way of showing the shape of what it missed.

The key itself, the way it is checked, and what happens when two records carry different versions of it, are each taken up separately. Step two involves one thing only: reading the key, writing it down, and holding it in mind while counting. The permanent account number (PAN) issued by the tax authority and the know your customer record (KYC) both sit behind records of this kind, and both are explained separately.

How many asset managers appear, and how many were expected?

The asset managers named are written out as a list, and the list is counted. Then, separately, the reader counts how many asset managers the household is believed to hold with. Two counts, made independently, and the second one is not in the document at all.

The count is the reader's check rather than the record's, and a check like that only works when it is made before the answer is looked at. Someone holding with three asset managers who finds only two named has landed on something precise: a single gap, pinned to a single asset manager, that can be typed into a letter and sent. A reader who never counted is holding the same document and has found nothing. With no expectation to compare against, every list of names looks equally complete.

A shopkeeper counts the day's delivery against the order slip rather than against the crates on the floor for the same reason. The crates on the floor always look like a delivery. Only the slip knows whether one is missing, and the slip carries information that came from somewhere else.

How are the accounts sitting under one manager found?

Under each asset manager the record names, look for the accounts, and expect more than one. Do not treat a second account under the same asset manager as a fault to be reported. Read it, note it, and move on.

An account started one year and another started later under a different set of names have no reason to fuse. Where one household turns up more than once beneath a single asset manager, the register is behaving normally rather than misfiring. A person opens an account in their own name in one year. Three years later the same person opens another, this time held jointly with somebody else in the same household. The two accounts carry different combinations of names. A human being would recognise them as the same household, and no system quietly consolidates them on that ground. The register keeps what it was told to keep.

The household in the worked example runs exactly this shape. Under Girnar Asset Management there are two accounts: one held singly, one held jointly. Both are in the Girnar Large Cap Equity Fund. Both are correct. A reader who expected one account, finds two and assumes an error wastes the next hour on a query that has no answer. Nothing has gone wrong.

On each scheme line, which of the two figures comes first?

The closing unit balance, and the rupee figure afterwards. The order is reversed from what almost every reader actually does, and reversing it is the single most useful move in the whole reading.

Nobody shifts the unit count without giving an instruction. The rupee amount beside it is that same count multiplied by a value struck on one particular day, and it will read differently tomorrow with nobody having lifted a finger. Take the number that shifts for a reason ahead of the number that shifts regardless. The closing unit balanceThe number of units standing in an account at the end of the stretch of time a record covers. is the holding. The rupee figure is a photograph of the holding taken under one day's light.

Take the invented household. Account one in the Girnar Large Cap Equity Fund carries 2,857.143 units, being Rs 1,00,000/- divided by Rs 35.00 a unit. Account two, held jointly, carries 1,428.571 units, being Rs 50,000/- divided by the same Rs 35.00. Adding the two counts, and stopping there for a moment, gives 4,285.714 units. The unit total is the number to hold on to, and it would be identical on a record printed on any other day this month. Only then comes the multiplication. At Rs 35.00 a unit, 4,285.714 units come to Rs 1,49,999.99, or about Rs 1,50,000/- by rounding rather than by equality.

Here is the everyday version. A vegetable seller counts sacks of onions at the close of the day, not the value of the onions. The sack count changes when somebody buys or delivers. The value changes overnight because a market moved in another district. Both are true numbers, and only one of them tells the seller what happened in the shop.

Why does multiplying the printed unit count back not return the money paid?

Because the printed unit counts have already been rounded, and a rounded number multiplied back cannot reconstruct the number it came from. The rounding is worth thirty seconds of attention. A reader who does not know about it will one day multiply, land a paisa short, and conclude that something is wrong when nothing is.

Take the division without rounding anything. Rs 1,00,000/- divided by Rs 35.00 is 2,857.142857 and keeps going, a quotient that never terminates. The record cannot print an unending decimal, so it prints 2,857.143. Multiplying that printed count back at Rs 35.00 gives Rs 1,00,000.005, half a paisa more than the money that went in. The same on the joint account: Rs 50,000/- divided by Rs 35.00 is 1,428.571428 and onwards, printed as 1,428.571, and multiplying back gives Rs 49,999.985, one and a half paise less. Put together, the pair comes to Rs 1,49,999.99 against Rs 1,50,000/- paid in, exactly one paisa short.

