TL;DR. “Save six months of expenses” is a rule of thumb, not a calculation. Runway is savings divided by a monthly gap: personal burn plus business costs plus a tax reserve plus the benefits your employer used to pay, minus new revenue. The gap shrinks as revenue ramps, so you need the sum of the gaps until break-even, plus a fallback long enough to get a job back if the bet fails, plus a margin. In the Federal Reserve’s 2025 Small Business Credit Survey, 25 percent of employer firms aged 0-2 years were operating at a profit, against 42 percent at 3-5 years. In September 2026 the median US jobseeker had been searching for 11.5 weeks and the mean for 24.8 weeks. In our worked example, a 24-month ramp and a six-month fallback require $103,860, or 23.1 months of personal spending: 3.85 times the six-month rule. The CEO move: set the quit date by the numbers, with a ramp already tested on the side. The student move: re-forecast the ramp every quarter and pull the fallback trigger you wrote down in advance.
This is the transition layer of our variable income money system for freelancers and solo operators, the derivation behind the “Runway” row of the freelancing versus full-time decision framework, and it takes its revenue target from how long it takes to replace a salary with independent income. It is general information, not financial, tax or legal advice.
Why is “save six months of expenses” the wrong number?
The rule divides savings by expenses and stops. Three things break it.
The drain is not your expenses. After quitting you also pay business costs, a tax reserve and the benefits your employer covered, and you earn some revenue. The net of those, the monthly gap, is what empties the account.
The drain changes shape. In month one the gap is close to your full spending. If the business grows, the gap shrinks every month. A fixed number of months cannot describe it.
There is no exit. Six months of expenses is a bet on reaching break-even within six months, with nothing set aside for getting back to a salary if the bet fails.
Many households could not fund even the rule. The Federal Reserve’s Survey of Household Economics and Decisionmaking for 2025 (fielded October 2025, published May 2026) updates the 2024 figures in our hub guide, with the headline shares unchanged: 55 percent of adults had a rainy day fund covering three months of expenses, and 63 percent would cover a $400 expense with cash or its equivalent. Thirty percent could not cover three months of expenses by any means, including borrowing or selling assets. Among adults aged 30-44, 49 percent had the three-month fund. Self-employment also changes the month itself: 58 percent of self-employed adults said their income varied from month to month and 22 percent struggled to pay bills because of it, against 28 and 10 percent of people working for someone else.
How long does a new business take to cover its owner?
Your own ramp is unknowable in advance; the base rate is not. The Federal Reserve Banks’ Small Business Credit Survey reports profitability by firm age.
Table 1. Profitability by firm age, US small businesses (figures as published)
| Group | Operating at a loss | Operating at a profit | Survey round |
|---|---|---|---|
| Employer firms, 0-2 years | 55% | 25% | 2025 survey, end of 2024 |
| Employer firms, 3-5 years | 35% | 42% | 2025 survey, end of 2024 |
| Employer firms, 6-10 years | 33% | 48% | 2025 survey, end of 2024 |
| Employer firms, 21+ years | 24% | 60% | 2025 survey, end of 2024 |
| Nonemployer startups, 0-2 years | 57% | 21% | 2023 survey, end of 2022 |
| Older nonemployers, 3+ years | 37% | 40% | 2023 survey, end of 2022 |
Sources: Federal Reserve Banks, 2026 Firms in Focus: Chartbook on Firms by Age of Business, page 7; 2024 Report on Startup Firms, pages 7 and 9. The rest of each row operated at break-even. Convenience sample, weighted to Census data.
Most young firms are not profitable yet. A quarter of employer firms aged 0-2 years and about a fifth of nonemployer startups operated at a profit.
Profit is a lower bar than your break-even. A sole proprietor’s own pay is not a business expense, so a firm “operating at a profit” may still not cover its owner’s household. For a solo owner, Table 1 is an optimistic reading of how many young firms reach the break-even defined below.
The owner is the bank. Among employer firms with financial challenges in the 2025 survey, 67 percent of those aged 0-2 years used the owner’s personal funds, against 39 percent of firms aged 21 years or more. In the 2023 survey the figure was 77 percent for nonemployer startups. Uneven cash flow troubled 53 percent of employer firms aged 0-2 years. A runway is the planned version of what most new owners do anyway; the owner-pay waterfall covers what happens once the business can pay you.
Check your own ramp against this outside view: break-even within 12 months means beating most of Table 1.
How long does getting a job back take if the bet fails?
If the ramp fails, you must fund a job search at full spending. The public anchor is how long unemployed people have been looking.
