What Does an 85% Chance of Success Actually Mean?
Open the Monte Carlo retirement calculator and the first number many people freeze on is the success rate. The live landing shows an example readout of Success Rate 85% — labeled as an example on the page, not a score for your plan, and not a verified result for any specific portfolio below. That percentage is also the number people treat like a weather forecast or a guarantee. It is neither.
An 85% chance of success, in this tool, is a count of model paths. It answers: under the assumptions you typed, what share of the simulated timelines still had money at the end? It does not answer: will the real future cooperate 85% of the time?
This article defines that percentage the way the product defines it, contrasts a fixed 7% line with a fan of paths, explains why planners often talk about roughly 80–90% instead of 100%, and walks a clear hypothetical you can re-run yourself. Hypothetical only. Not advice. Not a full replacement for a paid financial plan.
Success Rate = Share of Paths Still Holding Money
On this site, probability of success (the success rate) is the percentage of simulations that ended with money remaining — still have money at your target date.
If you run 1,000 paths and 850 of them finish with a positive balance, that is an 85% success rate in this model, with these inputs. The other 150 paths hit $0 (or otherwise fail the "money remaining" test) before or by year N.
That is a simple count. It is not:
- A promise that markets will deliver an 85% outcome for you
- A historical survival rate from past U.S. retirement start years
- A tax-, Social Security–, or Medicare-aware plan score
- Insurance against sequence risk in the real world
Change spending, expected return, volatility assumptions, or the timeline, and the percentage moves. That movement is the point of the tool.
Fixed 7% Line vs a Fan of Paths
A traditional projection applies one return every year. Enter 7%, and you get 7%, then 7%, then 7%. The chart is a single line. Easy to read. Tidy in a way real years are not.
Returns bounce. Order matters when you are withdrawing. A bad stretch early, while you are selling to fund spending, hurts more than the same average delivered smoothly. One average hides that. (The crash-year walkthrough is in Sequence of Returns: Why a Crash in Year 1 Hits Harder Than Year 5 or Year 10.)
Monte Carlo on this site replaces the single line with a fan of paths:
- Parametric, not historical. Random annual returns are drawn from a Gaussian (normal) distribution centered on the expected return you enter, using Box-Muller to generate the draws. Paths are not pulled from a historical market tape.
- 100 to 1,000 runs — not "thousands" as a marketing word, and not an unlimited cloud. Max is 1,000.
- Percentile bands so you can see spread, not just a success headline. The free landing widget surfaces the 10th, 50th, and 90th percentiles. The full projector can show a fuller set (10th, 25th, 50th, 75th, 90th). Either way, bands are outcomes in this set of runs, not a floor on reality.
If you want FIRECalc- or cFIREsim-style replay of real past sequences ("what if I retired in 1966?"), that is a different method. We do not do historical backtests. The method split is in Parametric vs Historical Monte Carlo.
Correcting a common reading of Monte Carlo pages: "hundreds or thousands of scenarios based on historical market behavior" sounds right in casual speech and is often wrong for this product. We are parametric around your expected return, capped at 1,000 runs. History is not walked through your plan year by year.
Why ~80–90% Shows Up So Often — and Why 100% Is Usually Too Stingy
People often want a 100% success rate. That impulse is understandable. Running out of money is the failure mode the chart is counting.
Advisors and planners frequently have to explain that chasing 100% can be overly conservative — especially when the Monte Carlo assumes no flexibility: spending stays fixed, allocation stays fixed, goals do not bend when markets are rough. Real households can cut spending, delay a big purchase, work a bit longer, or change the mix. A model that never allows that will punish plans that would have been fine with course-correction.
Industry practice is not one sacred number. Utilized success-rate thresholds often fall somewhere between about 85% and 95%, and many advisors use a fixed threshold near ~90%. Our own FAQ already says it in plain language: most planners consider 80–90% success rates acceptable. We align with that range as a common planning conversation — not as a new industry survey we invented, and not as a rule you must hit.
Aiming for 100% can force very low withdrawal rates so the plan survives even the worst modeled paths. In most of those same runs, that means large leftover wealth you never spent — an upfront spending cut that may never have been needed if adjustment were allowed. The success rate went up by shrinking life in the median path to protect the left tail of a rigid model.
That framing — clients wanting 100%, advisors explaining over-conservatism when flexibility is ignored, common thresholds in the mid-80s to mid-90s, and 100% leaving unspent wealth in most scenarios — tracks the discussion in a Kitces.com guest post by James Yaworski on refining Monte Carlo thresholds with volatility tolerance: Using Volatility Tolerance To Refine Monte Carlo Thresholds. We are not Kitces. We are paraphrasing publicly discussed planning points, not quoting the article as our product doctrine.
A high success rate on a rigid model is not the same as a good life under a flexible one. A middling success rate with room to adjust is not the same as "the plan failed." Read the percentage as a stress gauge, not a moral score.
A Worked Example: Same $1M, Two Spending Levels
The landing's 85% is still an example UI readout. The table below is a different pair: two runs we recorded on the live free widget with the inputs shown. Monte Carlo is random — click Run again and nearby numbers can shift. Treat this as one recorded illustration, not a permanent product constant.
Shared inputs (both runs)
| Input | Value |
|---|---|
| Starting portfolio | $1,000,000 |
| Horizon | 30 years |
| Expected return | 7% |
| Volatility (std. dev.) | 15% |
| Simulations | 500 |
Setup A — 4% starting withdrawal — Annual spending $40,000 → starting withdrawal 4%.
