The split below is the 50/30/20 rule, made popular by Elizabeth Warren and Amelia Warren Tyagi in their 2005 book All Your Worth. It is a first sketch, not a household plan. A household count sits under the income input, so the same needs band can also be read per person, and a set of shortcuts fills that count in for a few common household shapes.
The Split #
Take-home is the amount that lands in the bank each month, after tax and provident fund.
- Needs 50%
- Wants 30%
- Savings 20%
The needs band spread across 1 person works out to each a month.
This is the needs band divided by the household count, nothing more. It is arithmetic, not a judgement on whether that amount is enough. For scale, the NSO put average urban spending at ₹6,996 per person a month in 2023-24, and that figure covers everything a person spends, not only the needs band.
The needs band is the part of the income an emergency fund has to replace. Three, six and nine months of it work out like this.
- 3 months
- 6 months
- 9 months
The savings band above is where the monthly contribution to this comes from, so the fund gets built over months rather than set aside in one go. This page does not pick which row applies. The multiple depends on how steady the income is and how many people depend on it, and three to nine months is a rule of thumb rather than a rule. The emergency fund step of the Ladder works through the sizing, where the money waits and why equity is never the buffer.
Where that money waits is a category question rather than a scheme question. These are the two categories a buffer normally sits in, counted from the scheme data on this site rather than typed in by hand.
All 31 overnight funds tracked here sit in the lowest cell of SEBI's potential risk class matrix, which is the lowest credit risk bucket paired with the shortest duration bucket. Of the 30 liquid funds, 8 sit in that same cell and the other 22 sit outside it. Both categories carry the same everyday shorthand, so the shorthand is not what settles the risk class. That shows in the liquid counts just given. Inside that set the class has to be read scheme by scheme rather than read off the category name. Median regular plan expense ratio runs 0.17 percent across the 26 overnight funds that disclose one against 0.26 percent across the 23 liquid funds that disclose one. The schemes left out of those two medians publish no regular plan total. The smallest first investment recorded across the two categories is ₹100 and the most common is ₹5,000, which is what decides whether a small monthly buffer contribution can go in at all. Scheme documents read up to 15 August 2026. Expense ratios as disclosed on dates up to 16 August 2026. What the letter and the numeral mean is set out in the debt fund risk class guide. Nothing here names a scheme or says which of the two fits a particular household.
If the whole savings band went into a SIP, that would be a month. A step-up raises the contribution once a year so it tracks a rising income instead of staying fixed. By the fifth year the same SIP would be putting in this much a month.
- 5% a year
- 10% a year
- 15% a year
Read these as contribution arithmetic and nothing more. No return is assumed anywhere on this page, so none of these figures is a projection of what an investment becomes. That distinction is the honest part. A stepped-up SIP finishes ahead of a flat one because more money went in, not because the money earned more, and the rate of return on both is the same. The step-up is a savings discipline rather than a performance feature. This page does not pick a rate either. The one that fits is roughly whatever the income actually rises by, which is why FinSet's own treatment anchors the increase to the appraisal month, where the higher amount comes out of money the household never adjusted to spending. The goal SIPs step of the Ladder works through sizing a SIP against a real goal, matching the horizon to the asset and the annual review.
Whatever that monthly figure comes to, it lands in a scheme that sells the same portfolio twice over. The regular plan and the direct plan hold identical securities and differ only in what they charge. These are the two shapes a long-horizon SIP usually takes, counted from the scheme data on this site rather than typed in by hand.
Across 492 actively managed equity funds the median regular plan charges 2.23 percent a year against 1.06 percent on the direct plan, and the typical per-scheme gap is 1.17 percentage points. Across 305 open-ended index funds the same comparison runs 0.89 percent against 0.34 percent, and the typical per-scheme gap is 0.45 percentage points. Both figures are the median of the per-scheme gaps rather than the difference between two medians. Exchange traded funds are left out of the index count because they are bought on an exchange and priced differently. Expense ratios as disclosed on dates up to 16 August 2026, which is the newest disclosure date in the set rather than a date every scheme shares. The expense ratios themselves are not a FinSet measurement. Each fund house files its own, plan by plan, with AMFI, which publishes the plan-level figures, and this site reads that disclosure and computes the gap from it. What is first-party here is the arithmetic on those figures rather than the figures. The direct versus regular guide works through the same comparison category by category.
