Fund Accounting: How a Scheme's Own Books Are Kept

A scheme's books set down what that scheme is holding, what it owes, what it is owed, what came in and what it ran up, plus the count of units standing against all of it. The books belong to the scheme, not to the organisation managing it, and an auditor appointed for the scheme examines them. One unit needs a value every working day, so the books are ruled off that often.

Every rupee below comes off a single invented record, set out here once so that no later block has to restate it. The scheme is the Girnar Large Cap Equity Fund, open ended and holding equities; the manager behind it is Girnar Asset Management Limited; and the five lines beneath are the only quantities that record supplies.

Two people appear by name and nobody else does. Kalyani Bhagat decides what the Girnar Large Cap Equity Fund holds. Sohail Merchant runs operations. The roles are the subject here, so every other party appears as a role rather than a name.

None of the following four gets rebuilt below, so take all four as already known: what a scheme is, what a unit is, how one unit gets given a value each day, and that the running cost bites on the scheme's assets rather than arriving as a bill. None of those four said out loud that all of it sits on top of an ordinary set of books, written up by people, checked by other people, and closed at the end of every working day. That set of books is the subject here.

What does a scheme's own set of books actually record?

Six things, and all six can be held in mind at once. Whatever the Girnar Large Cap Equity Fund is holding. The amounts it owes. The amounts owed to it. The income that came in. The costs it ran up. And how many units stand against everything above. Six lines, and the list is complete. Fund accountingThe bookkeeping done for a pooled investment scheme, kept for the scheme rather than for the organisation that manages it. is not a larger or stranger discipline than the bookkeeping a shop does; it is the same discipline pointed at a pool of assets instead of at a trading business.

Two of the six are the ones readers skip over, so name them slowly. A payableAn amount a party has already incurred and still has to hand over, recorded as owing from the moment it is incurred. is an amount the scheme has already incurred and has not yet handed over. A receivableAn amount a party has already become entitled to and has not yet collected, recorded as due from the moment the entitlement arises. is an amount the scheme has become entitled to and has not yet collected. Both sit in the books from the moment they arise, not from the moment money moves, and that single habit is what makes the day's figures answerable.

Then compare the list against a shop and the difference is immediate. A shop buys things and sells them to customers at a margin, so a shop's books have a sales line and a cost of goods line. A scheme has neither: it does not sell anything to anybody and it produces nothing, so there is no revenue line and no cost of production line in a scheme's books at all. What looks like income in a scheme's books is income the scheme received on what it was already holding, and what looks like cost is the running cost of holding it. The shape is different because the business is different, and a reader who arrives expecting a shop's shape spends the first ten minutes looking for lines that were never going to be there.

The sixth item is the odd one and it gets its own part later. Unit capitalOne total, standing in a scheme's own accounts for however many units happen to be outstanding when a day is ruled off. stands in the books as one number for the whole scheme, 120.00 crore units on the Girnar Large Cap Equity Fund. The same quantity sits split across 3,80,000 folios in a separate record kept by the registrar and transfer agent. One quantity, two records, kept by two parties from two different starting points. The pairing of those two records is the only check below that can actually fail.

What are the four operations a scheme repeats every working day?

Four, in order, and none of them can be skipped. Value the holdings at what the written policy says each is worth for that day. Bring in what the scheme has earned and set down what it has incurred, whether or not any money has moved either way. Add the whole thing up to a net figure. Divide that net figure by the units in issue. The fourth operation produces the number that gets published, and the first three exist to make the fourth honest.

Most readers have never framed that sequence this way, so say out loud what it actually is. The four operations together are a full accounting close, and the Girnar Large Cap Equity Fund performs one on every single working day rather than at the end of a month or a year. A shop counts stock and rules off its books once a month and takes a week over the year end. A scheme rules off its books before it goes home, every day it is open, and publishes the result. There is no quiet fortnight in which to find a mistake and no year end during which the world waits.

