Coast FIRE: The Retirement Number Nobody Coined (2026)

Coast FIRE is a checkpoint, not a finish line. You have reached it when your existing investments, with no further contributions, are projected to compound into your FIRE number by your chosen retirement age. The entire concept is one line of arithmetic. Take your FIRE number and discount it by growth: coast number = FIRE number ÷ (1 + r)^t, where r is your assumed annual return and t is the years until you want the money. it is just the compound-interest formula run backwards, simple enough for anyone to recompute.

Our 4% rule guide could trace its number to a 1994 journal article and name its author. Coast FIRE has no such trail: no paper, no author, and, as far as we can trace, no coiner — the earliest use we can find already describes the term as something the writer had "seen called" coast FI. The label is folk taxonomy; the math is checkable arithmetic; this guide treats the two accordingly. It is the next entry in our pedigree series, which traces famous numbers to their primary documents — this time, a number whose document does not exist.

Before anything else: this guide is informational. Personalised projections about retiring on specific savings are regulated financial advice in both Singapore and Malaysia, and nothing here tells you what your numbers should be. For those, talk to a licensed adviser.

The arithmetic

The FIRE number itself comes from the 25× convention: 25 × annual retirement spending, which is just 1 ÷ 4%. The coast number inherits every assumption from the FIRE number, so all the caveats in our 4% rule guide apply here too. Our FIRE number guide, published alongside this one, takes that multiple apart on its own.

A worked example, done twice to expose the trap. A 30-year-old targets retirement at 60 (t = 30) on spending of $40,000 a year in today's dollars, so the FIRE number is $1,000,000 in today's dollars.

Path A, real return. Assume 5% real, meaning after inflation. Then the coast number is 1,000,000 ÷ 1.05^30 = $231,377, in today's dollars. The check runs forward: $231,377.45 × 1.05^30 = $1,000,000.01.

Path B, nominal return. Assume 8% nominal and 2.5% inflation. Now the target must be inflated first: spending in 30 years is 40,000 × 1.025^30 = $83,903 a year, so the nominal FIRE number is $2,097,568, and the coast number is 2,097,568 ÷ 1.08^30 = $208,451.

The two paths agree once you notice they assume slightly different real rates. The real return implied by 8% nominal over 2.5% inflation is 1.08 ÷ 1.025 − 1 = 5.366%, and discounting $1,000,000 at that rate for 30 years gives the same $208,451. Consistent frames, consistent answers.

Real versus nominal, the qualifier that does the work

The classic error is to take the nominal rate and the today's-dollars target together: 1,000,000 ÷ 1.08^30 = $99,377. That is less than half the consistent answer. A saver who declares victory at $99,377 and coasts will reach $1,000,000 in nominal dollars — which, at 2.5% inflation, buy roughly what $477,000 buys today. The plan quietly delivered half the retirement it promised, and no single year's statement would have flagged it.

The rule is to keep your frames consistent. If your return assumption is nominal, inflate the FIRE number. If it is real, the result is already in today's dollars. Never mix the frames. Our Coast FIRE calculator takes a real return, labelled "after inflation", precisely so that its output stays in today's dollars end to end.

How much the rate assumption moves the answer

The coast number is highly sensitive to r, because the divisor compounds. Same $1,000,000 target in today's dollars, real rates:

At 30 years out, moving the assumption from 4% to 7% real cuts the required pot by 57%. Nothing about the saver changed; only the bet on future markets did. The coast number is mostly that bet.

Which return assumptions are defensible

Two primary datasets anchor the range. Aswath Damodaran's NYU Stern tables cover US returns from 1928 to 2025 — 98 annual rows. His page prints the annual returns but no period averages, so we recomputed them from every row: the S&P 500 including dividends returned a geometric 10.02% a year nominal (arithmetic 11.86%), against 3.37% for 3-month T-bills and 4.53% for 10-year Treasuries. For a multi-decade compounding projection the geometric mean is the relevant one; the arithmetic mean overstates terminal wealth. The 2025 entry, for the record: 17.78%. The table carries no inflation column, so real returns need a source that publishes them.

The UBS Global Investment Returns Yearbook 2026 (Dimson, Marsh and Staunton, DMS Database 2026) provides that inflation data, covering 1900–2025. US equities: 9.8% a year nominal, 6.6% real, against 2.9% average inflation; real bonds 1.6%, real bills 0.5%. Its summary puts the compounding plainly: "An initial investment of USD 1 grew to USD 124,854 in nominal terms by end-2025." In real terms, $1 grew to $3,296.

