EPF: How It Accumulates, What the Employer Adds and When It Vests

A provident fund accumulates from two contributions, one deducted from pay and one added by the employer, with interest credited on the balance. The employer's side is part of what the job pays even though it never appears in take-home. Vesting decides when the accumulated balance becomes the employee's to take, and the rules for reaching it are set by the scheme.

Underneath that sits one picture, and the rest of this guide is a set of consequences of it. The whole arrangement is one account with two taps running into it and a credit added on what has gathered. Almost every difficulty people have with a provident fund comes from never having seen the second tap. Money the employer puts in is money the employee earned by working there, so reading a payslip as though take-home were the whole of the pay understates the pay of that job by exactly whatever the second tap carries.

The household that carries the arithmetic here is the same one that has carried it throughout. Meghna Bhosale is 36. She has worked for eleven years at Sahyadri Freight Services Private Limited, an invented freight business. Her husband Ashok Bhosale runs a tailoring counter with no employer, no payslip and no scheme attached to it. Their daughter Ira is at school. Every figure below is theirs, invented, and not one of them is a rate, a rule or a threshold.

What is a provident fund, mechanically?

How EPF Works: one account, two taps and a credit

Set the abbreviation aside for a moment and picture a steel drum in the corner of a room. Two pipes run into the top of it. The first pipe is fed by the person whose drum it is: every month, before her pay reaches her, a fixed amount is diverted into the drum. The second pipe is fed by the establishment she works for: every month, separately, the employer pours in an amount of its own. Nobody may take water out of the drum on an ordinary afternoon. And periodically somebody comes along and adds a measured amount on top of whatever has gathered, worked out on the level in the drum rather than on what went in that year.

The drum is the entire mechanism. A provident fundA workplace account that accumulates contributions from the employee and the employer, with interest credited on the balance under the scheme's own rules. is an account into which two parties contribute on a schedule, on which interest is credited, and out of which money comes only in the circumstances the scheme sets. The Employees' Provident Fund (EPF) is that same arrangement as it is run in India for employees of covered establishments.

Three properties follow immediately, and between them they explain most of what a first-time reader finds strange about it.

The first is that it is not voluntary the way a monthly savings habit is voluntary. The deduction happens before the money arrives. Because it does not land in the account the electricity bill is paid from, it never competes with the electricity bill. Nobody has to remember it on the twenty eighth of the month and nobody has to resist spending it. The most powerful property of the whole arrangement is that the money is gone before it is ever seen. Balances build up in accounts their holders barely think about for exactly that reason.

The second is that it is slow on purpose. A provident fund is built to be hard to reach, and that difficulty is a feature rather than a defect. The difficulty is the reason a balance can survive eleven years of a household that has had cash flow trouble, a card balance and a two-wheeler loan across the same period.

The third is that the balance is a running total rather than a row of separate deposits. Interest is credited on what has gathered, not on this year's contributions alone, and that single property makes the last years of a long account look nothing like the first ones. Credited interest on a running total is taken up below.

Where do the contributions come from?

Two places, and they behave completely differently in a household's experience of them.

The first is the employee contributionThe amount deducted from pay and put into the fund, which the payslip shows as a line between gross pay and take-home.. On Meghna's invented payslip, gross pay of Rs 46,000/- has three amounts taken off before anything moves: provident fund Rs 3,120/-, professional tax Rs 200/- and tax deducted Rs 2,880/-, leaving take-home pay of Rs 39,800/-. That first line is tap one. A reader who has met that payslip before knows the shape of it, and nothing about it changes here.

Tap one is easiest to understand from what it is not. It is not a charge. It is not a fee for having a job. It is her money, moved sideways into an account with her name attached to it, and the only thing the deduction does is decide which of her accounts it lands in. On the day of the deduction her wealth does not fall by Rs 3,120/-. Her reachable money falls by Rs 3,120/- and her balance in a slow account rises by the same amount. Wealth and reachable money are two different things, and only the second of them fell.