Table 2. Duration of unemployment, United States (figures as published)
| Measure | September 2026 | September 2025 | 2025 average | 2025, full-time jobseekers |
|---|---|---|---|---|
| Median, weeks | 11.5 | 10.1 | 9.9 | 11.1 |
| Mean, weeks | 24.8 | 24.1 | 23.0 | 24.7 |
| Share 27 weeks or longer | 27.1% | 23.6% | 23.2% | 25.4% |
| Share 52 weeks or longer | not in A-12 | not in A-12 | 13.8% | 15.0% |
Sources: BLS, The Employment Situation, September 2026 (released 2 October 2026), Table A-12, seasonally adjusted; BLS, CPS annual averages 2025, Table 30. The 2025 figures are 11-month averages that exclude October 2025, when data were not collected because of the federal government shutdown.
Two facts matter. BLS measures how long people “have been continuously looking for work”, and these are still-in-progress spells, not completed ones. And the mean is more than twice the median because some searches last very long: in September 2026, 27.1 percent of the unemployed, about 1.9 million people, had been looking for 27 weeks or more.
For a reserve the median is the wrong anchor: half of current searchers have already outlasted it. The mean is the base case and the 52-week tail the stress test. At 52/12 weeks per month, the September 2026 median is 2.65 months and the mean 5.72 months.
Table 3. Fallback anchors by group, United States, 2025 (CEOtudent analysis of BLS CPS annual averages, Tables 31 and 32)
| Group | Mean weeks | Share 27 weeks or longer | Base fallback, months |
|---|---|---|---|
| All unemployed | 23.0 | 23.2% | 6 |
| Age 25-34 | 22.7 | 23.3% | 6 |
| Age 35-44 | 25.0 | 26.0% | 6 |
| Age 45-54 | 27.0 | 28.6% | 7 |
| Age 55-64 | 26.7 | 25.0% | 7 |
| Management, business and financial occupations | 24.1 | 25.4% | 6 |
| Professional and related occupations | 21.6 | 22.7% | 5 |
| Information industry (last job) | 28.7 | 26.3% | 7 |
Weeks as published. The 27-week share is computed from published counts. Base fallback = mean weeks divided by 52/12, rounded up to a whole month (an editorial rule).
Outside the United States the tail is heavier. Eurostat measures duration as the time since a job search began, or since the last job if that is shorter.
Table 4. Share of the unemployed out of work about 6 months or longer and 12 months or longer, 2025 (CEOtudent analysis of Eurostat lfsa_ugad and une_ltu_a, and BLS CPS Table 30)
| Area | About 6 months or longer | 12 months or longer |
|---|---|---|
| United States (27 / 52 weeks) | 23.2% | 13.8% |
| EU27 | 48.0% | 31.5% |
| Germany | 45.7% | 27.9% |
| France | 40.1% | 23.0% |
| Spain | 47.4% | 32.1% |
| Italy | 64.9% | 50.3% |
EU shares computed from published counts; the 12-month shares match Eurostat’s published long-term shares. Eurostat flags France and Spain as “definition differs”, and 27 weeks is about 6.2 months: read the US-EU gap as large, not precise.
Nearly half of the EU’s unemployed in 2025 had been unemployed for six months or more. Among member states the 12-month share ran from 13.7 percent in the Netherlands to 63.4 percent in Slovakia; in Türkiye it was 18.2 percent. Use your own country’s data.
What is the Runway Equation?
The Runway Equation, a CEOtudent editorial framework, replaces “months of expenses” with three blocks of money: the ramp, the fallback and a margin for being wrong. Monthly inputs: B, personal burn at the level you will actually live on; C, business costs; H, replaced benefits such as health cover and a retirement contribution; tau, the tax reserve rate on business profit; and R(t), revenue in month t after quitting, with R0 the side revenue at the quit date.
- Gap: gap(t) = B + C + H + tau x max(R(t) – C, 0) – R(t).
- Break-even revenue: BE = C + (B + H) / (1 – tau), the revenue at which the gap is zero.
- Straight-line ramp over N months: R(t) = R0 + (BE – R0) x t / N.
- Ramp cost: G0 x (N – 1) / 2, where G0 = (1 – tau) x (BE – R0) is the gap at the quit date. This is exact when R0 is at least C. We call it the triangle rule: the ramp costs half the starting gap multiplied by one less than the number of ramp months.
- Fallback reserve: F x GF, where GF = B + C + H is the gap at zero revenue and F is the fallback period in months.
- Required savings: S = (ramp cost + fallback reserve) x (1 + m), with m the safety margin.