That gap is not a charge the fund house keeps. It is the distribution commission, which the scheme pays out to whoever holds the account. On a FinSet account that is FinSet, AMFI-registered under ARN 180462, so the number above is a price this site is paid rather than a leak it is warning about. Read plainly, the regular plan buys a person who is accountable for the plan. Somebody who sizes the SIP against a real goal, who says no to the fund that happens to be selling well this quarter and who is still answering the phone in the year the market falls. What it costs is those points a year on the money invested, every year it stays invested. An investor already doing all of that unaided is paying for work they are doing themselves. An investor who is not doing it is rarely choosing between two prices. The choice there is between a plan and no plan. A cheaper plan has never by itself made anybody hold on through a bad year. This page does not pick a side of that for anyone. It states the cost, names who receives it and leaves the judgement where it belongs.
Some of what the needs band already counts as monthly does not arrive monthly. Insurance premiums, school fees and the festival months land as bills once or twice a year. Setting a year of them aside a month at a time costs this much, depending on what they come to.
- Half a month of needs
- One month of needs
- Two months of needs
This money comes out of the needs band rather than on top of it. The premiums and fees are already counted inside needs, so setting them aside monthly moves what is already there instead of adding a fourth claim on the income. That is the difference from the emergency fund above, which the savings band pays for. This is also not the emergency fund in what it is for. That one is for the income stopping, which may never happen. This one is for bills that are certain and only irregular, which is why it is set aside rather than insured against. This page does not pick a row, because it cannot know what a household's own annual bills come to, and nothing here is a claim about what a premium or a festival typically costs. The multiple is a way of scaling the arithmetic to an income, not a measurement. The honest use of it is to add up last year's actual once-a-year bills and read the row nearest that.
The needs band already counts the contractual minimum on any loan, so a repayment is not a fourth claim on the income. It is a claim on the first band. Entering what those repayments come to each month shows how much of that band has gone before the rent and the groceries are paid.
Count the minimum due on every loan running, including home, car, personal and the fixed instalment on a card. Anything paid above the minimum is a choice and files under savings, so leave that part out. Leave the box at zero if there are no loans.
- Share of take-home
- Share of the needs band
- Needs band after repayments
Lenders read the first of those figures as a ratio of their own, the fixed obligation to income ratio, and each lender sets for itself the limit it will lend against, whether the borrowing is a home loan, a car loan, a personal loan or a card. The housing loan circular, the only circular this page cites that would carry a regulatory limit if such a limit existed, is named in the sources below. It sets no such limit. This page sets no limit and reads no verdict off these numbers. It divides what was entered by the income and by the needs band, and stops there. The third row is what the needs band has left for rent, utilities, groceries, transport, school fees and insurance premiums once the repayments are out, which is the figure the other rows on this page are quietly assuming is intact. Where it looks thin, that is the arithmetic showing which part of the split is doing the work, not this page saying anything about the borrowing itself. The glossary entry on FOIR sets out how lenders use the ratio and why it is a convention rather than a rule anyone is bound by, and the cited explainer on FOIR shows what the housing finance circular does and does not contain.
Rent is the largest single bill inside the needs band for most salaried renters, and it is the one that moves most from one city to the next. This calculator asks for it rather than assuming it. Entering what rent comes to each month shows how much of the needs band has gone before anything else is paid.
Count the rent itself plus any fixed monthly maintenance or society charge that comes with the tenancy. Leave the box at zero for a household that owns where it lives, because a home loan repayment belongs in the box above instead.
- Share of take-home
- Share of the needs band
- Needs band after rent and repayments
The third row nets both boxes off the needs band at once, so it is what is left for utilities, groceries, transport, school fees and insurance premiums once rent and the loan repayments are out. The row above it in the previous block nets the repayments only. This page sets no ceiling on any of these figures and reads no verdict off them. It divides what was entered by the income and by the needs band, and stops there. Why this page asks instead of assuming. The obvious alternative is a city multiplier that scales the needs band up for a metro and down for a smaller town. That is not applied here, and the reason is worth stating rather than hiding. A multiplier of that shape is a level ratio between two cities at one date, while the published Indian rent series measure growth off a base year, so no dated source exists that could produce the four numbers such a feature would need. Beyond that, two households in the same city paying ₹18,000 and ₹65,000 do not have the same needs band, and a city average would be wrong for both. What was actually paid is the better number, and the reader already knows it.