The pressure of a daily close is the reason the operating arrangement around a scheme looks the way it does. Every party feeding the daily closeThe end of day exercise in which a set of books is ruled off and the day's figures are finalised. has to be ready on the same clock: the holdings and their values, the day's earnings, the day's costs, and the day's movement in units. A value per unit is expected whether or not the inputs behaved, so miss one input and the fourth operation still runs. A bad input running through unnoticed is precisely why a scheme keeps checks against records it did not itself produce, and why the arithmetic of the close is deliberately dull.

Think of a household kitchen at the end of a wedding day. Somebody has to know what was spent, what is still owed to the tent supplier, what the caterer has not yet billed and what is left in the tin, and they have to know it before anybody can say how the household stands. A scheme is that household, except the wedding is every working day and the answer is published.

Why is the running cost accrued on every single working day?

Because a cost that has been incurred belongs in the books on the day it is incurred, whatever day it is settled on. The habit of recording on the day of the obligation has a name. An accrualRecording an amount in the books on the day it is incurred or earned, rather than on the day cash changes hands. puts an amount into the record as owing from the moment the scheme has run up the obligation. A charge therefore shows in a scheme's books well before anybody hands money over for it. The Girnar Large Cap Equity Fund runs at 1.65 per cent of net assets a year, so on any given working day the scheme has run up one day's worth of that, and one day's worth goes into the books on that day as an amount owing.

Work it in rupees so that it stops being an idea. On net assets of Rs 4,200 crore, 1.65 per cent for a year is Rs 69.30 crore. Spread across a year taken as 365 days, that is about Rs 0.18986 crore a day, or about Rs 18,98,630/- set down as owing on an ordinary working day before the fourth operation runs. Across 120.00 crore units that same amount is about Rs 0.0015822 a unit for the day, under a sixth of one paise on a value per unit of Rs 35.00.

Here is the consequence readers most often miss, and it is worth sitting with. Each of those days ended with one unit given a value only once that day's amount had already gone in, so somebody who held units of the Girnar Large Cap Equity Fund across part of a year has borne exactly the cost that went in on the days they were there, no more and no less, and no charge was ever raised against their folio at any point. Forty days in the scheme is 8 parts in 73 of a year, and it is 8 parts in 73 of the year's running cost per unit held, arrived at without anybody computing anything for that holder specifically. The arithmetic did itself, one day at a time, inside the published figure.

The everyday version is a shopkeeper who raises the price of a kilo of rice by two paise on a Tuesday. Nobody is billed two paise. Nobody sees a two paise line. Every customer who buys rice that Tuesday pays it, every customer who does not, does not, and the shopkeeper never has to work out who owes what. A scheme's running cost travels the same way: into the price, not onto an invoice.

Why does unit capital behave unlike a company's share capital?

Because a company's share capital sits still and a scheme's unit capital does not. A company changes its share capital when it raises money or returns money, and both are events: they are decided on, announced, dated and recorded, and years can pass between them. The number in the books is stable enough that a reader treats it as a fixed frame around everything else.

The Girnar Large Cap Equity Fund has no such frame. Units are created when somebody subscribes and cancelled when somebody redeems, and both happen on any working day on which anybody chooses to do either. In an open ended scheme unit capital shifts on essentially every working day, without anybody deciding it should, and that is why these books cannot be read the way a company's are: the thing everything gets divided by will not hold still. One unit is worth the net figure over the unit count, and in a company's books that divisor would be among the steadiest numbers in the accounts.

Two things follow from that, and both are practical. The first is that tomorrow the divisor is different, so no figure in a scheme's books means anything without the date and the unit count that go with it. The second is that a change in the value per unit does not establish that the scheme grew or shrank: net assets and units in issue both move, and they do not have to move together. A scheme can take in a great deal of money on a day its value per unit fell, and there is nothing contradictory about that at all.

The household version is a shared cooking pot rather than a share certificate. Ten people put in and eat from one pot, and the eleventh joins at lunchtime with their own contribution. The newcomer added in proportion to what they took out, so the pot is bigger and each share is not automatically smaller. A company's share register is a list of certificates that mostly sits still. A scheme's unit count is the pot, recounted at the end of every day.

At what point do these books end and the manager's own books begin?

