Compute the success of various withdrawal rates with different portfolios.

## How does this work?

The FIRE calculator is calculating the success rate of a withdrawal rate for a given situation. The calculator will try a retirement simulation starting at every possible month in the given period between the start and end year. Here is what happens during the simulation:

- You start with a certain amount of money
- Your portfolio value is adapted based on the returns each month
- Money is withdrawn each month from your portfolio
- The withdrawal is based on the withdrawal rate and the initial portfolio.
- Your withdrawal is adjusted for inflation every month

For instance, if you simulate between 1950 and 2000 with 20 years of retirement, the calculator will try to a retirement simulation starting from every month between 1950 and 1980.

The data being used is the same data as the data from the Trinity Study. I am using U.S. Stocks and Bonds.

## How to increase my retirement chances?

There are several ways you can **increase your chance of success in retirement**.

First, you can choose a better portfolio for your case. Ideally, you want a large allocation to stocks if your risk tolerance permits it. This will allow your portfolio to grow more. Then, you want broad index funds with very low fees.

Another way is to choose a better withdrawal rate. Simply said, a lower withdrawal rate will increase your chances of success. This means lower returns will not hurt your portfolio too much.

## Using this FIRE calculator

First, choose how much money you have in your portfolio (the initial value). And, choose the period you want to consider for simulation by choosing the start and end year.

Then, choose how many years you want to simulate retirement. And choose how much of your initial portfolio you want to withdraw each year (your withdrawal rate).

Finally, choose your investment portfolio. Once you are done, click Simulate. After a short moment, the result should be presented to you.

Excellent work, thank you so much! I’ve discovered that you using monthly instead of annual data makes your calculator much more precise in terms of worst durations:

Example:

In cFIREsim (annual data), the worst duration of a 100% stock portfolio would be 24 years (starting year: 1966 and 1969).

In thepoorswiss (monthly data), the worst duration of a 100% stock portfolio would be 15 years (starting year/month: 8/1929)

It would be absolutely epic if your FIRE calc would allow for glide path testing! How can we seduce you into implementing this :)?

Thanks :)

Using monthly data makes it more precise on all counts indeed. Ideally, I would use daily data if I had all the data available :)

I am very surprised that something as well-known as cfiresim does not use monthly data.

I try to keep this calculator simple to let everyone use it. The calculator with cfiresim is bloated with features. I thought about doing an advanced version of my calculator with more options. Maybe I should push this up on my todo list.

As for glide path, I would need to research that further and write about it before I integrate it in the calculator. I will also push this up on my todo list.

Thanks, very much appreciated! Actually, I don’t know of any other FIRE calc using monthly data, so you’re really quite unique :)!

Regarding glide paths: It’s a very interesting subject imho, because Kitces/Pfau and BigERN say glide paths matter. But many claim just the opposite: backtesting portfolios with and without glide paths gets you the same success rates on cFIREsim. But: Regarding worst durations of portfolios, glide paths seem to perform a little better. I still don’t know what to believe, and having monthly data to verify or debunk the glide path concept would be immensely helpful. However, of course I understand that you might have some more important priorities in life :)!

Usually, I would trust Big ERN over most other people because I trust his results and his data. And Big ERN does his simulations with monthly data.

But this is something interesting, so this will definitely be covered on this blog, although I can’t say when.

This is a great calculator, thank you!

I was wondering how people factor in Social Security? For instance, if I am working with a 45 year old couple who will not collect SS until age 70, what are some best practices for including this?

Thank you!

Hi laura,

That’s a very good question. And I do not have a great answer yet! It’s difficult to factor this in. If you are close to getting social security, then it’s easy to simply remove it from your expenses.

But if you are far from it, it’s more difficult since you do not know exactly how much you will receive and when (it could change).

Personally, I currently entirely ignore it from my calculations. It’s definitely more conservative (likely too much), but it’s safer.

I will try to write articles on that!

Thanks for your reply. I’ll look forward to the articles!

Hi is the median terminal value in the USD at the end of the time interval or at the start.

I mean is it the initial buy power or the terminal buy power?

Hi,

I am not sure I understand, the terminal value is always at the end. It’s the buying power at the end, but the buying power itself is not adjusted for inflation. So, it’s the value at the end, corresponding to the buying power today. But all the withdrawals would be adjusted for inflation.

Super tool! But, maintaining 101% (for example) of the initial portfolio value as the terminal value (defined here as “success”) means that in real terms you have lost a lot of purchasing power over the retirement period? It seems like most people are interested in preserving the buying power of their initial pot of money at retirement when they think of a “perpetual” withdrawal rate. So shouldn’t “success” be defined as 100% of the REAL (escalated by inflation) value of the initial amount?

In many cases, you end up with a significant amount of money at the end of the simulation.

But success in my tool is defined as having more than 0 USD at the end of the simulation. So, if you have 1 cent at the end, that’s a success.