Both datasets describe one country's unusually good century. Wade Pfau's test of 17 developed markets over 1900–2008 found the worst-case withdrawal rate exceeded 4% in only four of them; the same home-bias warning applies on the accumulation side. A 100%-equity portfolio riding US history is the ceiling, not the base case. Real portfolios hold bonds at 1.6% real and cash at 0.5%, pay fees, and may not be US-listed. That is why this guide's sensitivity band runs 4–7% real, with 5% a reasonable centre for a diversified portfolio. Pick the top of the band and you have roughly halved your required pot by assumption alone.

Who coined it? Nobody we can find

Put plainly, every early use of the term we can trace already attributes it to some prior usage that nobody can find. There is no first post, no named coiner, no academic or regulatory source. The trail simply starts mid-sentence, in mid-2016.

A method note, because provenance claims deserve their own provenance. Reddit blocks automated retrieval, so we located the comments below through PullPush, an independent archive of Reddit, queried oldest-first on 22 July 2026, with permalinks rebuilt from the archived ids. An archive is not guaranteed complete, so "earliest traceable" means the earliest we found, not provably the first; a fuller record could only push the origin further back into the same fog.

One origin story circulating in search results claims the term came from a 2016 ChooseFI podcast episode. We could not locate any such episode; the page we could fetch that surfaces alongside the claim (nickwolny.com) makes no origin claim itself; and the June 2016 Reddit usage predates ChooseFI's podcast catalogue as far as we can trace. Our own Coast FIRE calculator repeated a version of that story until today's audit removed it. Treat the claim as circulating and unsupported.

Barista and Coast, born entangled

The single earliest comment offers "barista FI or coast FI" as interchangeable names. The two labels were born entangled; the distinction now drawn between them is later convention, not anyone's definition. As it settled: Coast means you stop contributing, keep working to cover current living costs, and let compounding finish the retirement job; Barista means deliberately downshifting to part-time work that covers part of your expenses — the name, in Nick Wolny's telling, "a nod to Starbucks, which offers employees health insurance if they average a 20-hour work week". The Fioneers put the Coast side as "Coast FI means that you already have enough saved and invested for your traditional retirement", and WalletBurst now maintains separate Coast and Barista calculators. The distinction has hardened into software, which is as official as a folk term gets.

What coasting does not protect against

The formula answers one narrow question — what pot, growing untouched at exactly r, hits the target on schedule. Everything else remains live risk.

• Sequence of returns at the far end. Arriving at your FIRE number on schedule says nothing about the withdrawal decade that follows. Morningstar's replay of rolling 30-year windows finds starting safe rates from 3.9% (retiring late 1968) to 10.5% (mid-1982) for a 50/50 portfolio, and Bengen's description of the mechanism stands: an early bear market "sucks a lot out of the portfolio at the same time that you're drawing from it". The 4% rule guide covers this end in full.
• The coast years are themselves a sequence bet. The model assumes a smooth r with no contributions. A 20-year stretch that averages 5% real but front-loads a lost decade still reaches the target only if you never touch the money mid-path — and a coaster has, by construction, no new contributions buying the dips. Our DCA/SIP calculator models the contributions-continue counterfactual if you want to see what coasting gives up.
• Stopping contributions stops more than savings. In Singapore, CPF contributions ride on employment: without an employer, MediSave is the only mandatory contribution — CPF Board's self-employed guidance is framed around "the absence of regular MediSave contributions from an employer" — and employer inflows to the OA and Special/Retirement Accounts cease. Employment-tied group insurance lapses with the job, a cash cost the coast budget must absorb. In Malaysia, statutory EPF contributions likewise stop without employment, though voluntary schemes such as i-Saraan exist.
• The target drifts. The coast number is a function of assumed retirement spending, so lifestyle inflation repriced the goal while you were coasting. At the 4% convention, every permanent $1 a year of new spending adds $25 to the FIRE number; a $5,000-a-year upgrade adds $125,000 to the target and, at 5% real with 30 years to run, about $28,922 to the coast number you were supposed to have already cleared.
• The label protects nothing. No regulator defines Coast FIRE and no standard exists to comply with. Declaring yourself "coasting" has exactly the legal and financial force of declaring yourself tall.

Singapore and Malaysia: guaranteed floors change the equation

Coast math is usually run against market assumptions, where r is a hope with a confidence interval. Singapore and Malaysia add accounts with guaranteed or declared rates, a different kind of variable that changes what the formula means.