At a hard line, and the whole arrangement rests on it. The Girnar Large Cap Equity Fund's assets are not Girnar Asset Management Limited's. Nor are the scheme's books the manager's books. The scheme's books and the manager's books are two separate sets of records, kept for two separate things, describing two different questions, and the fact that one organisation writes both up does not merge them any more than a bookkeeper writing up two shops merges the shops.

The running cost is the clearest crossing available, so follow it across that line. About Rs 18,98,630/- is set down inside the Girnar Large Cap Equity Fund on an ordinary working day as an amount the scheme owes. Once it is settled, it is gone from the scheme's books completely. A different set of books, covered separately, records what happens to that amount afterwards, which parties receive what share of it, what any of them keeps and what any of it does to anybody's profitability. The scheme's books stop at the accrual. The scheme's books were never going to say anything else.

The stopping point is not a gap in the teaching but the structure working. One test settles almost every borderline question in a single move: ask what document would answer the question. If the answer is a company's own income statement, the question is about a business rather than about a scheme, and no amount of reading the scheme's books will produce it. If the answer is the scheme's own statement of what it holds, owes, earned and spent, the question is about the scheme and the books are the right place to look.

Run the test on a few live examples. Ask what the scheme spent on running costs this year. The scheme's books. Ask what the manager earned from running the scheme. A company's income statement, so not here. Ask what the scheme's value per unit is today. The scheme's books. How profitable is the manager as a business? A company's income statement, so not here. The line holds every time, and it holds because the two sets of books were never one set.

Who writes these books up, and who tests them?

Different parties, on purpose. The operations team at the asset manager, headed by Sohail Merchant, prepares the scheme's books: it applies the written valuation policy, records what has been earned and incurred, and produces the day's figures. Kalyani Bhagat decides what the scheme holds and does not keep the record of it. The split between deciding and recording is ordinary and it is deliberate: the person who makes the decisions does not write up the account of them.

Above the preparer sit two more roles, both named here by role only. The scheme auditorAn auditor appointed in respect of a particular scheme, forming an opinion on that scheme's own accounts rather than on the accounts of the organisation managing it. is appointed for the scheme and forms an opinion on the scheme's accounts. The trustee company holds the scheme on behalf of the people holding units in it, and answers for whether the record was kept the way it ought to have been. Neither is an employee of the preparer, and that is the whole point.

The scheme having an auditor of its own is the structural expression of the boundary above rather than a formality bolted onto it: one record gets one opinion, and a second record needs a second opinion before the split between the two means anything in practice. Two sets of books written up in one office and signed off by one opinion would be two sets of books in name only.

One thing has to be said plainly. Books that are kept properly make a record complete and checkable; they do not make the value struck from that record correct, and they do not stand between a holder and a loss. A complete, checked and audited record of a holding that fell in value is a complete, checked and audited record of a fall. Bookkeeping answers whether everything that happened was recorded, in the right amount, on the right day, and against the right line. Bookkeeping does not answer whether the values used were the right ones to use, and it never answers whether holding the thing was a good idea.

What comes out of these records that anyone outside ever sees?

Three things, and all three are worth carrying away. A value per unit, published for every working day. A list of what the scheme holds, published at intervals. And financial statements for the scheme, prepared periodically. A holder sees the first constantly, the second occasionally and the third rarely, and most readers never connect them.

One record produces all three, so a mistake inside that record cannot surface in one of them and stay buried in the other two. The mistake surfaces in all three together, and catching it in any one of them is catching it in every one. A wrongly carried holding does not merely make the holdings list wrong. The same error makes the day's net figure wrong, so the value per unit is wrong, so the statements built on those figures are wrong. The three were not built separately, so no version of the story has them disagreeing on that ground.

Turn that around and it becomes useful rather than alarming. Two of the three agreeing is weak evidence rather than strong. Both came from the same place, so consistency was never in doubt. The strong evidence comes from a record that was kept somewhere else by somebody else and still agrees. Separating a genuine check from a restatement is worth more than any single figure above.

How does one day of running cost actually work out in rupees?