Setup B — higher spend — Annual spending $55,000 → starting withdrawal ~5.5%.
| Run | Annual spending | Starting withdrawal | Success rate (recorded) | 10th %ile | Median | 90th %ile |
|---|---|---|---|---|---|---|
| A | $40,000 | 4% | 93% | $152K | $2.87M | $10.23M |
| B | $55,000 | ~5.5% | 73% | $0 | $1.08M | $6.31M |
Success moved from 93% to 73% when spending rose — same model, same 500 Gaussian paths around 7%, only the withdrawal changed. The 10th percentile also flipped from a leftover balance to $0: more of the left tail ran out.
Why higher spending lowers the share of paths that still have money: every year you take more dollars out. Bad-return years force larger share sales at low prices. Recovery has less capital left to compound. Across the fan, more paths hit $0 (or fail the money-remaining test) by year 30. The success rate is literally counting those endings. That is arithmetic under volatility — not a verdict that 4% is "safe" or 5.5% is "unsafe" for you, and not a claim your next click will reprint 93% / 73% exactly.
Re-run both yourself in the free Monte Carlo widget. Change one knob at a time.
For the starting-rate conversation on longer FIRE horizons, see 4% vs 3.5% SWR for Early Retirees. For runway language ("how long does this pile last?"), see How Long Will My Money Last?.
Optional next step after the widget: open a fuller plan in the projector. The Dual-Income FIRE sample is a savings / accumulation sample aimed at a $1.5M goal — not a retired withdrawal plan, and not the $1M / $40k example above. Use it when you want accounts, contributions, and a goal on a timeline; do not misread it as the worked retirement case.
Common Mistakes
Mistake 1: Treating 85% as a promise about your life
The landing example says Success Rate 85%. Your run might say something else. Either way, the number is the share of simulated paths that still had money under your inputs. It is not a forecast probability issued by the market.
Mistake 2: Chasing 100% by default
A 100% rigid-model score often means you cut spending (or assumed returns) until even the worst paths in the fan survive. That can leave a pile of unspent money in most scenarios. If you can adjust spending or work when markets are ugly, the model's "failure" paths are not all real-life failures. See the Kitces-linked framing above; do not treat 100% as the only responsible target.
Mistake 3: Reading our paths as historical market behavior
We draw Gaussian returns around the expected return you type (Box-Muller), 100–1,000 runs. We do not replay 1929 or 2008 as sequences. If that is the question, use a historical tool. Method notes: Parametric vs Historical Monte Carlo.
Mistake 4: Ignoring the expected return you typed
Parametric Monte Carlo wraps randomness around your assumption. Raise the expected return and success rates usually rise because you moved the center of the bell curve — not because the market promised that rate. Run a base case and a more conservative return.
Mistake 5: Looking only at the success headline
A plan can show a middling success rate with a terrible 10th percentile, or a high success rate with a skinny lifestyle. Read the percentile bands (widget: 10 / 50 / 90; full app can add 25 / 75) and the spending level that produced them.
Mistake 6: Confusing this free stress test with a full paid plan
We do not model taxes, Roth conversions, Social Security claiming, or Medicare. We are not a full replacement for paid planners. Use the success rate as one honest stress gauge among others.
Frequently Asked Questions
It means that 85% of the simulated paths — in this model, with these inputs — ended with money remaining at your target date. It is a count of model paths, not a promise about the real future. Change spending, expected return, volatility, or the timeline, and the percentage moves.
Not necessarily. Chasing 100% can force very low withdrawal rates to survive even the worst modeled paths. In most of those same runs, that means large leftover wealth you never spent. Most planners consider 80–90% success rates acceptable when the model assumes no spending flexibility.
No. This site uses parametric Monte Carlo: Gaussian (normal) random returns drawn around the expected return you enter, using Box-Muller to generate the draws. Paths are not pulled from a historical market tape. If you want FIRECalc- or cFIREsim-style replay of real past sequences, that is a different method. The difference is explained in Parametric vs Historical Monte Carlo.
Because most Monte Carlo models assume no flexibility — spending stays fixed, allocation stays fixed, goals do not bend when markets are rough. Real households can cut spending, delay purchases, work longer, or change the mix. A model that never allows adjustment will punish plans that would have been fine with course-correction. A middling success rate with room to adjust is not the same as "the plan failed."
A fixed projection applies one return every year — 7%, then 7%, then 7%. The chart is a single line. Monte Carlo replaces that with a fan of paths: random annual returns drawn from a distribution around your expected return. You see a range of outcomes and a success rate, not just one line.
100 to 1,000 runs — not "thousands" as a marketing word, and not an unlimited cloud. Max is 1,000. Each run is a full timeline of Gaussian-drawn returns.
The Bottom Line
An 85% success rate is a count: 85% of the simulated paths, under the inputs you typed, still had money at your target date. The other 15% ran out.
- It is not a promise. The model does not know the real future. It knows what you entered.
- It is not historical. We draw Gaussian paths around your expected return, not FIRECalc-style replays of past markets.
- 100% is usually too stingy. A rigid model that assumes no spending flexibility will punish plans that would have adjusted. Most planners consider 80–90% acceptable.
- The percentage moves when inputs move. Raise spending, and more paths run out. Lower expected return, and more paths run out. That is the point — see how sensitive the plan is.
Run the Monte Carlo calculator with your numbers. Change one knob at a time. Read the success rate as a stress gauge, not a verdict. Then, if you need a crash-year test, open the Sequence of Returns Calculator with the same plan. The two views are complements, not substitutes.
Check Your Numbers
Run a parametric Monte Carlo with your starting portfolio, spending, and expected return. See the success rate and percentile bands. Free, no signup.
Run the Monte Carlo CalculatorOr open the main app and model the same plan there. Data stays in the browser by default.
Disclaimer: This article is for informational purposes only and does not constitute financial advice. The projections and examples discussed are hypothetical and based on general assumptions. Investment returns are not guaranteed, and past performance does not predict future results. Consult a qualified financial advisor for personalized guidance based on your specific situation.