Employer Contribution: the part take-home never shows

The second is the employer contributionThe amount the employer puts into the fund alongside the employee's own. Take-home never shows it, because it never passes through the employee's bank account., and most people have never seen the figure, so it is worth dwelling on. On Meghna's invented fund statement, Sahyadri Freight Services Private Limited adds Rs 3,120/- a month of its own, shown on a separate line from hers. The two happen to be equal on this statement. The two sides are equal because that is what this invented employer's statement shows, and for no other reason: the rates at which either side is worked out, and any ceiling applied to them, are set by scheme rules.

One question decides how to think about the employer's side. Whose money is that Rs 3,120/-?

It is hers. She cannot spend it this evening, but it is hers in the sense that matters when working out what a job pays. The employer does not put it in as a kindness, out of goodwill, or from a separate charitable budget. The money goes in because she turned up and did the work, exactly as her pay does. The contribution is a term of the job in the same way the pay is a term of the job. The employer's contribution is not a benefit the employer confers; it is a part of what the job pays, paid into a different account.

The second tap gives a household two figures instead of one, and both are worth writing down.

Take-home is Rs 39,800/- a month. Take-home is the money that arrives, the money the rent comes out of, the money the buffer is measured in. Every sum this household has done across the cash flow, the budget and the emergency buffer used that figure, and every one of them was right to.

The job pays her Rs 42,920/- a month, being the Rs 39,800/- that arrives plus the Rs 3,120/- that goes into the fund without her seeing it. Across a year, that second part is Rs 37,440/-. Thirty seven thousand four hundred and forty rupees a year has been earned by this household, in every year of the eleven, without ever once appearing in anything the household counted.

There is a third figure too, and it is the employer's view rather than hers. The job costs Sahyadri Freight Services Private Limited the gross pay of Rs 46,000/- plus its own Rs 3,120/- contribution, or Rs 49,120/- a month. Rs 49,120/- a month is sometimes called cost to companyWhat a job costs an employer in total, including amounts that never reach the employee's bank account. It is a larger figure than gross pay and a much larger one than take-home., and a household that has only ever seen a payslip has good reason to find it confusing: the four figures sit at four different levels for four different reasons.

What happens when two jobs are compared on take-home?

Now the practical consequence, and it is the one that costs households real money without anybody noticing.

Suppose an offer reaches Meghna from another freight business, also invented. The offer would pay her take-home of Rs 41,000/- a month, or Rs 1,200/- more than she takes home now. Because those are the two figures anybody can see, that is the comparison the household will make at the kitchen table: Rs 41,000/- against Rs 39,800/-. The offer is better by Rs 1,200/- a month, or Rs 14,400/- a year, and that is real money in a household spending Rs 42,770/- a month.

Then the fund statement arrives from the new employer, or somebody thinks to ask for the figure. The new establishment's provident fund contribution on its own side is Rs 1,800/- a month rather than Rs 3,120/-. Employers can differ on this, for reasons set by scheme rules, by the structure of the pay, and by choices the establishment has made.

Redo the comparison with both taps counted. The offer pays Rs 41,000/- plus Rs 1,800/-, or Rs 42,800/-. Her present job pays Rs 39,800/- plus Rs 3,120/-, or Rs 42,920/-. The offer that looked Rs 1,200/- a month better is Rs 120/- a month worse, and the entire flip happened in a figure that appears on neither payslip.

The claim is a narrow one. Whether she should stay or go is a separate matter. A job is not a number, and the arithmetic of two taps says nothing about the work, the hours, the travel or the people. The arithmetic says only one thing. The two figures she was about to subtract from each other were measuring different things, so the comparison was not a comparison at all.

How does interest work on the balance, and why does timing matter?

Credited interest is what separates a provident fund from a tin box on top of a cupboard, and being precise about it repays the trouble.