Work it through in the order the operations team would, on the Girnar Large Cap Equity Fund, holding net assets constant at Rs 4,200 crore, units in issue at exactly 120.00 crore and the year at 365 days. The three assumptions are stated because the arithmetic below is only as good as they are, and this record carries one asset figure and no series behind it.

Three routes, one answer, and the temptation now is to call that a triple check and move on. Do not. Route two is route one with the divisions reordered and nothing else. The order in which the division by the unit count is taken changes no result, so net assets times the ratio, over the year, over the units is arithmetically the same expression as net assets over the units, times the ratio, over the year. Route three is the same expression again with the folio count cancelling: the year's cost over the folio count, divided by the net assets over the same folio count, returns the ratio it started from.

The two impressions are both true and readers usually hold only one, so place the day's figure against the scale of the published number. Shift the published figure by a single paise and, across the scheme's 120.00 crore units, that is Rs 1.20 crore. A day's running cost is about Rs 0.18986 crore. Carried exactly rather than rounded, moving the published figure by a single paise costs 4 parts in 231 of a whole year of everything this scheme spends running itself. On a 365 day year that works out at about 6.32 days of it. A sixth of a paise a unit is nothing in the hand and upwards of six days of the whole cost base in aggregate, and both readings describe one figure.

One absence has to be stated rather than filled. Nowhere in this record is the running cost divided among the parties who share it, and no service provider carries a fee schedule here, so the account stops at the amount the scheme sets down. Where that amount ends up afterwards is a separate subject.

Which check on this record could actually have failed?

Only one, and it is not any of the three routes above. The books of the Girnar Large Cap Equity Fund carry unit capital as one total: 120.00 crore units in issue at the close of the day. The unit registerThe record listing every holding account in a scheme and the units standing in each, maintained separately from the scheme's own books. maintained by the registrar and transfer agent carries the same quantity built from the other end, as a balance for each of 3,80,000 folios added up. Two records. Two parties. Two different sets of source documents. One quantity that has to come out the same.

The tie between those two records is the only figure above that could have come out wrong, and that is exactly what makes it worth running. On this record the difference between the two totals is exactly zero. Nothing is being asserted by that zero except that on this invented record the two sides agreed. A zero here is a result, not a formality, and it is a different kind of statement from the three routes, all of which were always going to agree because they were always the same equation.

There is a second residue in this guide, and it goes the other way: it does not cancel and it is not supposed to. Suppose somebody hands the Girnar Large Cap Equity Fund Rs 1,00,000/- on a day when one unit is worth Rs 35.00. The division gives 20,000 parts in 7, or 2,857.142857 and onwards without ending. The scheme's illustrative treatment carries units to three decimals and rounds up, so 2,857.143 units are allotted. The rounding up added exactly 1 part in 7,000 of a unit, and 1 part in 7,000 of a unit at Rs 35.00 is exactly Rs 0.005.

Look at what that half paise is. Money received was Rs 1,00,000.00 and value handed out was Rs 1,00,000.005, so the signed residue is minus Rs 0.005, sitting exactly halfway between two paise. Because rounding up is one sided, that residue is always negative and never positive, so it cannot average away across many purchases the way a residue from rounding to nearest would. It is small, it is systematic, and the books have to put it somewhere rather than pretend it is not there. The market regulator sets the applicable convention on any given day, and the three decimals and the rounding up above are an illustration used to make the residue visible.

Who sets the recording and reporting duties, and why does none of them appear above?

From two places. Which records a scheme has to keep, the basis it keeps them on, what has to appear in its financial statements, the way the scheme's auditor comes to be appointed, and the deadlines by which anything has to be reported, are all matters for the market regulator. The accounting standards themselves, the technical rules about how a transaction is recognised and measured, sit with the accounting institute.

Requirements like these get revised. A printed requirement does not simply go out of date the day it moves: it goes from right to wrong while still saying it in exactly the same confident voice. The current text of any requirement, standard number, prescribed format, reporting interval, audit condition, threshold and effective date is the regulator's to state.

The mechanism set out above outlasts every one of those revisions: four operations run every working day, cost recorded on the day it is run up, unit capital shifting daily, and two records carrying two opinions. The shape does not depend on any current detail, and every current detail depends on what the regulator says today.

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