In the last post “Is 2 Percent the New Safe Retirement Withdrawal Rate?”, we explored how some experts have knocked the conventional 4% retirement withdrawal rate in favor of figures as high as 7% and low as 2%. In conclusion, I demonstrated using a simple linear example how I thought a 3% withdrawal rate with an assumption of 6% growth was my personal preference. My reasoning was because it offered a high probability for steady returns and seemed to maintain portfolio integrity even after adjustment for inflation.

So what’s the problem?

• Rule of thumb calculations are okay for estimations. But in real life, they hardly ever play out the way they do on paper. Think about it. In our previous example, the portfolio always grew year over year by 6%. But portfolios don’t go up and up forever. And when they go down, they can have a bigger impact on your portfolio than you’ll expect.

Consider the following scenario:

• Year 0: You start your retirement with $1,000,000 and withdraw 3%. So you take out $30,000 and are left with $970,000. Your brilliant model predicts that at the end of Year 1 you’re money will grow by 6% and you’ll have $1,028,000.

• Year 1: Oh no! The market tanked by negative 10%! Now your balance drops to $873,000. You withdraw another 3% of $26,190. Now your balance is all the way down to $846,810. That’s over $150,000 lost in just the first year! At this rate you’ll be broke in less than 10 years. How’s that 6% model treating you now?

So if you can’t generalize some constant rate of return to keep your proverbial cookie jar from running out of cookies, then what are you supposed to do to predict how long your retirement funds will last?

**A Better Approach:**

The truth is that your guess is just as good as anyone else’s when it comes to what the market will do next year, the year after that, and so on. In reality, you could take my example from earlier and put in any random number for the return rate each year.

So how can we build this into our model and find a way to generate random numbers automatically?

**Using a Monte Carlo Simulation:**

This is where a thing called a Monte Carlo simulation comes in. A Monte Carlo simulation is when you use random sampling to run a computer model to test your theory.

That’s exactly what we plan to do – test our theory about whether or not we’ll run out of money during retirement.

To begin, download my Microsoft Excel file: My Money Design Monte Carlo Worksheet Rev1

The picture above is an example of just one Monte Carlo simulation from my Excel worksheet (and because it’s made of our random numbers, you’ll never see it look like this ever again). As you see, it shows you how your money will do over the next 50 years. What’s our goal?

• Don’t go below $0 and run out of money!

The beauty of this experiment is that you can do it hundreds of times to try out an infinite combination of results. Every time you click “F9”, the random numbers update and the graph will change. Download it and try this out for yourself! It’s a lot of fun.

**How Does It Work?**

• You can enter any Starting Balance you want. I have it set to $1,000,000 to start with.

• You can enter any Withdrawal Rate you want. I have it set to my ideal 3%, but feel free to experiment with higher percentages. I think you’ll find that you’ll go below $0 more often the higher you set that number.

• Inflation has been set to 3%. You can change this if you want, but this is a pretty accepted figure.

• The random numbers in the “Return Rate” column come from the Return Rate Mean and Sigma variables based on the empirical rule (if you’re dying to know what that means, click here to read about it on Wikipedia). These figures come from real S&P 500 data between 1950 and 2012 under the Reference tab within the same Excel worksheet. If you want to change these, go ahead.

**Adjusted for Inflation – For Your Convenience:**

We will want to increase our withdrawal each year to account for the rate of inflation. This has already been built into the equations and model.

**How Did You Do?**

So after downloading my Excel file and trying it out multiple times, how did you do? Did any of these trials lead to a negative balance? What if you raise the withdrawal rate? At what point do you stop feeling safe about how much to withdraw?

**Related Posts:**

1) Are We Fools for Saving Our Money?

2) A Strategy for Maxing Out Your Retirement Savings

*Photo Credit: Microsoft Clip Art*

Rob Bennett says

This is Rob Bennett, MMD.

I wrote a comment on the earlier thread giving background on my work in the SWR area.

You might want to take a look at my calculator. It’s the first New School SWR calculator. I called it “The Retirement RIsk Evaluator”:

http://www.passionsaving.com/retirement-calculator.html

After I put up my famous SWR post at the Motley Fool site, there was a fellow who saw it there who devoted the next eight years of his life to researching the SWR question and related matters. His name was John Walter Russell and he was a fantastic guy. He was both smart and kind and was loved by just about everybody in all the communities in which he posted. I worked with John on a daily basis for years. He posted his research on a real-time basis at a discussion board I founded for that purpose (“The SWR Research Group”) and we benefitted from super feedback from lots of great people.

John died in October 2009. I miss him every day. But he posted his research at his web site — Early Retirement Planning Insights. So we all still have that to benefit from. The research John did was amazing. I think there will come a day when he will be known as one of the giants in this field, bigger than Bogle or even Shiller. I think it is very cool that we were able by working with our fellow community members to come up with all sorts of wonderful stuff that the big names in this field were never able to discover (our edge is that we were not concerned with marketing considerations, only in finding out what truly works in the long term).

Here’s a link to John’s site:

http://www.early-retirement-planning-insights.com/

There’s a lot of super material there. The best place to start is the “Foundations” section or the “Guidelines” section. I also love the Letters to the Editor.

Enjoy!

