F.I.R.E. Analysis: the mathematics of financial independence

✍️ AlphaFrame Editorial Team•Published 4/11/2026•Updated 4/11/2026•18 min read

F.I.R.E. Analysis: the mathematics of financial independence

The F.I.R.E. (Financial Independence, Retire Early) movement has transformed personal financial planning into a rigorous quantitative discipline. Born in English-speaking online communities in the early 2010s — with academic roots dating back at least to 1994 — the core idea is straightforward: accumulate enough wealth to sustain your lifestyle indefinitely by periodically withdrawing from investment returns. This article examines the mathematical models underlying FIRE, their assumptions, their limitations, and how to apply them in practice using the AlphaFrame F.I.R.E. Calculator.

1. Origins: from the 4% rule to the FIRE movement

The empirical foundation of FIRE is the seminal work of William Bengen (1994), published in the Journal of Financial Planning. Analysing historical returns of American stock-bond portfolios from 1926 to 1992 over 30-year horizons, Bengen concluded that a withdrawal rate of 4% per annum of the initial portfolio value — adjusted annually for inflation — had never depleted a portfolio in any historical 30-year period. This result became known as the 4% rule.

In 1998, Cooley, Hubbard and Walz — professors at Trinity University — published the study that became famous as the Trinity Study. They extended Bengen's analysis by introducing the concept of portfolio success rate: the percentage of historical periods in which the portfolio was not depleted. The Trinity Study has been updated multiple times (2009, 2011, 2021) incorporating more recent data.

In the 2010s, researchers such as Wade Pfau (2011) and Javier Estrada (2020) refined the model to account for market valuation (via Shiller's CAPE) and horizons exceeding 30 years — relevant for those retiring at 35–45. In parallel, Karsten Jeske (known as Early Retirement Now, ERN) developed from 2018 a Block Bootstrap simulation methodology that represents the current state of the art for Sequence of Returns Risk analysis.

2. The FIRE Number: how much do I need?

The starting point of any FIRE analysis is computing the target wealth, called the FIRE Number. The formula is immediate:

FN = \frac{C}{SWR}

where $C$ is the annual expenditure sustainable in retirement (in real terms, i.e. at constant purchasing power) and $SWR$ is the Safe Withdrawal Rate. With $SWR = 4\%$ :

FN = \frac{C}{0.04} = 25 \times C

The "25× multiplier" has become the FIRE rule of thumb: you need a portfolio equal to 25 times your annual expenses. With $SWR = 3.5\%$ the multiplier rises to approximately 28.6 — more conservative and appropriate for horizons exceeding 40 years.

Critical assumption: the FIRE Number assumes the portfolio is fully invested in financial assets with a positive expected real return. It excludes owner-occupied property, labour income, state pension, or other illiquid assets. Ignoring these sources makes the FIRE Number conservative — which may be desirable.

3. The Trinity Study: historical success rates

Cooley, Hubbard and Walz (1998, updated 2021) systematised the concept of portfolio success rate: across all historical periods of length $T$ years available, what percentage did not exhaust the portfolio? Their dataset covers the American stock and bond markets from 1926 to the present (S&P 500 + long-term US government bonds).

Key results for a 50/50 stock/bond portfolio with inflation-indexed withdrawals show success rates of approximately 95–98% for 30-year horizons with $SWR \leq 4\%$ . For 40-year horizons results are more uncertain and heavily dependent on allocation: a 100% equity portfolio performs better over long horizons (Trinity Study 2021).

Limitations of the Trinity Study:

US-only dataset: does not capture international diversification or underperforming markets such as Japan in the 1990s.
Overlapping historical periods are not statistically independent: the success rate is not a true frequentist probability.
Uses historical US inflation, not directly applicable to EUR-denominated portfolios.
No modelling of taxes, transaction costs, or variable spending over time.

4. CAPE-based SWR: market valuation and withdrawal rates

Wade Pfau (2011) showed that the safe withdrawal rate is not a constant: it depends on market valuation at the time of retirement. Pfau estimated a linear model using Shiller's CAPE (Cyclically Adjusted Price-to-Earnings, ten-year moving average of real earnings) as a predictor:

SWR_t \;\approx\; \alpha + \beta\,\cdot\, \frac{1}{CAPE_t}

The intuition is direct: expensive markets at retirement (high CAPE) imply lower expected future returns, hence a lower SWR to ensure sustainability. With CAPE around 30 (S&P 500 level in 2024), Pfau's model suggests $SWR \approx 3.0\%\text{–}3.5\%$ .