I have never done simulations where I would preserve the capital since I find this too conservative.

Fair enough and thanks for the quick response! For many people, handing down a specific inheritance is important, and they want to know “can I live off the real returns.” But for most, as you say, that’s a nice surprise rather than a requirement.

Yes, it all depends on whether you want generational wealth or not!

Hey, this calculator is fantastic! Terrific work. I’m wondering if it would be at all possible to somehow add a max drawdown that one would experience, as well as having a CAGR input (could be one for stocks and one for bonds or just one total). This could allow one to see the max drawdown, and see what CAGR one would need to obtain certain results in various scenarios. I also think somehow seeing a historical path value of account could be interesting (maybe just for best case, worst case, and median/average). Also was wondering if this was all done through c++?

Hi Tony,

Would you like the max drawdown as the output or the input?

I don’t see as CAGR would make sense as an input. It’s too simple, it would not work well. My simulations take into account monthly returns, including drops. CAGR would only go up and it would very easy to get success rate.

I could try to output some graphs, that’s a good idea!

The backend is entirely in C++ and the frontend is just a very simple form calling the backend.

Hey thanks for getting back to me! I agree graphs would be awesome. I would like to see max drawdown/ maybe even standard deviation experienced as an output.

Thanks so much again!

It’s on my list for a future update of the tool, but it may take a long time to come, I have very little time for the blog these days.

Hi,

Congratulations for the excellent content!

Would it be possible to support Swiss Franks and Swiss inflation? Same US portfolios.

For someone that intents to retire in Switzerland it would make more sense.

I understand this will not be easy, especially predicting the CHF/USD exchange rate.

Maybe add a parameter for 2-3 scenarios as you analyzed below

https://thepoorswiss.com/4-percent-rule-swiss-stocks/

e.g. Scenario1–> historical exchange rate

Scenario2 –> 1-1

Schnario3 –> the “reasonable” assumption you made in the above article

You can also take some ideas from the study below:

https://ideas.repec.org/p/ngi/dpaper/10-12.html

The outcome of the “4% rule” (or any other rate) differs significantly between countries. Though I think this analysis focuses on domestic markets.

Hi huskarl,

That’s a good idea. I have a plan to make an advanced calculator in the future, I could definitely include Swiss francs in the mix.

I can’t give you an idea of when that will be, I have very little time these days and need to complete the French translation first.

What would be really great is to be able to adjust the WR to reflect various life stages. For example – an initial WR of say 4.5% for 10 years then reducing to 3.5% (when I can access my state pension) and then after another 10 years, then reducing to say 3% for 10 years as I’ll be old and less active.

Hi David,

That’s indeed interesting. However, starting to add such advanced options to the calculator would make it complicated. And I don’t really how to integrate in the interface part to make it user-friendly.

Hi TPS,

There is a bug in your WTV result.

When setting the withdrawal rate at 0, the WTV doesn’t change and is fixed at 2,435.00 when starting in 5/1902 even if the Initial Value changes.

Thanks,

I never thought of testing with WR 0 :)

Thanks for letting me know, I will fix that!

What do “The best terminal value” and “The worst terminal value” represent?

How can the best be better than the median?

Hi Don,

The best terminal value is the maximum terminal value, the highest amount of money you can end up at the end of the retirement time.

The worst terminal value is the minimum terminal value, the lowest amount of money you can end up at the end of the retirement time.

The median is always between the maximum and the minimum.

Hello TPS

first of all, thanks for all your great work! Not least the calculators are very cool to play with.

What I noticed though: with best terminal value (BTV), there seems to be three digits missing. When I change the input, I get very different median and average TV’s, but BTV is a five or six-figure number even if I reduce the year from 40 to 20… So I see the same as Don.

Kind regards – and keep up the good work

Tom

Hi Tom,

I have looked into it and have not found an issue except that the numbers after the command seem to always be zero. But the best (maximum) terminal is expected to be much higher than than the median terminal value. This is just because timing the market is very effective but can’t be done in the future.

Best Terminal Value: 892,393.00 when starting in 5/1940

Worst Terminal Value: 10,250.00 when starting in 6/1872

Median Terminal Value: 116,121,465.00 USD

Average Terminal Value: 213,020,550.00 USD

This is the result of some test I ran in the calculator. I get the same issue as the others – if worst is 10K and best is 892K, how can median be 116M and average be 213M?

Hi Dan,

Thanks for reporting that, that’s indeed impossible :)

I am pretty sure that the median and average are multiplied by 1000 but they should not. I will look into that when I get a moment.

What did you use for fund fees for the 100% stock portfolio ? Thank you

Hi Jr,

The fees are 0.1% TER for each portfolio.

I may do an advanced calculator version in the future that would let the user configure the fees. Do you think that would be useful?

That would be great!

Hello,

May I know what is the inflation rate you are using pls?

Thks

Hi,

These are historical inflation rates from the United States (CPI).