Retirement Corpus Calculator

Ask twelve Indian calculators how much you need to retire and they will answer between eleven and forty-eight times your annual spending. They all assume similar inflation. The gap comes from one number almost none of them shows you, and this calculator puts it in front of you instead.

The multiple is an answer, not a rule

25x, 30x and 33x are not properties of a country. They are what falls out of the assumptions you feed a formula. The field treats them as rules handed down from somewhere, and then contradicts itself within a single page.

On 27 August 2026 we read twelve ranking India pages and reproduced each one's arithmetic from its own published figures. Three of the largest recommend one multiple in their text and deliver a different one in their own worked example.

The insurer's sentence is worth quoting exactly: it recommends "a lower 30-33x", which is higher than the 25x it is contrasting against. The prose and the calculator on these pages were written by people who never spoke to each other.

What actually decides the number

The post-retirement return. Not inflation, not the multiple, not your city. Every page agrees inflation is around 6% and agrees you must carry today's spending forward. The eleven-to-forty-eight spread comes from what they assume your corpus earns once you stop adding to it.

One page in twelve says this out loud, and it belongs to an independent analyst with a doctorate and no product to sell. His reasoning is the best in the field: a retired portfolio cannot hold much equity, so it is safer to assume it earns less than inflation. Two of the biggest tools instead assume you keep earning 8 to 12% for twenty-five years after you stop working, and one of them recommends holding 65 to 75% in debt on the same page.

Our formula reproduces that analyst's published figure to the rupee, which is how we checked the model before trusting any of our own numbers.

Where the 4% rule came from, and why it does not transfer

The 4% rule is American research from the 1990s, and its own author excluded tax from it.

Bengen's 1994 paper tested a portfolio of half US stocks and half intermediate Treasuries against American data from 1926. His conclusion is more fragile than the folklore: 4% "should be safe", while 4.25% "could exhaust a portfolio in as little as 28 years". And the assumption that matters most for an Indian reader is stated plainly in the paper, that all assets were assumed to sit in tax-deferred accounts, so capital gains tax was not a concern. He adds that with taxable assets "the conclusion might have been different".

The 1998 Trinity study, which gave the rule its name, says the same thing in one line: it "did not adjust for taxes or transaction costs". It also never claimed certainty. At 4% over thirty years its success rates were 95% for all stocks, 98% for a 75/25 split, and 20% for an all-bond portfolio.

There is one published study on Indian data, and it concludes the rule does not hold here. Raju and Saraogi presented it to the Institute of Actuaries of India in November 2024: Monte Carlo and circular bootstrap simulation, 50,000 iterations each, horizons of 25, 30 and 35 years, equity at 40 to 50%, and tax modelled at 30% on fixed deposits and 10% on equity. Their finding is that 3.0 to 3.5% is the appropriate Indian range, which is 28.6 to 33.3 times annual spending.

Their statement of the tax drag is the clearest anywhere: a nominal 5% return against 7% inflation means an investor in the 30% bracket loses about 3.5% a year in real terms once tax is paid on the nominal gain.

The part of the corpus you cannot spend

NPS forces part of your money into an annuity at exit, and the share is 20% for private-sector subscribers, not the 40% almost everyone publishes.

PFRDA's Exits and Withdrawals Regulations, as last amended on 20 July 2026, require a non-government subscriber to put at least twenty percent of accumulated pension wealth into an annuity on exit at or after 60. Government sector subscribers remain at 40%, and a premature voluntary exit still requires 80%. PFRDA's own FAQ page has not caught up and still shows 40% with a five lakh threshold, which is where most of the repetition comes from.

The bigger problem is what the locked money buys. An Indian annuity currently costs about twelve times the annual pension it pays, against a typical actuarial twenty-two, and increasing annuities are, in the words of the actuarial presentation above, "conspicuously missing from the Indian pension sector". So the payment never rises.

At 4% inflation, a pension fixed at retirement is worth 82% of its face value after five years, 68% after ten, and 47% after nineteen. Not one calculator in the field models this. The tool above does, if you tell it what share of your corpus sits in NPS.

How long to plan for

Life expectancy at 60 in India is 18.6 years, to roughly age 79, and planning to that number leaves half of people short.

That figure comes from the SRS Abridged Life Tables 2020-24, published by the Registrar General. It is very different from life expectancy at birth, which is 70.6, and using the wrong one is a real error in this field: one large broker's page assumes you die at 70 and therefore need only ten years of retirement.

Bengen put the correct response plainly, and it is the sentence the whole question turns on: by definition you have a 50% chance of outliving your life expectancy, so it is wise to add years. The tables are also period tables, which do not project future improvements in mortality, so they understate the horizon for someone retiring today. The default here plans to 90 for that reason.

Across the twelve pages we read, the assumed terminal age ranged from 70 to 100, and not one benchmarked it against any actuarial source. One page carried three different figures in three different places.

What this calculator does not do

It does not model tax on withdrawals or on the income the corpus generates. That is a deliberate omission rather than an oversight: the Income-tax Act 2025 renumbered the relevant provisions and the department's website could not be reached to verify the current position, and a confidently wrong tax figure here would be worse than none. The actuarial study above embeds tax in its return assumptions, which is why its numbers are higher than most.

It does not model continuing EPF contributions or gratuity. Existing savings are treated as a lump sum that grows, the same simplification every tool in this field makes.

It assumes one steady return and one steady inflation rate, which no economy has ever delivered. Read the range rather than the middle figure.

And it does not tell you whether your plan is adequate, what to invest in, or when to retire. Those are questions for a SEBI-registered investment adviser who can see your whole position.

Pair this with the guide

For the movement that popularised the 25x rule and what it actually claims, the FIRE movement explained covers the origins. For the mechanism doing the compounding on the way in, what compound interest actually does is the foundation, and the SIP calculator works the accumulation side with the fund fee and the tax subtracted. If you are relying on NPS, our NPS explainer covers the scheme itself.

Frequently asked questions

Sources

The research papers and regulations below were read in their own text on 27 August 2026. The full record, including a list of widely repeated figures we could not verify and have therefore left out, is kept in the repository as a working document.

• William Bengen, "Determining Withdrawal Rates Using Historical Data", Journal of Financial Planning, October 1994, for the origin of the 4% figure, the 50/50 US portfolio it was tested on, and its explicit assumption of tax-deferred accounts
• Cooley, Hubbard and Walz, AAII Journal, February 1998, the Trinity study, for the 1926 to 1995 test period and the statement that it did not adjust for taxes or transaction costs
• Raju and Saraogi, "Balancing Acts: Safe Withdrawal Rates in the Indian Context", as presented to the Institute of Actuaries of India, 12th Techtalk on Retirement Benefits, 8 November 2024, for the 3.0 to 3.5% Indian range, the simulation method, and Indian annuity pricing
• Pension Fund Regulatory and Development Authority, Exits and Withdrawals under the NPS Regulations 2015, last amended 20 July 2026, for the 20% non-government annuity floor, the 40% government floor and the 80% premature-exit floor
• Office of the Registrar General of India, SRS Abridged Life Tables 2020-24, for life expectancy at age 60 of 18.6 years
• Reserve Bank of India for the 4% inflation target and 2 to 6% tolerance band under section 45ZA, retained on 25 March 2026, and the August 2026 Monetary Policy Committee minutes
• Ministry of Statistics and Programme Implementation for CPI inflation of 4.45% and CPI health inflation of 1.34% in July 2026
• Observations about how other retirement calculators handle multiples, returns, longevity and the NPS annuity come from a live review of twelve ranking India pages on 27 August 2026, reading served markup and JavaScript bundles rather than marketing copy