Rs 1,00,000.005 sits exactly halfway between Rs 1,00,000.00 and Rs 1,00,000.01, so neither residue can be stated to the paisa. Any answer to the paisa would be a rounding rule quietly chosen rather than a fact read off the record. The same is true of Rs 49,999.985, which sits exactly halfway between Rs 49,999.98 and Rs 49,999.99. The residues stand as check rows, unresolved. Resolving them would mean choosing a rounding rule rather than reading a fact.

So when a total is quoted as about Rs 1,50,000/-, the word about is doing real work and is not politeness. The exact product of the printed counts is Rs 1,49,999.99, and the one paisa is not a fee, not a loss and not an error in the record. The odd paisa is the shadow of a division that does not come out evenly.

How are the movements the statement lists checked?

Each entry is taken in turn and held against the instructions the household remembers giving. Checking each entry that way is the whole of step six. The two ways a record and a memory can disagree are not the same problem and do not get the same first question.

A movementAn entry showing units going into or out of an account, as against the balance those entries leave behind. that is listed and is not recognised is worth asking about straight away. One that is remembered and cannot be found is far more often a query about how far this document reached, or about the dates it spans, than about the instruction itself. The asymmetry is worth holding on to because it saves the wrong phone call. Something on the record that nobody asked for is a question about the account. Something missing from the record that was asked for is, far more often, a question about the document.

Neither reading is a verdict. A holder who finds an entry they do not recognise has not established that anything improper happened, and a holder who finds one of their own instructions missing has not established that it failed. Both have found a question with a name attached to it, and a named question is exactly what step nine asks for.

How is a holding that never appears at all found?

Deliberately, and last. The list made at step three and the accounts found at step four are set against what is known to exist. Every account that did not appear is then written down. Writing down the absences is the whole of step seven, and it is the step readers skip.

Step seven is the one step that runs on knowledge the record in front of the reader never held, and that is exactly why it drops out: every other step rewards looking harder at the document, and this one does not. An absenceA holding that exists but never appears on a record, so the record itself carries no mark showing that anything is missing. leaves no mark. There is no grey box, no footnote and no warning. The record simply ends, and a record that ends looks finished.

Now say plainly what an absence means and what it does not. Units held under a key the search never matched, or inside an account whose verification entry was left unfinished, fall outside a report while staying exactly where they were in the account itself. Nothing has been taken. Nothing has lapsed. An absence on a record is a question about the report, not a verdict on the holding.

Run the invented household through it as a second reading of the same document. Suppose the search ran on INVENTED-KEY-0004-A and the jointly held account was opened under INVENTED-KEY-0004-B. The record then prints one account, 2,857.143 units in the Girnar Large Cap Equity Fund, and about Rs 1,00,000/-. The record has reached two thirds of the position, or about 66.7 per cent of it, computed as Rs 1,00,000/- over Rs 1,50,000/-. The missing third is 1,428.571 units, being Rs 50,000/- of units, and those units have not moved anywhere at all. The units are in the account. The units are not in the report.

What will the statement never state, however carefully it is read?

Four things, and naming them is step eight. Whether the bank instruction attached to the accounts still points at an account that is open. Whether a nomination stands. Whether the contact details on the register still reach the holder. And what the scheme charged the holder, a large enough matter to be taken up on its own.

The first three share one shape. A record reports positions and movements. A record reports what is there. A record does not report what is not there. None of those three fields produces an entry when it fails, and failing quietly is exactly what those fields do. A bank instruction pointing at a closed account does not print a warning. A nomination that was never recorded does not print an empty space with a label on it. An address that stopped working does not announce itself, least of all to the person it stopped reaching.

A document that records no absences cannot be made to yield one however long it is stared at, so the standing of those fields has to be requested from the asset manager directly and in plain words. None of that is a caution or a failing on anybody's part. The silence is a fair statement of what this class of document is built to do, and knowing it saves a reader from a search that has no ending. And if any of those fields has lapsed, that is how record systems behave over a decade rather than how a person fails: a form existed, nobody said why it mattered, and a household got on with its life.

How large is the charge that the statement does not print?

Large enough to be worth knowing, and completely absent from the document, and both of those facts come from the same place. Work the size from the invented record, naming the base every time a ratio appears.

