The only safe withdrawal rate for early retirement

The question of how much money can be safely withdrawn from a portfolio has been around for as long as portfolios have been around. Various academic analysis over the years have attempted to address this question, and the most reference is the Trinity Study, published in 1998. This study lead to the widely quoted rule of thumb that a 4% withdrawal rate will cause your money to last for at least 30 years. You put 20% to 80% of your money into stocks, the rest into bonds, and increase the amount you withdraw each year by the rate of inflation.

Ok, so far, so good. Looks like our friends in academia have solved our problem for us. We follow the advice above and we have a worry-free retirement, right?


There are a number of very serious issues with the 4% withdrawal rate. For one thing, it’s only designed to make your money last 30 years. That’s fine if you retire at 65, but not so fine if you retire at 45. In addition, many of the conditions that existed when the 4% rule was created no longer apply.

Issue #1 – ridiculously low bond yields

The Trinity study was done at a time when bond yields were significantly higher than they are today. In 1998 the yield on a 10 year Treasury bond was 5.54%. In July of 2016 the rate was 1.46%. That 4% difference essentially wipes out the entire 4% safe withdrawal! From 1973 to 1993 the 10 year Treasury rate never dipped below 6%. That’s a 20 year run with very high returns. Now bonds are yielding a mere 26.4% of what they were returning when the study was done.

Of course, since we are adjusting our withdrawals for inflation we only really care about real returns (bond interest rate minus inflation). If we look at real returns for TIPS (Treasury Inflation Protected Security) we get the following returns:

Date Value
29-Jul-16 -0.03%
1-Jan-16 0.67%
1-Jan-15 0.27%
1-Jan-14 0.63%
1-Jan-13 -0.61%
1-Jan-12 -0.11%
1-Jan-11 1.06%
1-Jan-10 1.37%
1-Jan-09 1.91%
1-Jan-08 1.47%
1-Jan-07 2.44%
1-Jan-06 2.01%
1-Jan-05 1.72%
1-Jan-04 1.89%
1-Jan-03 2.29%

(This shows the inflation adjusted return for TIPS on January 1st of each year from 2003 to 2016, plus the rate as of July 16, 2016)

TIPS work by providing a return that’s adjusted for inflation. If inflation goes up then your return goes up. If inflation does down then your return goes down. Buying a 10 year TIPS today means locking in a -.03% real return per year for 10 years, regardless of what inflation does.

Given that the 4% rule was based on rolling 30 year periods from 1929, the study used the higher real returns for bonds from that time period. Yields today are lower than in the past. Lower yields today result in less income. Less income requires lower withdrawal rates to ensure your money lasts for a minimum of 30 years.

Issue #2 – ridiculously high stock valuations

As I mentioned in a previous post, the S&P is clearly overvalued today.  The CAPE (Cyclically Adjusted PE, which is calculated using today’s price and an average of the last 10 years of earnings) puts today’s market at 60% overvalued. If the market reverts to the mean over the next 10 years and we have 5% growth in profits we’ll be looking at -0.73%/year returns after inflation. If we have a portfolio that’s split 50/50 between stocks and bonds we’d be looking at total return of:

Total return = 50% bond returns + 50% stock returns = (.5)(-.03%) + (.5)(-.73%) = -.38%/year

So, based on current stock and bond valuations, we can expect -.38% returns per year for the next 10 for a portfolio of 50% bonds and 50% stocks. As you can imagine, pulling out 4% of your original portfolio value per year after LOSING .38% per year from your investments is not a formula for long-term financial security. If you started with a $1,000,000 portfolio, then after 10 years you’d be down to $569,414 in inflation adjusted dollars.

At that point your $40,000 yearly withdrawal is 7% of your remaining portfolio. There’s no way you can expect to average 7%/year after taxes from a combined bond/stock portfolio. Even a very optimistic 5% combined return after inflation has you going broke at year 36, and again, this is a VERY optimistic scenario.

So where does that leave us?

Well, somebody already did an analysis on the success rate of various scenarios using the same time methodology used in the Trinity Study:




This shows the success rate for various combinations of stock allocation, withdrawal rates, and time periods. Note that for any stock allocation of 50% or higher a 3% withdrawal rate was successful 100% of the time through 40 years.

When I run the numbers in the scenario I presented above (-.38%/year return for 10 years, then 5%/year after that) using a 3% withdrawal rate we end up with $973,553.49 in today’s dollars even after 45 years. If appears that a 3% withdrawal rate can reasonably be expected to last forever.


Given the current environment of high stock valuations and low bond yields, a 4% withdrawal rate can no longer be considered safe. The longer the planned retirement, the lower the safe withdrawal yield. For early retirees looking at long (40+ years of retirement), a 3% withdrawal rate is safe in today’s environment of low bond yields and high stock prices.



6 thoughts on “The only safe withdrawal rate for early retirement

  1. Much like you, I’m 40 years old, and anticipate retiring by 45 with a very safe withdrawal rate. I could be done by now if we maintained our current spending and were comfortable with a 4% withdrawal rate.

    For a variety of reasons, including the fact that I mostly like my job, I am planning to work at least a few more years, and barring a market disaster, should be able to proceed with an early retirement and a withdrawal rate of less than 3%.

    Thank you for doing the math, and showing us what we might be able to expect.


    1. Just to clarify – are you saying you’re planning on working for a few years beyond today, or a few years beyond 45, when you believe you should be able to comfortably retire?

      I’m absolutely convinced that retiring early and expecting to be able to withdraw more than 3% has a very high chance of ending poorly for most people. People who do so are just setting themselves up for failure down the road. It sounds like you have the right plan – work until you can live on a 3% withdrawal rate.

  2. If someone is contemplating retiring in their 40s, I tend to think 4% withdraw rate is fine. I wouldn’t let any concern like that prevent me from making the leap. Most people who retire or are seriously think about it at that age, assuming one doesn’t have health related problems, are likely burnt out, completely bored/unmotivated & want a career change, or maybe just have a desire to take few years off. Even after retiring, everyone will either work part time at their own pace, become self employed, make a career change, or decide early retirement isn’t for them and return to work force. Hence they’ll continue to augment their portfolio and savings with other earnings even without a full time job. I doubt anyone can really just read, watch TV, travel for 40 years without doing some sort of work.

    Once someone reaches their 40s, unless you’re one of the lucky few who truly love their job or maybe are senior executive level at fortune 50 companies , most people understand time is more valuable than padding their bank account working at a job they’re not particularly excited about.

    1. James – the scenarios you are mentioning aren’t really retirement. Taking a few years off or making a career change is a different beast than permanently unplugging from a steady paycheck. I tend to agree with you that 4% is a reasonable withdrawal rate IF you have other money coming in. That other income gives you options when there’s a market drop or if there’s a hit to your income in some way. But permanently unplugging in your 40’s with a 4% withdrawal rate is, in my opinion, too risky.

  3. Here are the results, back-tested 1976 to 2016 with dividends re-invested on the S&P500
    SWR Years
    6.57% 40
    7.41% 39
    8.20% 38
    9.25% 37
    9.90% 36
    9.77% 35
    13.83% 34
    10.81% 33
    12.84% 32
    11.69% 31
    9.96% 30
    9.16% 29
    11.10% 28
    10.34% 27
    10.25% 26

    1. I’m not exactly sure what these numbers are. Is this just the average rate of return for each of these time periods? Because that’s very different than the safe rate of return for each of these time periods, especially once you adjust for inflation.

Leave a Reply

Your email address will not be published. Required fields are marked *