- Quit rule: quit only when liquid savings are at least S, when R0 and the ramp slope come from paid side revenue rather than a forecast, and when the fallback reserve sits in a separate account with a written trigger.
Two US anchors help. For H, the KFF 2025 Employer Health Benefits Survey puts the average premium for single coverage at $9,325 a year, about $777 a month, of which covered workers paid $1,440. That payroll deduction never appeared in a burn measured from take-home pay, so the whole replacement premium is a new line; real quotes vary by age, location and plan. For tau, self-employment tax alone is 12.4 percent plus 2.9 percent on 92.35 percent of net earnings, about 14.1 percent before income tax; set the full rate with an adviser. The salary-to-freelance-rate calculation prices the rest of the employer package.
What does the math look like month by month?
Every input below is an illustrative assumption, in US dollars because the benefits anchor is American; the method works in any currency.
Table 5. Worked example inputs
| Input | Value |
|---|---|
| B, personal burn | $4,500 a month (illustrative assumption) |
| C, business costs | $400 a month (illustrative assumption) |
| H, replaced benefits | $900 a month (illustrative assumption) |
| tau, tax reserve | 25% of revenue minus business costs (illustrative assumption) |
| R0, side revenue at quit | $1,600 a month (illustrative assumption) |
| N, ramp | 24 months, straight line (illustrative assumption) |
| F, fallback | 6 months, the all-unemployed base fallback in Table 3 |
| m, margin | 20% (illustrative assumption) |
So BE = 400 + (4,500 + 900) / 0.75 = $7,600 a month. Revenue rises (7,600 – 1,600) / 24 = $250 a month. The gap at the quit date is G0 = 0.75 x 6,000 = $4,500, and the gap at zero revenue is GF = $5,800.
Table 6. The 24-month ramp (CEOtudent calculation on illustrative assumptions)
| Month | Revenue | Gap |
|---|---|---|
| 1 | 1,850 | 4,312.50 |
| 2 | 2,100 | 4,125.00 |
| 3 | 2,350 | 3,937.50 |
| 4 | 2,600 | 3,750.00 |
| 5 | 2,850 | 3,562.50 |
| 6 | 3,100 | 3,375.00 |
| 7 | 3,350 | 3,187.50 |
| 8 | 3,600 | 3,000.00 |
| 9 | 3,850 | 2,812.50 |
| 10 | 4,100 | 2,625.00 |
| 11 | 4,350 | 2,437.50 |
| 12 | 4,600 | 2,250.00 |
| 13 | 4,850 | 2,062.50 |
| 14 | 5,100 | 1,875.00 |
| 15 | 5,350 | 1,687.50 |
| 16 | 5,600 | 1,500.00 |
| 17 | 5,850 | 1,312.50 |
| 18 | 6,100 | 1,125.00 |
| 19 | 6,350 | 937.50 |
| 20 | 6,600 | 750.00 |
| 21 | 6,850 | 562.50 |
| 22 | 7,100 | 375.00 |
| 23 | 7,350 | 187.50 |
| 24 | 7,600 | 0.00 |
Gap = B + C + H + 25% of (revenue – C) – revenue. Month 1: 4,500 + 400 + 900 + 0.25 x (1,850 – 400) – 1,850 = $4,312.50.
The totals: a ramp cost of $51,750 (exactly 4,500 x 23 / 2), a fallback reserve of 6 x 5,800 = $34,800 and a margin of 20 percent of $86,550, or $17,310. Required savings: $103,860, or 23.1 months of personal burn.
The six-month rule would hold 6 x 4,500 = $27,000, which this ramp exhausts in month 8 with nothing left for a fallback. The funded plan reaches break-even still holding $52,110, the fallback reserve plus the margin. Note that the gap falls by $187.50 a month, not $250, because a quarter of each extra dollar of profit goes to tax, and that $39,375 of the ramp cost, 76.1 percent, is spent in the first 12 months.
How much does the answer move with ramp speed and fallback length?
Table 7 keeps the Table 5 inputs and varies only the two unknowns. Fallback columns: 3 months is the median search rounded up, 6 months the mean rounded up, 12 months the 52-week tail.