Why these numbers? Open any line to see the arithmetic and the dated source behind it.
Needs,
The arithmetic. Monthly take-home multiplied by 50 percent.
Where the 50 comes from. The 50/30/20 frame in All Your Worth (2005). It is a teaching heuristic written for American households and it carries no India calibration of its own.
What the Indian data says. In the NSO Household Consumption Expenditure Survey 2023-24, released 27 December 2024, the average urban person spent ₹6,996 a month, with food at 39.68 percent, conveyance at 8.46 percent, rent at 6.58 percent, education at 5.97 percent and medical care at 5.85 percent. Those five item groups alone come to about 66 percent of urban spending, and every one of them sits in the needs row of this page. So 50 percent is a target to aim at, not a description of what urban India already manages.
Wants,
The arithmetic. Monthly take-home multiplied by 30 percent.
Where the 30 comes from. The same 2005 frame. It is the discretionary band, so it is the residual once needs and savings are set.
What that means in practice. This is the band that absorbs the squeeze first. If the needs band runs over, as the NSO shares above suggest it often does, the overflow comes out of here before it comes out of savings.
Savings,
The arithmetic. Monthly take-home multiplied by 20 percent.
Where the 20 comes from. The same 2005 frame.
What the Indian data says. The Reserve Bank of India Annual Report 2025-26, released 29 May 2026, puts net household financial savings at 7.0 percent of gross national disposable income for 2024-25, up from 5.8 percent the year before. The 20 percent band is close to three times the national rate, which makes it a target rather than a norm.
Needs per person,
The arithmetic. The needs band divided by the household count, where the count is adults plus children plus parents or in-laws supported.
What it deliberately does not do. It does not weight a child differently from an adult, because no dated Indian source gives a per-dependent weighting this page could cite. A made-up weighting would look more precise and be less true.
The scale reference. NSO Household Consumption Expenditure Survey 2023-24, average urban spending of ₹6,996 per person a month. That covers everything a person spends, not only the needs band, so it sits below the figure above by construction.
Emergency fund,
The arithmetic. The needs band multiplied by three, six and nine months. Nothing else feeds it, so it moves only when the income box moves.
Why the needs band and not the whole income. A buffer has to replace what still has to be paid when income stops, not what gets spent when it does not. The needs row on this page is rent, utilities, groceries, transport, school fees, insurance premiums and the minimum on any loan, which is the same base the Ladder sizes against. Wants stop in a crisis, so they are left out.
Why three figures and not one. The multiple is the part a flat rule cannot set. FinSet's own treatment puts three months against a steady dual income, six against a single income or dependents and nine against variable or self-employed pay, with Indian hiring gaps stretching that end closer to twelve. Choosing one of those rows off an income box would be a judgement this page has no basis for, so it shows the arithmetic for each and links to the reasoning instead. What the buffer is actually spent on once pay stops is worked through separately in the guide to budgeting after a layoff, which runs the same needs row forward across a gap rather than sizing it.
Step-up SIP,
The arithmetic. The savings band raised by the step-up rate four times over, because the fifth year follows four annual increases. At a 10 percent step-up that is the starting amount multiplied by 1.10 to the fourth power, or about 1.46 times. Only the income box feeds it, so the household count does not move it.
Why there is no corpus figure here. Turning a SIP into a future value needs an assumed rate of return, and this page has no dated source for one. Rather than invent a rate, it shows only what goes in, which is the half of the calculation that is arithmetic rather than a forecast. What comes out depends on markets and is never assured.
Why three rates and not one. The right step-up is roughly the rate the income itself rises by, which the page cannot know. So 10 percent sits in the middle here as a round illustration rather than a measured figure, with a slower and a faster case on either side, and the choice is left to the reader. None of the three is a claim about what rises typically are.