Under CPF's current structure (rates for 1 July – 30 September 2026), the Ordinary Account earns 2.5% a year, its legislated minimum. The Special, MediSave and Retirement Accounts earn 4%, a floor extended through 31 December 2026. Extra interest adds 1% on the first $60,000 of combined balances (OA share capped at $20,000), and members 55 and above earn a further 1% on the first $30,000 — up to 6% on that first tranche. The backing is stated flatly: "CPF savings are invested in Special Singapore Government Securities (SSGS) which are guaranteed by the Government."

One structural change matters for anyone coasting past 55. On 19 January 2025 the Special Accounts of about 1.4 million members aged 55 and above were closed. SA savings moved to the Retirement Account up to the member's cohort Full Retirement Sum, where they keep earning the long-term rate; the CPF Board's release states the rest plainly: "Any remaining SA savings have been transferred to their Ordinary Account (OA) where they will earn 2.5% interest per annum, and members have the flexibility to withdraw them when needed." Members can still transfer OA to RA for the 4% rate, but the default guaranteed leg of an above-FRS coast projection is now 2.5%, not 4%, for that cohort.

Malaysia's EPF declares a dividend annually rather than guaranteeing a floor in advance. For 2024 it declared 6.30% for both Simpanan Konvensional and Simpanan Shariah, a total payout of RM73.24 billion, announced 1 March 2025. For 2025: 6.15% for both, RM67.1 billion plus RM12.5 billion for a total of RM79.6 billion, announced 28 February 2026. Those are handsome numbers — and they are declared after the fact, in nominal ringgit, with no forward promise. Our EPF vs CPF guide compares the two systems in full.

Now run the coast formula against the floors. Pot required today to reach $1,000,000 in 30 years:

Read that table with both eyes open, because it mixes frames on purpose to show the trap. The CPF floors are nominal guarantees; the 5% market line is real. Side by side without that caveat, the comparison flatters the market case — the guarantee removes return risk, not inflation risk, and its real value depends on what inflation does over your 30 years. The other half is equally true: money coasting on a government-guaranteed 4% cannot suffer a 1970s-style equity drawdown, which is precisely the risk the sequence section above describes. A floor account buys certainty of the nominal path at the price of a much larger required pot.

Our calculators, and what we fixed today

We audited our Coast FIRE calculator against this guide's sources. As of today (22 July 2026), its explanatory text is corrected. An unverified origin claim is gone, a mismatched percentage ladder now matches the correct figures, and several worked examples have been recomputed. The calculation engine itself was already right. It implements this guide's arithmetic exactly: coast number = (annual expenses ÷ withdrawal rate) ÷ (1 + r)^years to target age. The calculator takes a real return, labelled "after inflation", so results stay in today's dollars. It then projects your current pot forward at five-year intervals with zero contributions and estimates the years until you reach coast using year-end contributions. It deliberately models no taxes, fees, variable returns, or CPF/EPF floors; to approximate a floor account, convert the floor to a real rate under your own inflation assumption first — the nominal-versus-real rule again.

The FIRE number calculator handles the 25× step that feeds this one, and the investment returns calculator is the bare (1 + r)^t engine if you want to check any figure in this guide by hand. All of them are aids for exploring the arithmetic, not advice engines.

FAQ

Who coined the term "Coast FIRE"?

Nobody identifiable. The earliest use we can trace, a Reddit comment of 22 June 2016, already calls it "what I've seen called a 'barista FI' or 'coast FI'", and every early instance after it likewise points to prior usage nobody can produce. The circulating claim that it began on a 2016 ChooseFI podcast episode is unsupported: we could not locate any such episode.

Is my coast number in today's dollars?

Only if your return assumption is a real (after-inflation) rate. If you use a nominal rate, you must inflate the FIRE number first, and the result is in future dollars. Mixing a nominal rate with a today's-dollars target understates the required pot — in this guide's worked example, $99,377 against the consistent $208,451.

What is the difference between Coast FIRE and Barista FIRE?

By current convention, Coast means you stop retirement contributions and work only to cover present living costs; Barista means downshifting to part-time work that covers part of your expenses, with the name a nod to Starbucks' health insurance for 20-hour-a-week employees. The distinction is convention, not definition — the earliest traceable comment used the two labels interchangeably.

Can I run Coast FIRE math on CPF or EPF savings?

The formula applies to any pot, but mind the frames. CPF's floors (2.5% OA; 4% SA/MA/RA through 31 December 2026) are nominal guarantees, so a coast projection built on them removes return risk but not inflation risk. EPF dividends (6.30% for 2024, 6.15% for 2025) are declared after each year in nominal ringgit, not promised in advance. And employer CPF/EPF inflows cease with employment, which a contribution-stop plan must absorb. Decisions here belong with a licensed adviser.