Credited interestInterest added to a fund balance under the scheme's own rules, worked out on what has gathered rather than on what was paid in during the year. is worked out on the balance. Not on the year's contributions, not on the employer's half, not on the original sum: on whatever has gathered in the account. The scheme sets the rate, sets how often the working is done, and sets the method for handling money that arrived part way through a period. All three are set by rules that change and are published by the authority that runs the scheme.

The shape of the credit can be described without touching a rate at all, and the shape is what people get wrong.

Take a household example with no finance in it. Two neighbours each put a bucket under a tap that drips at the same speed. The first neighbour puts her bucket out at dawn; the second puts an identical bucket out at dusk. At midnight the first bucket holds far more water, and not because her bucket is better or her tap faster. The first bucket was simply under the drip for longer. Interest is not a reward for the size of a contribution; it is a payment for the length of time an amount sat in the balance being counted.

Now put arithmetic on it without inventing a rate. Meghna's contributions arrive in twelve instalments across a year, one a month. The first instalment is in the balance for twelve months before the year ends. The second month's instalment is in for eleven. The twelfth is in for one. Add those up: twelve plus eleven and so on down to one is seventy eight. Call each one a month of an amount sitting in the balance, and a year of monthly contributions buys seventy eight of them.

If exactly the same Rs 74,880/- had arrived in a single lump in the first month, every twelfth of it would have sat for the whole twelve months, and the total would have been twelve times twelve, or one hundred and forty four. The same money, paid in across a year rather than at the start of it, spends seventy eight month-units in the balance rather than one hundred and forty four, a little over half.

None of that is an argument for paying in early. The month-unit is the mechanism behind two things met elsewhere. The early contributions were both smaller and, at the time, sitting in a small balance, and a balance built over eleven years is therefore weighted heavily towards its last years. And it is why delay on a long goal costs so much more than the contributions actually skipped: it is not the money that is missed, it is the month-units.

What does eleven years of it actually look like?

Now the worked instance, and it is the household's own position at 31 March of year two.

Meghna has eleven years of serviceTime worked, usually with one employer or under one scheme. Several scheme rules are measured against it rather than against age. at Sahyadri Freight Services Private Limited. Her fund statement at that date shows a balance of Rs 4,12,000/-. The structure of that balance matters more than the total, so take the three parts one at a time.

The statement splits that balance into three parts, as a fund statement does.

Her own contributions across the eleven years come to Rs 1,68,000/-. On this statement the two sides have matched each other in every year, so the employer's contributions come to the same Rs 1,68,000/-. Contributions to date are therefore Rs 3,36,000/-. Interest credited across the eleven years comes to Rs 76,000/-. Add the two: Rs 3,36,000/- plus Rs 76,000/- is Rs 4,12,000/-, the balance on the statement.

Stop on the first surprise before going further. Her contributions today run at Rs 3,120/- a month, or Rs 37,440/- a year, and eleven years of that would be Rs 4,11,840/- from her side alone. Her side has actually contributed Rs 1,68,000/-. The gap is not an error: the current contribution is worked out on the largest pay she has ever had, so it is the largest she has ever made, and for most of the eleven years both were far smaller. When she joined, her contribution was under a sixth of what it is now.

The size of that gap is the second thing a long balance teaches, and it is the one that surprises people who have never opened a statement. The eleven years are not eleven equal blocks. The three most recent years contributed Rs 1,73,280/- of the Rs 3,36,000/- total, more than half of everything contributed across the whole eleven. The first three years contributed Rs 40,320/- between them, under an eighth.

Two columns of that table are worth reading against each other. The contributed column rises because her pay rose, in steps rather than smoothly, with the sharpest steps in the last two years when she moved onto the despatch desk. The credited column rises for a different reason: it is worked on a balance that keeps getting larger, so it climbs even in a year when nothing about her contribution changed. Contributions grow because a career grows; credit grows because a balance grows, and only the second of those keeps working when a career stops.

The credited column also moves for a third reason. The rate at which the credit is worked is set by the scheme and is not the same in every year. The last year on this invented statement therefore carries more than twice the credit of the year before it rather than a little more. A reader who tries to work a rate backwards out of two rows of a statement will get a different answer from every pair of rows, and the published position sits with the scheme rather than with the arithmetic.

One row is worth marking for later. A statement always starts from where the last one stopped. The balance at the end of year ten is Rs 3,08,900/-, and that is the figure the statement for year two opens with. Reading such a statement line by line is covered under reading a fund statement.

Why was this balance missing from the household's own sheet?

Here is a question a careful reader will already have asked. The household's balance sheet listed everything it held: two bank accounts at Rs 41,887/-, a recurring deposit at Rs 64,000/-, a public provident fund at Rs 84,000/-, gold at the household's own estimate of Rs 1,40,000/- and a two-wheeler at Rs 38,000/-, or Rs 3,67,887/- in all, against Rs 71,594/- owed, leaving a net worth of Rs 2,96,293/-. There was no line on it for Rs 4,12,000/-.

The sheet was not wrong. The sheet was answering a particular question. It was built to show what the household could reach and what it could count: what could be turned into money, how long the buffer lasted, what would happen if income stopped. A provident fund balance cannot be reached on the afternoon it is needed, so it answers none of those questions. Leaving it off a resilience sheet is a choice, and for a sheet about resilience it was the right choice.

There is an apparent inconsistency here, and it is worth naming rather than hiding. The public provident fund at Rs 84,000/- is on that sheet, and it is not quickly reachable either. The public provident fund went on because the household had the passbook in front of it with a figure printed on it, and a sheet lists what it has documents for. The provident fund balance had never been asked for. Nobody in the household had opened the statement, and a figure nobody has looked up cannot be listed. A balance sheet shows what its builder went looking for, so the boundary of a sheet is a decision rather than a fact, and it is worth writing that decision down beside the total.

Now put the fund balance on the sheet. Holdings become Rs 3,67,887/- plus Rs 4,12,000/-, or Rs 7,79,887/-, and net worth becomes Rs 7,79,887/- less the Rs 71,594/- owed, or Rs 7,08,293/-. The household's net worth on paper more than doubles, from Rs 2,96,293/- to Rs 7,08,293/-, and not one rupee of that changes what it can do next month. The buffer still covers 0.73 months. Nothing has been gained and nothing has been discovered except a figure that was always there.

Both sheets are true and they answer different questions. The first says what this household can reach. The second says what this household has accumulated. A household that has been feeling behind on everything is often reading only the first, and the second is frequently a much less bleak document.

Take three settings of the panel in turn. At eleven years, the household's own position, the balance is Rs 4,12,000/- and credit is Rs 76,000/- of it, being 18.4 per cent. At thirty years with no rate assumed at all, the balance is Rs 18,34,720/-, of which Rs 17,58,720/- is contribution and the credit band is still the original Rs 76,000/-, being 4.1 per cent. At thirty years on an assumed 7 per cent, the balance is Rs 45,11,093/-, of which the same Rs 17,58,720/- is contribution and Rs 27,52,373/- is credit, being 61.0 per cent. The contribution figure is identical in those last two settings, so every rupee of the difference between Rs 18,34,720/- and Rs 45,11,093/- is the assumption rather than the scheme.

The gap between those two settings is worth saying in plain words. Any long projection of a provident fund balance has two ingredients. The first is money that will actually be paid in, knowable and small. The second is credit worked on an assumed rate, unknowable and eventually much larger. Anybody presenting a single confident figure for a balance decades away has chosen the second ingredient on the reader's behalf. The honest form of the sentence names the assumption first.

When does the balance become the employee's to take?

Vesting: the word that answers it

Somebody reading a fund statement for the first time asks a reasonable question about the employer's column. If that money came from the employer, can the employer take it back? Suppose she leaves after two years, or is asked to leave, or the establishment closes.

The word that answers all of those is vestingThe point at which an accumulated balance becomes the employee's to take, decided by the rules of the arrangement rather than by who paid the money in.. Vesting is the point at which an accumulated balance becomes the employee's to take. Before that point an amount may be sitting in an account with her name on it and still not be hers to walk away with; after it, it is hers, and the party that contributed it has no further say.

Two things about that definition do more work than they look like they do.

The first is that vesting is a property of the arrangement, not of the money. The rules attach to where each amount came from rather than to where it now sits, so two amounts sitting in the same account, contributed on the same day, can be treated differently. The question is never what the balance looks like; it is what the rules of the scheme say about each part of it.

The second is the one that catches people. Vesting settles whose money it is, and vesting is set by the rules of the scheme rather than by the source of the rupees, so who paid it in settles nothing. The instinct runs the other way: the employer put it in, so presumably the employer keeps some claim on it. Sometimes an arrangement does work like that and sometimes it emphatically does not, and the only way to know is to read the rules of the particular scheme.

A household example makes the shape clear without any scheme at all. A chit collector on a lane takes a monthly amount from each member of a group, and what each member gets, when, and what happens to somebody who stops paying after four months, is set by the arrangement the group agreed at the start. The rupee notes do not carry the answer; the arrangement does.

Vesting conditions, qualifying lengths of service and the exceptions to them are set by the rules of the scheme, they differ between the parts of what looks like one account, and they change. The Employees' Provident Fund Organisation publishes them at epfindia.gov.in, and that is where they are read. One question survives every rule change, and no answer to it does: what does vesting say about this part of the balance? Any answer will be different in five years.

What happens to the balance when somebody changes jobs?

People change jobs. Over a working life of thirty five years, several times is ordinary rather than unusual, and the arrangement has to survive that or it would be useless.

The mechanism is a transferMoving an existing fund balance so it continues under a new employer rather than being left behind or started again.. In broad shape, a provident fund balance belongs to the person rather than to the job, so on changing employers the existing balance can generally be carried across and continue accumulating, with the new employer's contributions joining the same running total. The conditions for doing that, what has to be done and by whom, and what happens if nothing is done, are set by scheme rules and are read at the Employees' Provident Fund Organisation at epfindia.gov.in. Why it matters so much follows directly from the section on timing.

If a balance is carried across, the whole accumulated amount keeps sitting in the balance being counted, and every future credit is worked on the larger figure. If a balance is instead left behind and a fresh account is started, the person now has two balances, and the arithmetic of the second one begins from the bottom of the ladder rather than from the top. A balance not carried across does not vanish, so the cost is not the balance itself. The cost is that one accumulated total becomes several, and every rule that rewards a large balance rewards several small ones less.

There is also a plain administrative cost that has nothing to do with arithmetic. Several small balances at several former employers, each with its own paperwork and its own record of service, are harder to find, harder to claim and easier to forget than one. People do lose track of them. A household that has moved cities twice and changed employers three times may hold money it does not remember having.

The teaching point is a question rather than a procedure. On changing jobs, the useful thing to ask is what happens to the existing balance and what has to be done to keep it in one place. The procedure on changing jobs is set by scheme rules and changes; the question survives every change, and it is worth asking on the day rather than eleven years later.

What does a household with no employer have instead?

Everything above has assumed a payslip. A great many working people in this country do not have one, and the household described here contains one adult who does and one who does not.

Ashok Bhosale runs a tailoring counter. He works six days a week, sometimes seven in the wedding months, and the counter supports this household as much as the freight job does. He has no employer. There is therefore no payslip, no deduction before the money arrives, no second contributor, no fund statement arriving once a year, and no balance quietly accumulating in an account nobody thinks about.