Committed loan repayments,
The arithmetic. The repayments entered, divided by the take-home income for the first row and by the needs band for the second. The third row is the needs band minus the repayments. Only the income box and the repayment box feed it, so the household count does not move it.
Why the needs band is the second yardstick. The classification stated further down this page files the contractual minimum on a loan under needs, because it is a bill that arrives whether or not anyone decides anything. So a repayment does not sit beside the three bands. It sits inside the first of them, and the useful question is how much of that band is left once it has been paid.
What it deliberately does not do. It sets no ceiling and it does not say whether a share is high. Lenders judge this ratio, the fixed obligation to income ratio, against a limit each of them sets for itself. The one instrument this page cites that could carry a regulatory limit is the Reserve Bank of India's Master Circular on Housing Finance, RBI/2025-26/16 dated 1 April 2025, and it does not carry one. What it prescribes for a housing loan is the loan to value ratio and the risk weight, under the heading Quantum of Loan, which are limits on the size of the loan against the price of the property rather than on the share of an income that may go to repaying it. So there is no dated figure to put here, and putting an undated one in would be a number in the arithmetic with no source behind it, which is the same reason the city rent multiplier and the per-dependent weighting are absent.
Rent entered,
The arithmetic. The rent entered, divided by the take-home income for the first row and by the needs band for the second. The third row is the needs band minus the rent and minus the loan repayments together, because both come out of the same band. Only the income box, the rent box and the repayment box feed it, so the household count does not move it.
Why the reader is asked for it. Every other figure on this page is either arithmetic on the income or a share taken from a dated survey. Rent is neither, because the number that matters is what one household actually pays and no published figure carries that. The alternative would be a city multiplier, and the reason there is not one is that a multiplier is a level ratio between cities at a point in time while the Indian rent series that exist measure growth off a base year. A growth index cannot produce a level ratio however authoritative it is, so any such multiplier would be a number inside the arithmetic with nothing behind it.
What it deliberately does not do. It does not shift the 50, 30 and 20 shares. A rent that eats the whole needs band moves the third row to zero and says so plainly, rather than quietly recalibrating the split to make the rent look affordable. The bands stay where the rule puts them and the arithmetic shows what is left, which is the honest order of those two things.
Lumpy annual bills,
The arithmetic. The needs band multiplied by the row, then divided by twelve. One month of needs set aside across a year is the needs band divided by twelve every month. Only the income box feeds it, so the household count does not move it.
Why the needs band is the yardstick. Not because a year of lumpy bills is known to come to a month of needs. It is because the needs row on this page already counts insurance premiums and school fees as though they were paid every month, when most of them arrive once or twice a year. Setting them aside monthly is what makes that assumption true rather than a rounding error the budget absorbs in one painful month. Scaling the estimate to the needs band keeps it proportionate to the income the rest of the page is built on.
Why three rows and nothing chosen. The total depends entirely on which premiums fall due, how many children are in school and how the household keeps a festival, none of which this page has any way of knowing and none of which has a dated Indian source it could cite if it guessed. The rows are arithmetic on a multiple, not an estimate of anyone's bills. Adding up last year's actual once-a-year payments beats any of them.
What Each Band Covers #
| Band | Share | What sits here |
|---|---|---|
| Needs | 50% | Rent, utilities, groceries, transport, school fees, insurance premiums and the minimum on any loan. |
| Wants | 30% | Eating out, subscriptions, travel, gadgets and everything that lifts the standard of living without being essential. |
| Savings | 20% | Emergency fund, SIPs, retirement contributions and repaying debt faster than the minimum. |
Where the Bands Are Hard to Call #
The table above is the easy half. Most of the work in using a split is deciding which row an item belongs in, and a few things in a salaried Indian household do not sort themselves. This section states how this page files them, so the arithmetic above can be reproduced rather than guessed at. It is a stated convention, not a rule anyone has to follow.