Rob

MMD says

Hi Rob: I received your email. I apologize, but my blog is setup to hold back any comments containing hyperlinks (such as the one you just posted). Also, I noticed your comments on Twitter.

Thank you for the link to John’s website. Early retirement is a common theme (among many) on my blog. I definitely plan to have a look around and see what I can learn from it!

Aaron Hung says

Very nice points, I’ll have to check the sheet when I get home. I’ve always wondered about that, are we going to take out more than what the investment will earn us?

MMD says

Aaron: Thanks! Please let me know how you like it and how you do with your withdrawal rate.

Alik Levin says

MMD,

My father was huge fan of Monte Carlo long ago when i was a kid, and I am huge fan of my dad so you brought me sweet memories with this post 😉

Downloaded the spreadsheet and looks like easy to use tool, all i need now is $1M to play with 😉

MMD says

You and me both could use that $1M! I’d like to think that all this devout saving and strategizing will help me achieve it. I hope you enjoy my spreadsheet. I’m glad it brought up great memories. And I hope it helps you to plan for many more great ones to come.

Rob Bennett says

my blog is setup to hold back any comments containing hyperlinks (such as the one you just posted).I of course understand, MMD. Thanks again for hosting a discussion of this important topic.

[i]I definitely plan to have a look around and see what I can learn from it![/i]

That’s kind. Please know that I am happy to help in any way if John’s work prompts any questions on your part.

Rob

Nick says

Ooohhh… a new toy! Thanks. Very interesting premise. Looking forward to playing around with the file.

Rob Bennett says

Thanks for your interest in the concept, Nick. I can tell you that lots of blood, sweat and tears went into that one. John Walter Russell gets half the credit.

Rob

MMD says

Thanks Nick. I hope you enjoy the file!

Rob Bennett says

Sorry!

When I saw Nick’s words, I jumped to the conclusion that he was referring to the Risk Evaluator, which is obviously the tool that is always uppermost in my mind in these discussions.

I am sure that you will enjoy a great learning experience studying the file put forward my MMD, Nick. And thanks much for providing us all with that tool, MMD.

Rob

Matt says

Thanks for posting the MC simulator — gaining lots of insights from running it with different assumptions, tweaking the variables, etc. One quick question — how are you treating taxes in your modeling? For example, I’ve been doing work with some other tools attempting to model a “forever” withdrawal rate that essentially ensures I never run out of money, even if I live 100 more years (hah!). I tend to get to something like a 3% rate if I assume the withdrawals must cover all my taxes, and a 2% withdrawal rate if I assume taxes are already covered and I’m just covering all my other expenses on the withdrawals. I believe your modeling assumes your annual withdrawals must cover all tax expenses, as well as all other “living” expenses each year. Is this correct? Thanks again.

MMD says

Matt: You are correct. Whatever withdraw rate you enter assumes that YOU are going to take care of the tax expense. In other words, this model is pre-tax. I apologize for not having including this very important characteristic of retirement planning within this spreadsheet. Perhaps I can update it at a later date. I suppose an easy way to figure this out would be to add another column where we subtract the taxable amount from the “Withdrawal Amount” column to give us an idea of how money we’d really have to play with each year.

Bill says

Your spreadsheet calculates withdrawal as a percentage of each year’s starting balance so the withdrawal amount can drop unless your return and/or inflation create a constant or increasing balance. I believe that a better approach is to set an initial withdrawal amount ($) and increase this each year by inflation. I believe that other models us this approach.

(example:

MMD says

Interesting spin on this model! I have never heard anyone characterize it in this manner. But it does seem to make sense. I will have to read through your article and then try out these calculations myself.

Bill says

For example, let’s say you have $1M and need $40k from your investments to live in your 1st year of retirement (4%). If you plug in 3% inflation, 8% mean return and 15% sigma (close to historical norms), you will see (after a few taps of the F9 key), some results where the withdrawal amount drops to only about $20k in perhaps as few as 15 years or less. Many of us could not reduce our spending that much as we age. If you use 6% inflation and 4% return(likely going forward), then you stand a much greater chance of missing your target of $40k in today’s dollars. In reality, I think most people will need a nearly constant withdrawal in inflation corrected dollars to live a normal life. There may be times when expenses go down (mortgage paid off) and others when expenses go up (illness) but they rarely continuously go down. In practical terms, if you need $40k to retire today, you better have much more than $1M or pray for a much better economy (high return, low inflation and low sigma).

MMD says

Hi Bill, thanks for the nicely laid out example. And I understand your point.

Going back to your last post, I was researching Bill Bengen (who initiated the 4% rule) and his findings. And yes, it does say to start with 4% and adjust each year for inflation. A few other articles validated this. So I do agree!

I will have to revise my model and spreadsheet! Thanks for bringing this to my attention.

Bill says

Looking forward to your revised spreadsheet! I suppose that you could also add a mean and sigma for inflation but then again life is complicated enough:) I enjoy your web site. One question off this topic: can you explain the difference and relationship between “YTD Return” and “SEC Yield”? I think SEC Yield is the annual expected cash from dividends whereas YTD Return includes stock/fund appreciation plus YTD dividends…but I’m not sure.

Thanks!

MMD says

Good question – I’ll have to look into that one!

MMD says

I was also excited to see the article in the link state that 3% will potentially last 50 years. It just again validates my choice to have a higher safety margin.