Table 7. Required savings, with months of personal burn in brackets (CEOtudent analysis; margin 20%)
| Ramp to break-even | Ramp cost | Fallback 3 months | Fallback 6 months | Fallback 12 months |
|---|---|---|---|---|
| 12 months | $24,750 | $50,580 (11.2) | $71,460 (15.9) | $113,220 (25.2) |
| 18 months | $38,250 | $66,780 (14.8) | $87,660 (19.5) | $129,420 (28.8) |
| 24 months | $51,750 | $82,980 (18.4) | $103,860 (23.1) | $145,620 (32.4) |
| 36 months | $78,750 | $115,380 (25.6) | $136,260 (30.3) | $178,020 (39.6) |
Required savings = (ramp cost + fallback months x $5,800) x 1.2; months of burn = required savings / $4,500.
No cell is near six months. The best corner, a 12-month ramp with a 3-month fallback, needs 11.2 months of burn. The “more than 6 months” band in the freelance-readiness scorecard is a floor, not a target.
An exit month costs more than a ramp month. Each extra ramp month adds half the starting gap, $2,250 before margin; each extra fallback month adds the full $5,800.
Shape matters as much as length. If revenue stays flat early and arrives late (rising with the square of time over the same 24 months), the ramp costs $69,718.75 instead of $51,750, 34.7 percent more, and required savings rise to $125,422.50, or 27.9 months of burn.
Side income before quitting is therefore the cheapest runway. Holding the $250 monthly slope and a six-month fallback, quitting with no paid side revenue means a 31-month ramp and required savings of $142,380 (31.6 months of burn). Quitting with $3,200 a month means an 18-month ramp and $74,655 (16.6 months): $67,725 less, and a slope that was measured rather than guessed. The one-person revenue stack maps which streams can start while you are still employed.
When should you pull the fallback trigger?
A runway is a forecast, and forecasts are wrong. The CEOtudent fallback rule (editorial framework; every threshold is an editorial parameter) has three parts.
- Quarterly checkpoints. At months 3, 6, 9 and 12, record monthly revenue and the slope since the quit date.
- Re-forecast, do not compare. Recompute the remaining requirement with the measured slope: (remaining ramp cost + fallback reserve) x (1 + m). If it exceeds the current balance, you have 30 days to make a change that closes the gap in the re-forecast (lower burn, higher prices, a signed retainer) or you start a full-time job search.
- A hard floor. If the balance falls to the fallback reserve while the gap is still positive, the search starts that week. No re-forecast overrides the floor.
Suppose the example business grows by $125 a month instead of $250 (an illustrative assumption). At the month 6 checkpoint, revenue is $2,350 against a plan of $3,100, or 75.8 percent of plan, and the balance of $78,828.75 is only $1,968.75 behind plan.
The re-forecast is far worse. At $125 a month, break-even is 42 months away. The remaining ramp costs $80,718.75; with the $34,800 fallback reserve the requirement is $115,518.75, or $138,622.50 with the margin. The plan is unfunded by $59,793.75 and the trigger fires at month 6. Carrying on instead, revenue would reach only $3,975 by month 19, 52.3 percent of break-even, and month 20 would break the floor.
The balance lags; the slope leads. This is the logic of kill criteria: fix the signal, threshold, date and action before you start, and give the trigger to someone who is not the most invested person in the room.
The runway worksheet and pre-quit checklist
Table 8. Runway worksheet (CEOtudent editorial framework; ranges are editorial parameters)
| Line | Item | Formula or source | Example |
|---|---|---|---|
| 1 | Personal burn, B | Higher of trailing 3 months and planned budget | 4,500 |
| 2 | Business costs, C | Written quotes, committed costs only | 400 |
| 3 | Replaced benefits, H | Real quotes for cover you will buy | 900 |
| 4 | Tax reserve rate, tau | Local adviser | 25% |
| 5 | Side revenue now, R0 | Average of last three paid months | 1,600 |
| 6 | Monthly increase, s | Measured side trend, last 6 months | 250 |
| 7 | Break-even revenue, BE | Line 2 + (line 1 + line 3) / (1 – line 4) | 7,600 |
| 8 | Ramp months, N | (Line 7 – line 5) / line 6, rounded up | 24 |
| 9 | Gap at quit, G0 | (1 – line 4) x (line 7 – line 5) | 4,500 |
| 10 | Ramp cost | Line 9 x (line 8 – 1) / 2 | 51,750 |
| 11 | Gap at zero revenue, GF | Line 1 + line 2 + line 3 | 5,800 |
| 12 | Fallback months, F | Table 3 or national data; stress test 12 | 6 |
| 13 | Fallback reserve | Line 11 x line 12 | 34,800 |
| 14 | Margin | 15% to 25% of (line 10 + line 13) | 17,310 |
| 15 | Required savings, S | Line 10 + line 13 + line 14 | 103,860 |
| 16 | Months of burn | Line 15 / line 1 | 23.1 |
| 17 | Liquid savings today | Reachable within a week | |
| 18 | Quit test | Is line 17 at least line 15? |
Line 10 is exact for a straight-line ramp when side revenue covers business costs; for a lumpier path, add the monthly gaps as in Table 6.