Money that never reaches the account. Where an employer deducts provident fund and wherever tax is deducted at source, that money is gone before the salary is credited. This page asks for take-home pay for exactly that reason, so none of it sits in any of the three bands. The practical effect is that a provident fund balance is already being built outside the 20 percent, and a household that also counts it inside the savings band is reading its own saving rate high. Nothing here says what to do about that. It says only that the figure entered above does not contain it. How much tax is deducted before the credit depends on which regime the salary is taxed under, and the two rate tables that decide it are set out in the guide to the old and new regimes.
A loan repayment lands in two rows at once. The contractual minimum is a bill that arrives whether or not anyone decides anything, so it files under needs, as the table says. Anything paid above that minimum is a choice, and it reduces what is owed rather than what is consumed, so it files under savings. The same rupee changes rows depending on whether it was compulsory. That is why a household carrying a large loan can read as squeezed on needs and healthy on savings in the same month.
Insurance premiums. A premium that keeps protection in force is a recurring bill like rent or electricity, so it sits in needs. Where a policy mixes cover with an accumulation component, this page has no way to separate the two halves and does not attempt it. The whole premium goes to needs, which understates the savings band by whatever part of it was accumulation rather than cover. A savings figure produced from such a policy reads as a floor rather than a total.
School fees against an education corpus. A fee already payable for a child already enrolled is a need. Money set aside for a course that starts in ten years is savings. The purpose is the same and the row is different, because one is a dated bill and the other is a decision that can be revisited next month.
Groceries against eating out. This is the line most household budgets blur. Food bought to be cooked at home is a need, and food bought because cooking was inconvenient is a want, even though both leave the account as food. The NSO's 39.68 percent food share quoted below counts the two together, which is one more reason that survey figure and this needs band are not directly comparable.
Living in family-owned housing. Where no rent is paid, the needs band reads unusually low and the split looks easy to keep. The honest reading is that the space between the bands is a housing subsidy rather than efficiency, and it can end with a transfer or a marriage. This page cannot see any of it, so the comfortable-looking needs figure is doing less work than it appears to.
The once-a-year bills are handled above, not here. Insurance renewals, festival spending and school admission fees are lumpy rather than hard to classify, so they sit in the set-aside rows inside the calculator instead of in this list. Those rows are arithmetic on a multiple. This section is about which band an item belongs to in the first place. Where those bills are festival and wedding spending rather than renewals, the guide to sinking funds works through what a year of them tends to come to and how the set-aside is kept.
How The Bands Behave As Income Rises #
The three shares above are fixed. The rupee figures they produce are not. This section sets out what that difference does across the range the calculator accepts, which runs from ₹10,000 to ₹10,00,000 a month, because a proportional rule behaves differently at the two ends and the arithmetic on its own never says so.
Every band moves by the same multiple. A take-home of ₹30,000 produces a needs band of ₹15,000, a wants band of ₹9,000 and a savings band of ₹6,000. Ten times that income produces ten times each figure, ₹1,50,000, ₹90,000 and ₹60,000. Nothing in the rule bends. Set against the average urban spend of ₹6,996 a month per person quoted below, the first needs band covers about two such people and the second about twenty-one. The proportion held. The distance from the benchmark did not.
The rule assumes necessities cost a constant share of income. That assumption is the whole of its calibration, and nothing in the arithmetic tests it. Where a household's necessary spending sits closer to a fixed rupee amount than to a fixed fraction, the needs band and the actual need move apart as income rises, so the same 50 percent reads tight at one end of the range and loose at the other. This page carries no dated source for how that gap changes across income levels, so it states that the assumption is there rather than correcting for it.
Which row binds changes with the income. At the low end the needs row is the one that will not fit, and whatever it overruns comes out of the savings row, because the emergency fund and the SIP are the only two lines in the split that nobody sends a bill for. At the high end the needs row has room and the 30 percent wants band is the loose one, since a ceiling that large fills without anyone deciding to fill it. The rule reports the same three numbers in both cases and does not flag which of them is doing the work.
A raise leaves every ratio where it was. Where income and spending rise together, the split cannot see the change at all. A household spending 30 percent on a wider set of wants after a raise and a household spending 30 percent on the same set at higher prices both read as on target, and so does a household whose wants have simply grown to fill the band. The measure is a proportion, so it is blind to what the proportion buys. The rupee figures in the calculator move. The reading does not.