Put the asymmetry as plainly as it deserves. Meghna's arrangement has two contributors and requires nothing of her each month. Ashok's has one contributor, and that one contributor has to decide, every single month, out of money that has already arrived and is already being asked for by the rent, the fabric supplier, the school and the card. The gap between the two positions is not effort and not intention; it is structure, and structure is doing a job in one case that willpower has to do in the other.

Three consequences follow, and each is worth naming because each is often mistaken for a personal failing.

The first is that nothing happens by default. A provident fund accumulates whether or not anybody thinks about it in a difficult month. A tailoring counter accumulates nothing at all in a month when nobody thinks about it, and difficult months are exactly the months nobody thinks about it.

The second is that there is no second contributor. Whatever Ashok sets aside is the whole of what goes in, with no line on any statement adding the same amount again, and for a household comparing its two adults that halving is invisible and easily read as one person being better with money than the other.

The third is that there is no protection from reach. Being hard to get at is precisely what keeps a balance alive through a bad quarter. Money in an ordinary account is reachable, a strength when a machine breaks and a weakness in every other month.

He has access to the same schemes anybody can open on their own, some of which are covered separately, and nothing that arrives automatically. The difference is not a small one. None of it is a verdict. A person running a counter that supports a household is not behind on anything; the arrangement around them is simply doing less of the work, and knowing that is more useful than feeling bad about it.

How do people actually use this, outside a lesson?

Four users, and each reads the same statement for something different.

A lender reads it as evidence. A bank assessing a loan application is trying to work out whether income is likely to continue, and a long unbroken run of provident fund contributions is a record of continuous employment that is hard to manufacture. The record says less about the size of the pay than the payslip does and more about the steadiness of it, and steadiness is often what a lender actually wants to know.

Somebody reading an establishment's accounts uses it as a cost. The employer's contribution is real money leaving the business every month, so a business with a hundred employees on the books is spending on this line whether or not anybody in the business thinks of it as pay. When employment costs are compared between two establishments, the contribution line is part of it.

A household uses it in two quite different ways, and both matter more than they sound. The first comes at the moment of comparing jobs, and it is the failure described above: two figures, each hiding a different amount, and one question that fixes it, namely what each employer puts in. The second is quieter and comes up more often. A household that feels it has nothing set aside for a goal decades away frequently has more than it thinks, sitting in a statement nobody has opened, and finding out costs nothing.

The fourth user is the person negotiating their own pay, and this is where knowing the arithmetic changes a conversation. A discussion about pay conducted entirely in take-home figures leaves out a term of the job that is worth Rs 37,440/- a year in this invented case. The useful question in that room is not what will I take home, but what does the job pay in total, and those two questions have different answers on every payslip in the country.

What about a provident fund cannot be settled in advance, and why?

No rate and no condition of any kind survives being written down: not the rate at which either side is worked out, not the rate at which interest is credited, not any ceiling on the wage those calculations run on, and no vesting period, qualifying service, withdrawal circumstance, transfer deadline or tax treatment. Every one of them is set by scheme rules that have changed before and will change again, and conditions are exactly what people act on. The shape of the mechanism is the part that lasts, so everything carrying a number or a date is named and pointed at rather than reproduced.

Reading a statement line by line is covered separately. Whether the scheme is a good one is not a question answered here about any arrangement, and it would be a strange one to ask about something that is not optional for a covered employee anyway. Whether anybody should contribute more than the arrangement requires, or open anything, or move anything, or take anything out, is likewise outside the subject.

And nobody's position is treated here as avoidable. A reader who is fifty with no scheme, no statement and nothing set aside has read an account of how a thing works, not an account of what they should have done. The two accounts are different, and only the first of them is any use.

The Bhosale household, Meghna Bhosale, Ashok Bhosale, Ira Bhosale and Sahyadri Freight Services Private Limited are invented, as is the second employer whose offer appears above.
Educational material. Not advice on any investment, tax, budget or market position.