Model assumption: the CAPE–SWR relationship is estimated on US historical data. In a European or global context, CAPE has different predictive power. The model also assumes the portfolio is periodically rebalanced and that allocation remains stable over time.

5. Long horizons: Estrada's contribution (2020)

Javier Estrada (2020, Journal of Financial Planning) extended the FIRE analysis to 40-, 50- and 60-year horizons using an international dataset of 21 countries (1900–2019). Key findings:

For 40-year horizons, a 3.5% SWR delivers success rates above 90% in the vast majority of markets studied.
For 50+ year horizons, 3.5% remains reasonable but cross-country dispersion is high: markets like Japan in the 1990s show much lower success rates.
International diversification significantly reduces depletion risk compared to single-country portfolios.

The AlphaFrame calculator uses SWR = 3.5% as the default for long retirement horizons, consistent with Estrada's and Pfau's literature.

6. The five FIRE variants: mathematical definitions

The term "FIRE" encompasses several distinct goals, all derived from the base formula but with different multipliers or conditions on spending. In all cases, expenditure $C$ is relative to the individual's own lifestyle — not an absolute threshold.

6.1 Regular FIRE

The base case: maintain current spending exactly in retirement.

FN_{\text{regular}} = \frac{C}{SWR}

6.2 Lean FIRE

Frugal lifestyle in retirement, expenses reduced 30% from current levels:

FN_{\text{lean}} = \frac{0.70 \cdot C}{SWR}

Lean FIRE requires less wealth but implies real trade-offs. It suits those with already low expenses or who plan to optimise drastically (e.g. relocating to low cost-of-living areas, eliminating subscriptions, lifestyle redesign).

6.3 Fat FIRE

Premium lifestyle, expenses increased 50% from current levels:

FN_{\text{fat}} = \frac{1.50 \cdot C}{SWR}

Fat FIRE includes a significant buffer for discretionary spending, travel, giving or emergencies. It requires 2.14× the wealth of Lean FIRE, for the same SWR.

6.4 Barista FIRE

Semi-retirement: the individual continues working part-time with annual income $I_{pt}$ , funding only the gap from the portfolio:

FN_{\text{barista}} = \frac{\max(0,\; C - I_{pt})}{SWR}

If $I_{pt} \geq C$ , the FIRE Number is zero: part-time income fully covers expenses. Barista FIRE is particularly attractive for those wanting to exit full-time work without exiting work entirely, while preserving benefits such as health coverage, intellectual stimulation and social connection.

6.5 Coast FIRE

The most mathematically elegant variant: to what wealth $P_t$ must I arrive today (at age $t$ ) so that, without any further contributions, compound growth alone brings the portfolio to the FIRE Number by target age $T$ ?

CoastFN = \frac{FN_{\text{regular}}}{(1+r)^{T-t}}

where $r$ is the expected real annual return and $T - t$ is the years remaining to target retirement. Once Coast FN is reached, one can stop saving (or dramatically reduce contributions) and let compounding do the work.

Example: with $C = \$36{,}000$ , $SWR = 3.5\%$ , $r = 5\%$ , current age 30, target age 55:

FN = \frac{36{,}000}{0.035} = \$1{,}028{,}571

CoastFN = \frac{1{,}028{,}571}{(1.05)^{25}} = \frac{1{,}028{,}571}{3.386} \approx \$303{,}780

7. The years-to-FIRE formula

Given current wealth $P$ , monthly contribution $A$ and monthly real return $r = (1+r_a)^{1/12} - 1$ , the future value of the portfolio after $n$ months is:

FV_n = P\,(1+r)^n \;+\; A\,\frac{(1+r)^n - 1}{r}

Setting $FV_n = FN$ and solving for $n$ :

(1+r)^n \;=\; \frac{FN + A/r}{P + A/r}

n \;=\; \frac{\displaystyle\ln\!\left(\frac{FN + A/r}{P + A/r}\right)}{\ln(1+r)}

Years to FIRE are therefore $n/12$ . This is the exact closed-form solution for any combination of initial wealth and periodic contribution. The formula reduces to two limiting cases:

A = 0 (no contributions, compounding only): $n = \ln(FN/P) / \ln(1+r)$
r = 0 (zero return): $n = (FN - P) / A$ months

Note: the formula assumes monthly contributions that are constant in real terms. In practice contributions grow with career progression — this simplification is conservative (overestimates time needed).