The per-person line is the one figure that does not flatten out. The needs band divided by household size climbs in step with income while the household stays the same size, which is why that line sits further from the survey benchmark at higher incomes and closer to it as the count goes up. It is the only place on this page where two of the inputs meet, and even there it is a division rather than a calibration.
What the calculator does about any of this is nothing, deliberately. The same three shares are applied at ₹10,000 and at ₹10,00,000. Shifting the shares by income level would need a dated source for where they ought to move and by how much, and this page puts no number into its own arithmetic that it cannot point at a source for. That is the same reason the city rent multiplier and the per-dependent weighting are absent from it.
The Same Split Worked Through At Four Income Levels #
The section above sets out why a proportional rule reads differently at the two ends of the range without the arithmetic ever saying so. These four pages carry that out in rupees at four points along it, each one opening the calculator with its own figure already filled in. The shares are identical on all four. What changes is the size of what those shares leave behind, and which row runs out first.
- A budget on a ₹40,000 take-home, where the needs row is usually the one that will not fit.
- A budget on a ₹1 lakh take-home, the point at which the savings row first has real room in it.
- A budget on a ₹2 lakh take-home, where the 30 percent wants band becomes the loose one.
- A budget above ₹5 lakh a month, which is a band rather than a figure, so that page names the income it works from before it counts anything.
Each of those pages runs the same three shares on a stated income and nothing more. None of them says what a household at that income ought to spend, because the split is arithmetic on a number that was entered rather than a reading of what the number means.
When Needs Run Past The Fifty Percent Band #
The three shares always add to a hundred, so the calculator above can never report a shortfall. Necessary spending is not bound by that arithmetic. Where it runs past the 50 percent the rule allocates, none of the figures on this page change, because none of them are measured. What changes is the distance between what the rule allocates and what is actually left, and the calculator has no input that would let it see the difference. This section states what the model does in that case, which is nothing, and sets out the arithmetic that follows from its own shares.
Because the shares are fixed, the residual is fixed with them. Where the wants band is filled at the full 30 percent, whatever the needs row takes above 50 comes out of the savings row one point at a time. The table below reads the needs share at its real level rather than at the allocated 50 and shows what the rule's own numbers leave behind it.
| Needs actually taking | Left for wants and savings | Savings left once wants fill 30% |
|---|---|---|
| 50% | 50% | 20% |
| 55% | 45% | 15% |
| 60% | 40% | 10% |
| about 66% | about 34% | about 4% |
| 70% | 30% | nothing |
| 75% | 25% | five points short |
The row where the savings band reaches zero is 70 percent, and that follows from the wants share alone rather than from anything about Indian incomes. The about-66-percent row is on the table only because it is a figure this page already carries. The five NSO item groups filed under needs, food and conveyance and rent and education and medical care, sum to about 66 percent of average urban spending, and it sits just below the row where the rule runs out. Those five shares add to 66.54 precisely, so the true distance is nearer three and a half points than four, and the residual on that row is nearer three and a half percent than the rounded figure the table carries.
That comparison is looser than it looks and is worth reading as such. The survey figure is a share of what an average urban person spends. The bands on this page are shares of what a household takes home. Those two denominators part company wherever spending and take-home pay are not the same number, which is most households in one direction or the other. The 66 percent is also an all-urban average across owners and renters and every income level, so it describes no particular household, and nothing here turns it into one.
What the calculator does about any of this is nothing, and that is deliberate rather than an omission. It asks for one number, take-home pay, and it has no field for what a household actually spends, so it cannot detect an overrun and does not flag one. The savings figure it reports stays an allocation rather than a residual, and the emergency fund and step-up rows above are sized off that allocation. Where real needs run above the 50 percent band, those rows describe the rule and not the month.
Modelling the overrun instead of describing it would need a distribution of households by the share their necessities take, and the survey behind this page publishes item-group shares of average spending rather than that distribution. No other source cited below carries it either. Putting a correction in without one would mean a number inside the arithmetic that this page cannot point at a source for, which is the same reason the city rent multiplier and the per-dependent weighting are absent from the calculator.