Start at the scheme. The Girnar Large Cap Equity Fund carries an expense ratioA running charge set against a scheme's own assets, quoted as a percentage of those assets for a year. of 1.65 per cent of net assets, and its net assets are Rs 4,200 crore, so the scheme gives up Rs 69.30 crore across a year. A year's charge of Rs 69.30 crore is about Rs 0.1899 crore on any one day of a 365 day year. Bring that down to one account by two routes that share no arithmetic. Route one divides: Rs 69.30 crore across 3,80,000 accounts is about Rs 1,824/- an account for the year. Route two multiplies: the average account is Rs 4,200 crore over 3,80,000 accounts, about Rs 1,10,526/-, and 1.65 per cent of Rs 1,10,526/- is Rs 1,823.68.

The two routes are not identical. Route one, carried unrounded, gives Rs 1,823.6842 and onwards. Route two starts from an average already rounded to Rs 1,10,526/- and so lands about half a paisa lower at Rs 1,823.68. Run on the unrounded average of Rs 1,10,526.31578 and onwards, route two returns Rs 1,823.6842 as well, and the check passes. Two routes agreeing to the fifth decimal place once the rounding is taken out is what establishes that the arithmetic is right, rather than merely plausible.

Now the household described here. Its holding across both accounts is about Rs 1,50,000/-, and 1.65 per cent of Rs 1,50,000/- is Rs 2,475/- for the year, or about Rs 6.78 on any one day of a 365 day year. Rs 2,475/- is a real amount of money. The sum is roughly what the household would spend on a week of vegetables.

Not one of those figures shows up anywhere on the record, and nothing is being hidden. The charge is absorbed inside the value per unit and never billed to anybody, so no line exists for it to sit on and a search of the document has no end. The nature of that charge, the way it is levied and any ceiling on it are covered separately, and what may be charged at all is a matter for the Securities and Exchange Board of India (SEBI) at sebi.gov.in.

When is the reading finished?

On one of two endings, and not before either of them. Ending one: every account known to exist has appeared on the record. Ending two: a written list of the accounts that did not appear is in hand, with the asset manager each one should have come from named beside it. The pair of endings is the stopping ruleThe condition that tells a reader the reading is finished, so it ends on a test rather than on running out of patience., and a procedure without one is not finished, it is merely abandoned.

A named list is the thing any route can pick up, and an uneasy feeling is not, so finishing on the second ending counts as a full reading rather than a botched one. This is worth sitting with, because it inverts how most people feel at the end of a difficult document. A reader who finishes with two named gaps has done better work than a reader who finishes with a warm feeling and no list. The first can be acted on by somebody else. The second cannot be acted on by anybody, including the reader.

Notice what is not an ending. Running out of patience is not an ending. Deciding the total looks about right is not an ending. Neither of those is a test, and a reading that stops on a mood produces nothing anybody can carry forward. Step nine is written as a rule rather than as a suggestion for exactly that reason.

What is never a step in this reading?

Anything done with the rupee amounts beyond simply reading them. There is no step ten in which judgement is passed on the numbers, and no step at all in which the record states how a holding has fared.

A document of this kind keeps score of nothing: no line on it reports whether a holding fared well or badly, and no amount of further reading will make one do so. The rupee column exists so a holder can see the size of the position on one day. The rupee column is not a score and carries no comparison. A single closing figure has no starting point attached to it and no measuring stick beside it.

Performance stays outside this reading for the same reason. Working out how a holding has done needs a period, a basis, and a statement of whether costs are inside the figure or outside it, and none of those three is on a statement. Return questions are worked properly under scheme performance, and attempting them from a record would produce a number that looks authoritative and means nothing.

What does the whole reading look like on one statement?

Here is the invented household run end to end, one row a step, so the whole procedure is visible in one place. Girnar Asset Management Limited is the only asset manager on this record, and the Girnar Large Cap Equity Fund is the only scheme.

Now the same document a second time under the absent case, the reading worth practising. Suppose the account held jointly was opened under a key the search never reached. Step four now finds one account rather than two. Step five reads 2,857.143 units and about Rs 1,00,000/-. Step seven, and only step seven, catches it: the reader's own list said two accounts under this asset manager, the record shows one, and the gap is 1,428.571 units of the Girnar Large Cap Equity Fund. Step nine then ends on ending two, with one written line naming the account and the asset manager it should have come from. The document was identical in both readings, and only the reader's own list of what should be there separated a complete reading from a comfortable one.

Girnar Asset Management Limited, the Girnar Large Cap Equity Fund, Kalyani Bhagat, Sohail Merchant, the household described here and the identity keys printed on it are invented.
Educational material. Not advice on any investment, tax, budget or market position.