8. Savings rate and acceleration toward FIRE

The savings rate is the single most powerful variable in determining speed to FIRE. Defining net income as $Y = C + S$ (expenses + monthly savings):

SR = \frac{S}{Y} = \frac{Y - C}{Y}

The impact of SR is twofold: a higher SR increases the monthly contribution $A = S$ and simultaneously reduces expenses $C$ , lowering the FIRE Number. Increasing SR from 20% to 50% can reduce time to FIRE from over 40 years to approximately 15–17 years (at constant real return), as demonstrated in the foundational studies of the Mr. Money Mustache blog (Adeney, 2012).

9. Sequence of Returns Risk (SORR)

Sequence of Returns Risk is the phenomenon whereby the same average return produces radically different outcomes depending on the temporal order of returns. During the withdrawal (decumulation) phase, a sequence of negative returns in the early years of retirement is devastating: assets are sold at low prices to fund spending, permanently reducing the base that will benefit from the subsequent recovery.

Formally, let $W_t$ be wealth at year $t$ :

W_{t+1} = W_t \cdot (1 + r_{t+1}) - C

If $r_1 = -30\%$ (catastrophic year 1) followed by $r = +8\%/\text{year}$ for 29 years, the portfolio behaves dramatically worse than the same scenario with reversed returns (first +8%, then −30%). The annual average is identical; the final outcome is completely different.

SORR is the primary reason why:

The 4% SWR is lower than the historical average real equity return (7–8% real).
Maintaining a bond or cash allocation during the withdrawal phase is recommended.
"Guardrail" and "bucket" strategies have been developed to mitigate its impact.

10. Monte Carlo simulation and Block Bootstrap

Monte Carlo simulations quantify the distribution of possible outcomes, overcoming the limitations of deterministic analysis. The AlphaFrame calculator implements the Block Bootstrap method developed and popularised by Karsten Jeske (ERN, 2018–2024).

10.1 Naive Monte Carlo (and why we don't use it)

The simplest approach is to sample returns from a normal distribution:

r_t \sim \mathcal{N}(\mu,\, \sigma^2)

The problem is that real equity returns are not Gaussian: they exhibit fat tails, negative skewness and short-term autocorrelation. A normal distribution significantly underestimates the probability of years like 2008 (−40%) or 2022 (−18%), which are critical for SORR modelling.

10.2 Block Bootstrap

Block Bootstrap solves these problems by sampling contiguous blocks of actual historical returns rather than synthetic ones. For each simulation:

The historical series is divided into blocks of length $L$ (in our case $L = 3$ years).
Blocks are sampled with replacement until the desired horizon is covered.
The return sequence is applied to the portfolio, including monthly contributions.

W_{t+1} = W_t \cdot (1 + r_{t+1}^{*}) + A

where $r_{t+1}^{*}$ is the bootstrapped return. The 3-year block length ensures that correlated crises (e.g. 2000–2002, 2008–2009) appear jointly in simulations, preserving the temporal autocorrelation structure.

The dataset embedded in the AlphaFrame calculator comprises annual real returns of the MSCI World from 1970 to 2024 (55 observations), including: oil crisis (1973–1974), Black Monday (1987), dot-com crash (2000–2002), Great Recession (2008–2009), and bear market 2022. With $N = 2{,}000$ simulations and 3-year blocks, the method captures both average volatility and the fat tails of global equity return distributions.

11. Model assumptions and limitations

Every FIRE model rests on a set of assumptions that are essential to understand before making important decisions:

Constant real return — The deterministic model uses a fixed annual real return. Monte Carlo relaxes this assumption but uses historical data that may not replicate in the future (the "this time is different" problem).
Constant real spending — In reality expenses change with age (declining in later years, increasing for healthcare). The model does not capture this dynamic.
No tax modelling — In Italy capital gains on ETFs are taxed at 26% (with some exceptions). A nominal SWR of 3.5% becomes approximately 2.6% net of taxes — requiring a higher FIRE Number or higher gross SWR to compensate. In the UK, ISA wrappers allow tax-free growth; in the US, Roth accounts serve a similar role.
No alternative income — State pension, rental income or part-time work can significantly reduce the FIRE Number. Ignoring them is conservative but not always realistic.
Limited historical dataset — 55 years of MSCI World data are insufficient to precisely estimate fat tail behaviour. Block Bootstrap mitigates the issue but does not eliminate it.
Uncertain horizon — Life expectancy cannot be known in advance. For a retirement at 40, a 60+ year horizon is realistic; the classic FIRE model (30 years) may be insufficient.

12. How to use the AlphaFrame F.I.R.E. Calculator

The F.I.R.E. Calculator implements all the models discussed in this article. Follow these steps:

Choose your currency — EUR, USD or GBP. The default inflation adjusts automatically (2% for EUR, 2.5% for USD/GBP).
Enter your age and current wealth — Target age determines the horizon for the Coast FIRE Number.
Estimate your expenses — Use direct entry or the category wizard (housing, food, transport, etc.) which automatically computes the annual total.
Select your FIRE type — Regular, Lean, Fat, Barista or Coast. For Barista, enter expected monthly part-time income.
Advanced parameters — Adjust SWR (default 3.5%), expected real return (default 5%) and inflation if you have different estimates.
Run the Monte Carlo — The Block Bootstrap generates 2,000 possible paths. The chart shows the 5th, 25th, 50th, 75th and 95th percentiles of wealth over time.

Results display FIRE Number, estimated years to FIRE, projected FIRE age and implied savings rate, alongside a comparative table of all five FIRE variants.

13. References

Bengen, W. P. (1994). "Determining Withdrawal Rates Using Historical Data." Journal of Financial Planning, 7(4), 171–180.
Cooley, P. L., Hubbard, C. M., & Walz, D. T. (1998). "Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable." AAII Journal, 20(2), 16–21. [Trinity Study]
Cooley, P. L., Hubbard, C. M., & Walz, D. T. (2011). "Portfolio Success Rates: Where to Draw the Line." Journal of Financial Planning, 24(4), 48–60.
Pfau, W. D. (2011). "Safe Savings Rates: A New Approach to Retirement Planning Over the Life Cycle." Journal of Financial Planning, 24(5), 42–50.
Estrada, J. (2020). "Managing to Target: Dynamic Asset Allocation for Goal-Based Investing." Journal of Financial Planning, 33(2), 38–50. See also: Estrada, J. (2018). "Replacing the Failure Rate: A New Success Metric for Retirement Portfolios." Journal of Wealth Management.
Karsten, K. (ERN). (2018–2024). "Safe Withdrawal Rate Series." Early Retirement Now Blog. earlyretirementnow.com
Shiller, R. J. (1981). "Do Stock Prices Move Too Much to Be Justified by Subsequent Changes in Dividends?" American Economic Review, 71(3), 421–436. [CAPE ratio]
Adeney, P. (Mr. Money Mustache). (2012). "The Shockingly Simple Math Behind Early Retirement." mrmoneymustache.com.

Disclaimer and warnings

This article and the associated calculator are for educational and illustrative purposes only. The models presented are based on historical data and assumptions that may not replicate in the future. Past performance is not a guarantee of future results. The FIRE Number computed does not constitute an investment recommendation or personalised financial, legal or tax advice.

Applying FIRE models to one's own financial situation requires the evaluation of individual variables — time horizon, risk tolerance, tax position, pension entitlements, family obligations — that lie beyond the scope of any automated simulation. Retirement planning is a complex subject; before making material decisions it is advisable to consult a qualified independent financial adviser.

AlphaFrame Analytics accepts no liability for decisions made on the basis of the calculator's outputs. Monte Carlo simulations are probabilistic tools: even a 95th percentile outcome may prove insufficient in historically unprecedented scenarios.