What The Savings Band Assumes About Protection #
The calculator reports the savings band as a single figure and stops there. Read plainly, that figure is what is left once the necessary bills and the discretionary spending are taken out. It is not a statement that the whole of it is free to be invested, and on one specific point the arithmetic is blind in a way worth naming.
Where a household has no life cover in force, the savings band reads higher than the same household's band would read once cover is bought. The reason is the classification already stated above. A premium that keeps protection in force is a recurring bill, so it files under needs, which means buying cover moves money out of the savings row rather than adding a fourth row. A household comparing its own savings figure against another's is not comparing like with like unless both are carrying the same protection. Nothing on this page adjusts for that, because the page never asks whether any cover exists.
The figure usually quoted when the question turns to how much cover is a multiple of annual income, most often somewhere between ten and twenty times. That multiple is a practitioner rule of thumb rather than a level set by any Indian regulator, and no Indian survey cited on this page publishes a per-household cover norm that could be quoted here with a date against it. It is a starting point people use, and it is worth reading as exactly that.
This page does not compute a cover amount, and the omission is deliberate rather than unfinished. Working out what a particular household should hold is a need analysis, which depends on existing cover, liabilities, dependants and health, none of which this page asks for. Under the IRDAI rules that govern a point of sale person, that analysis sits with the licensed broker rather than with a distribution surface like this one. Putting a rupee figure here would mean a specific cover conclusion produced from one number, take-home pay, which is neither honest arithmetic nor within what this page is permitted to do.
What the calculator can do with protection it already does. Once a premium is known it is a needs bill like rent or electricity, and where it falls once or twice a year rather than monthly, the set-aside rows inside the calculator turn it into a monthly figure. That is the whole of the honest treatment. The sizing question belongs elsewhere, and saying so is more useful than a multiple dressed up as a result. The Ladder steps on term life cover and health cover are where that question is worked through.
What This Does and Does Not Tell You #
It does give a flat first frame that works the same for every income, which is exactly why it is a sketch and not a plan.
It does not yet shift the split itself for the two things that move an Indian household budget the most, the city rent burden and the number of people the income supports. The household count divides the needs band per person, which makes the squeeze visible, but the needs share stays at 50 percent whether one person or six are living on it. A metro renter and a small-town homeowner on the same salary do not have the same needs band. No rent calibration is applied here, so the needs band reads better as a floor than as a ceiling. The household the split strains hardest against is one earner carrying every dependent alone, and that case is worked through in the guide to budgeting for single parents.
What Urban India Actually Spends #
The rule's neat bands sit oddly against what Indian households report. In the NSO's Household Consumption Expenditure Survey for 2023-24, the latest round available, the average urban person spent ₹6,996 a month. Food took 39.68 percent of that, conveyance 8.46 percent, durables 6.87 percent, rent 6.58 percent, education 5.97 percent and medical care 5.85 percent.
The rent share looks impossibly low because it averages owners and renters across all of urban India. A salaried renter in a metro runs far above it. That is why the 6.58 percent is worth reading as an all-urban average rather than as anyone's actual rent.
The savings gap is wider still. The rule asks for 20 percent while the RBI's Annual Report 2025-26 puts net household financial savings at 7.0 percent of gross national disposable income for 2024-25, up from 5.8 percent the year before. The 20 percent band is a target, not a description of what most households manage.
FAQ #
Is the 50/30/20 rule realistic for an Indian salary?
It is a target to aim at rather than a description of what most households manage. In the NSO's Household Consumption Expenditure Survey for 2023-24, food alone took 39.68 percent of average urban spending, and adding conveyance, rent, education and medical care brings the five item groups this page files under needs to about 66 percent. On the other side, the RBI's Annual Report 2025-26 puts net household financial savings at 7.0 percent of gross national disposable income for 2024-25 against the rule's 20 percent. The bands are still useful as a first frame, which is why this page shows them alongside the survey figures rather than instead of them.
Should I enter my take-home pay or my gross salary?
Take-home. That is the amount that lands in the bank each month, after tax and provident fund. Gross salary includes money that never reaches the account, so splitting it would size every band above what is actually there to spend.
Why does the emergency fund show three rows instead of picking one?
Because picking a multiple off a single income box would be a judgement about a particular household, and this calculator does not make one. Three, six and nine months of the needs band are shown side by side so the whole range is visible and the choice stays with the reader. The Ladder step on emergency funds sets out the reasoning behind where in that range a household tends to sit.
Does the household count change the needs, wants and savings split?
No. The count divides the needs band by the number of people the income supports, which makes the per-person squeeze visible, but the 50, 30 and 20 shares stay the same whether one person or six live on that income. Shifting the shares by household size would need a dated per-dependent weighting, and this page carries no source it could cite for one.
Does the calculator adjust for city rent differences?
No. The rent share behind these bands is an all-urban average across owners and renters, so a salaried renter in a metro runs well above it. A city multiplier is not applied here because this calculator will not put a number into its own arithmetic that it cannot point at a dated source for. Reading the needs band as a floor rather than a ceiling is the honest workaround for now.
Are the step-up SIP figures a projection of returns?
No. The step-up figures show only what the contribution itself becomes after four annual increases. No rate of return is assumed anywhere on this page, so nothing here is a forecast of what an investment would be worth. A stepped-up contribution finishes ahead of a flat one because more money went in, not because the money earned more.
Is this financial advice?
No. It is a free educational calculator that runs the same arithmetic on whatever numbers are entered, so the same inputs give the same output for everyone. It is not personal advice, not a financial plan and not a recommendation to buy or sell anything. FinSet is an AMFI-registered Mutual Fund Distributor, ARN 180462. On insurance it is a Point-of-Sales Person (POSP) / sub-broker under PolicyBazaar Insurance Brokers Pvt Ltd, an IRDAI-registered insurance broker. FinSet is not a SEBI-registered Investment Adviser.
Sources #
- NSO, Ministry of Statistics and Programme Implementation, Fact Sheet on the Household Consumption Expenditure Survey 2023-24, released 27 December 2024. Statement 4 carries the urban item-group shares used above and the average urban spend of ₹6,996 a month, counted without imputing items received free through welfare schemes, mospi.gov.in. No newer round had been conducted as of MoSPI's July 2025 statement to Parliament.
- Reserve Bank of India, Annual Report 2025-26, Chapter II Economic Review, released 29 May 2026. Net household financial savings rose to 7.0 percent of gross national disposable income in 2024-25 from 5.8 percent a year earlier, with gross financial savings at 11.8 percent and liabilities at 4.8 percent, rbi.org.in.
- Reserve Bank of India, Master Circular on Housing Finance, RBI/2025-26/16, DOR.CRE.REC.No.12/08.12.001/2025-26, dated 1 April 2025. This is the consolidated instrument for bank housing loans. Section 3, Quantum of Loan, sets the loan to value ratios and risk weights and excludes stamp duty and registration charges from the property cost used to compute them. Nothing in it limits the share of a borrower's income that may go to repayments, which is why the committed repayments block above reports a ratio and sets no ceiling, rbi.org.in. The RBI reissues its master circulars each 1 April, so the edition above is the one to check first.
- The 50/30/20 frame itself comes from Elizabeth Warren and Amelia Warren Tyagi's All Your Worth (2005). It is a teaching heuristic and carries no India-specific calibration of its own, which is what the two sources above supply.
Disclosures #
This is a free educational tool. The numbers it shows are general illustrations based on common rules of thumb, not personal advice and not a financial plan. Your own situation may need different amounts. Check with a qualified professional before you act.
Mutual fund investments are subject to market risks. Read all scheme related documents carefully. Past performance is never a guarantee of future returns.
FinSet is an AMFI-registered Mutual Fund Distributor (ARN-180462), not a SEBI-registered Investment Adviser. That ARN can be looked up in AMFI’s own distributor register, which searches on name or ARN code. On insurance FinSet is registered as a Point-of-Sales Person (POSP) / sub-broker under PolicyBazaar Insurance Brokers Pvt Ltd, an IRDAI-registered insurance broker. IRDAI publishes the registration number, licence category and validity date of every registered broker in its own register of insurance brokers. Where cover is arranged through FinSet that is the licensed broker the protection section above points to. Figures are illustrative and rounded.