The reason it has taken me so long to get this posted is that these spreadsheets contain a lot of information and I didn’t know the best way to go about writing it for AFM readers.

So, here is part one of a several part series (I’m not sure how many parts will be in the series).

First, let me lay out what I did:

• I took the monthly total returns for the S&P 500 and a Bond Index (using returns found in Ibbotson’s SBBI Yearbook) and created a 50/50 portfolio with annual rebalancing. NOTE: These are index returns and not actual investment account returns. In other words, my numbers were higher because they do not take into consideration fees associated with mutual funds or brokerage accounts.

• I assumed a beginning retirement balance of $1,000,000.

• I assumed a 4% initial withdrawal for income needs. The first withdrawal occurred at the beginning of the first year. In other words, $960,000 was invested the first year.

• In this scenario, I assumed that annual income would be 4% of the investment balance at the end of each 12 months. I know a lot of people disagree with this strategy. Most people like to use the initial withdrawal and then adjust subsequent withdrawals for inflation. I have those numbers and will publish them later. My findings seem to show that the fixed percentage withdrawal led to more income over the years—though the income could be variable.

• I assumed a 30 year retirement.

• I began the first 30-year period in January 1926, which ran through December 1955. The second 30-year period was February 1926 through January 1956. (You get the idea, I hope.)

**Findings…**

Unlike other withdrawal strategies, the fixed 4% withdrawal never ran out of money. In fact, the smallest ending balance over 30 years of withdrawals, was $1.8 million. This makes sense when you think about it because the withdrawal is a relatively small 4% of the account balance. The downside as stated earlier is that the income can be quite volatile from one year to the next.

• There were 685 rolling 30-year periods from 1926-2012.

• The average income over each 30-year period was $2,547,750.

• The highest income over a 30-year period was $5,778,756 (October 1981 – September 2011).

• The lowest income over a 30-year period was $1,210,642 (September 1929 – August 1959).

• Each first year’s monthly income was $3,333.

• The average final year’s monthly income was $13,030.

• The lowest final year’s monthly income was $3,363.

• The highest final year’s monthly income was $16,052.

• The average monthly income over all 685 rolling 30-year periods was $7,077.

• Over all 685 rolling 30-year periods, the average number of years that the income decreased was 9.6 times during each 30-year period.

I created a 58-page PDF report that you can download **here**. There will be more reports to come.

Really, if you have the monthly returns, you can look at the withdrawals monthly. No-one is going to withdraw a year’s take the first day of the year.

Doing that, one is also more likely to start at the END of the month, since pre-retirement paychecks will still be coming in the first month.

Yes, I thought about doing it that way but instead assumed the withdrawal would go into a bank account and used during the year.

BTW, I wrote a FORTRAN program to do the same thing many years ago, but I do not have data going back as far as you do.

Wow, you rock when it comes to doing research and generating really useful information. We are in our mid-fifties and with interest rates being so low the past several years we’ve been concerned about what kind of withdrawl rate we can use when we retire. It’s good to know that 4% is sustainable (based on history). Thank you for this wonderful post. I hope you have a great weekend.

Thanks, K D! I wish more people felt that way. Next week I will post the results of a 5% withdrawal.

Interesting numbers, but… Isn’t it mathematically impossible (by definition) to run out of money with this strategy?

It’s sort of like Zeno’s Paradox except you’re only moving 4% at a time.

As your balance shrink so does your withdrawal so you can never reach zero. So not running out of money doesn’t really prove anything.

In the illogical extreme, if your portfolio collapses to a buck, you’ll adjust and only take out 4 cents.

What really matters is how low the monthly (or annual) payouts could get. Averages are all well and good but are there periods buried in there where it would be impossible to live?

It would also be helpful to inflation-adjust these numbers as you move from year 1 to 30 so you’d have a sense for spending power. You could just use a flat 3% or something to get a rough idea.

I don’t mean for this to come of as an attack. Just food for thought. Looking forward to seeing your future reports on the subject.

I understand what you are saying but, that wasn’t the point. The point was that the consistent 4% withdrawal meant more income and higher ending balances.

Like I said, this is the beginning of a series. There will be more. I have inflation numbers too.

Great Work. Will you consider income taxes and RMD’s in some of your other runs?

Thanks