Before resigning, all of these should be true:
- Line 18 passes.
- Lines 5 and 6 come from at least six months of paid side revenue.
- Line 8 is consistent with Table 1, or you can say in one sentence why your business differs.
- The fallback reserve sits in a separate account the ramp never touches.
- The tax reserve has its own account; mixing it with operating cash is one of the classic solopreneur money mistakes.
- Health cover is quoted and starts the day employer cover ends.
- The fallback trigger is written: checkpoint dates, the re-forecast rule, the floor, and the person who will hold you to it.
- Your CV and network are current, so a search could start within a week.
If the numbers pass and you still hesitate, the regret minimization framework helps separate fear from information.
What this does not mean
It does not predict your ramp. Table 1 describes groups of firms, not you.
It does not say the fallback takes six months. BLS figures describe searches still in progress and do not separate people returning from self-employment.
It is not country-neutral. In the EU, duration tails are longer (Table 4), health cover is often public and self-employment taxes differ. Replace each line with your country’s figures.
It is not advice. This is general information, not financial, tax or legal advice; the ranges are editorial choices, and a qualified adviser should set the tax reserve.
The CEO and the student in this decision
The CEO half is owning the number. A company funds a launch to break-even, holds a reserve for failure and names the review date. Set your quit date when line 18 passes, not on a good Monday or a bad Friday.
The student half is treating the ramp as a hypothesis: line 6 is your best current estimate, and every quarter gives a better one. Write the trigger while calm, re-forecast with real data, and act when the plan is unfunded. A trigger pulled on schedule is a good decision even if one more quarter would have worked.
Frequently asked questions
How many months of runway do you need before quitting your job?
Enough for the sum of the monthly gaps until break-even, plus a fallback, plus a margin. In the worked example that is 23.1 months of personal spending; across Table 7 it ranges from 11.2 to 39.6 months.
Is six months of expenses ever enough?
Only when the ramp is short and the fallback fast. If line 16 of the worksheet is six or less, it is enough.
Should the fallback use the median or the mean unemployment duration?
The mean, rounded up, as the base case, and 12 months as the stress test. In September 2026 the US median was 11.5 weeks and the mean 24.8 weeks.
What if the required amount is out of reach?
Change the inputs, not the rule: build side revenue first, cut burn, or choose a business with faster first revenue.
How often should the runway be recomputed after quitting?
Re-forecast quarterly. In the month 6 example the balance was $1,968.75 behind plan while the plan was unfunded by $59,793.75.
Sources
- U.S. Bureau of Labor Statistics. The Employment Situation, September 2026 (USDL-26-1549), Table A-12.
- U.S. Bureau of Labor Statistics. CPS annual averages 2025, Table 30.
- U.S. Bureau of Labor Statistics. CPS annual averages 2025, Table 31.
- U.S. Bureau of Labor Statistics. CPS annual averages 2025, Table 32.
- U.S. Bureau of Labor Statistics. CPS concepts and definitions.
- Eurostat. Long-term unemployment by sex, annual data (une_ltu_a), updated 10 September 2026.
- Eurostat. Unemployed persons by duration of unemployment (lfsa_ugad), updated 10 September 2026.
- Eurostat. Statistics Explained glossary: Long-term unemployment.
- Board of Governors of the Federal Reserve System. Economic Well-Being of U.S. Households in 2025, May 2026.
- Federal Reserve Banks. 2026 Firms in Focus: Chartbook on Firms by Age of Business (2025 Small Business Credit Survey).
- Federal Reserve Banks. 2024 Report on Startup Firms (2023 Small Business Credit Survey), December 2024.
- KFF. 2025 Employer Health Benefits Survey, Summary of Findings.
- Internal Revenue Service. Topic No. 554, Self-Employment Tax.
Tables 1 and 2 reproduce published figures (Table 1 read from the survey charts). Tables 3 and 4 are CEOtudent analyses of the BLS and Eurostat data named under them, recomputed by script. The Runway Equation, the triangle rule, the fallback rule and Table 8 are CEOtudent editorial frameworks; Tables 5 to 7 and the month 6 checkpoint are CEOtudent calculations on illustrative assumptions. The “three to six months of expenses” rule was not used as evidence because no primary source derives it, and business survival tables from sibling articles were not repeated.
This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.
This post is also available in:
