The death of the 4% rule

I’ve written about the 4% rule before. I’ve pointed out that the underlying assumptions behind the original “Trinity Study” (the basis for the 4% rule) don’t exist today, but everywhere I look people continuing to quote the “4% rule” as gospel. I’m sure you know the rule – if you withdraw 4% of the value of your next egg at retirement and annually update the initial withdrawal amount for inflation, you’ll never run out of money in retirement. You’ll play with unicorns all day and dance in meadows and have a perfect retirement, free of financial concerns.

It’s complete bullshit.

Research published in 2013 by Michael Finke of Texas Tech University, Wade Pfau of The American College, and David Blanchett of Morningstar Investment Management found that using historical interest rate averages, a retiree drawing down savings for a 30-year retirement using the 4 percent rule had only a 2 percent chance of running out of money. But using interest rate levels from January 2013, when their research was published, the authors found that retirees’ savings would grow so slowly that the chance of failure rose to 57 percent.

Reread that paragraph again. Using historical interest rate averages the 4% rule had a 98% chance of resulting in 30 years of income. But using interest rates from today the chance of failure rose to 57%.

“The 4 percent rule cannot be treated as a safe initial withdrawal rate in today’s low interest rate environment,” they concluded.

Interest rates – 2013 vs today

Well, that doesn’t sound good, does it?

“But wait!” you say. “That’s based on interest rates from 2013, not interest rates today.”

I can’t argue with you there. Interest rates in 2013 were at historically low levels due to the Great Recession. Surely they are higher now, which means the 4% rule DOES apply today…right?

Well, let’s check the US Treasury website to pull historical interest rates from 2013 and compare them to early 2017.

January, 20130.12%1.91%5.25%1.28%
December, 20130.08%1.81%5.06%1.05%
February, 20170.06%1.80%4.34%0.74%

Uh oh. That doesn’t look good for us at all, does it?

Interest rates are lower in every category than they were in 2013. This means we’d have even LESS income today from our bond allocation, meaning the 4% rule is even LESS likely to work today than it was in 2013.

Market valuations – 2013 vs today

“Ah”, you say, “but everybody knows bonds suck. Who would invest in bonds at today’s ridiculously low rates? Those lower interest rates don’t matter. Smart retirees will have higher equity exposure.”

Let’s investigate that line of thinking.

First, we have to understand the relationship between the market’s valuation and future returns. Thankfully for us, Vanguard has already done that for us. In October 2012 Vanguard published a study where they looked at a number of indicators used by market prognosticators to predict future market returns.

It’s worthwhile to note that nothing was very accurate in the short-term (1 year), but if you look at page 7 of the study you’ll see that the 10-year Shiller CAPE (Cyclically Adjusted Price/Earnings ratio) explained about 43% of the return of the market over the next 10 years. Here’s the graphic from that report:

Now look at Figure 5 from page 13:

This shows a regression analysis of the relationship between the 10-year Shiller CAP and the average real returns over the next 10 years. The word “real” is important here – this means we are looking at returns after inflation. If you’re not familiar with a regression analysis it’s just a statistical way of creating a formula/relationship/line on a graph that’s the best fit for various data points.

The regression analysis resulted in 2 lines on the graph above – the blue line is the fit for the Shiller CAPE 10 (which averages the last 10 years of earnings) and the red line is the fit for the Shiller CAPE 1 (which looks at just the most recent year of earnings).

The valuation is along the horizontal axis and the expected average return over the next 10 years is on the vertical axis. To use this graph you find the current valuation along the horizontal axis and then look up to see what the prediction is from the blue and/or red lines.

The downward sloping line shows that as valuations increase the expected return for the next 10 years decreases. For example, when the PE is 1o we expect to get around 10% real returns per year for the next 10 years (about a 2.6x total return). When the PE is 30 we see returns of around 0% annual returns according to the red line and 2.5% according to the blue line.

So what was the CAPE in 2013 and what is it now?

Well, from we have:

  • January 1, 2013 – 21.9
  • January 1, 2017 – 28.03
  • March 10, 2017 – 29.24

The valuation of the market today is 33.5% higher than in January, 2013.

And what did we learn from the Vanguard study? Higher valuations equal lower future expected returns. Looking at the red and blue lines we see that we can expect average returns of between 0% and 2.5% per year over the next 10 years based on today’s valuation.

Lower future returns result in higher chances of the 4% rule failing over a 30-year timeframe.

Fine, Pfau, and Blanchett determined that lower interest rates result in higher chances of the 4% rule failing over a 30-year timeframe.

So if the 4% rule had a 57% chance of failure at 2013 interest rates and the 2013 market valuation, what do you think the chance of failure is with today’s interest rates being 18% lower than in 2013 and the stock market valuation being 33% higher?

It’s hard to say without all of the data from the original study, but if the odds of failure were 57% when just considering higher interest rates, the addition on lower expected equity returns clearly increases the odds of failure even further.

Let me just reiterate what that means – if you retire today and withdraw 4% of your initial nest egg and then update that amount annually to match inflation, there’s more than a 57% chance you’ll run out of money in 30 years.

Implications for early retirees

I’m guessing that the vast majority of people reading this blog are aiming for early retirement or early financial independence. This means your money will almost certainly need to last for significantly longer than 30 years.

Checking the Social Security actuarial tables, we can see how much longer you can expect to live based on any given age:

  • Male @ 40 – 38.53 years
  • Female @ 40 – 42.43 years
  • Male @ 50 – 29.58 years
  • Female @ 50 – 33.16 years

A key thing to note – this is the average life expectancy, which means that 50% of the people will live longer and 50% will die sooner. I think it’s the height of folly to base your retirement plan on the average life expectancy – by definition there’s a 50% chance you’ll live longer (and possibly run out of money).

To be safe let’s add 10 years to each of the above estimates. This means the average man who retires at 40 should plan for about 48 years of retirement and the average woman should plan for about 52 years of retirement. If you’re 50 you’d expect to live about 40 more years for men and 43 more years for women.

In all of the above cases your planned retirement period is significantly longer than the 30 years of the Trinity Study. Let’s look at the actual conclusions from the Trinity Study:


For any given mix of stocks/bonds, we see 2 irrefutable trends:

  • As the withdrawal rate increases (as you move to the right in the table) the chances of running out of money increases
  • As the timeframe increases (as you move down the table) the chances of running out of money increases

As an example, look at the section for 75% stocks/25% bonds. For a 15 year time horizon we see that the money lasted 100% of 15 year scenarios at a 3% withdrawal rate, it lasted 96% of the time at a 7% withdrawal rate, and 46% of the time at a 12% withdrawal rate.

Looking at the same information we see that at a 7% withdrawal rate the money lasted for 96% of 15 year periods and 88% of 30 year periods.

(And remember – these calculations were all done in periods of higher bond rates and lower equity valuations).

You’ll note that 40 and 50 year timeframes aren’t in this table. I’m guessing this is because the authors of the study didn’t envision people needing to live off their retirement income for that long.

Now let’s put this all together.

  • Large numbers of people are pinning their retirement on the idea that a 4% withdrawal rate is sustainable in retirement. They are literally banking on the conclusion from the Trinity Study that showed that a 4% withdrawal rate was sustainable for 30 years for 98% of the studies periods
  • Today’s lower interest rates have, according to a recent academic study, raised the failure rate to 57% rather than 2%.
  • Today’s higher equity valuations strongly imply lower returns over the next 10 years. This further increases the chance of failure.
  • Early retirees will need to live off their investments for longer than the 30 year timeframe studied in the Trinity Study. This further increase the chance of failure.

Add that all up and I believe that an early retiree today who relies on the 4% withdrawal rate will have a 75%+ chance of running out of money in retirement.


Ok, we’ve established that retiring today and relying on the 4% rule to get you through retirement is not going to work. So what should you do?

First, you should rely on something lower than 4% as your withdrawal rate. I’d recommend somewhere around 3%. If you look at the Trinity Study results again you’ll see that a 3% withdrawal rate succeeded in every timeframe and with every mix of stocks and bonds.

Based on my research I believe a 3% withdrawal rate is safe indefinitely. A 3% withdrawal rate should literally last forever.

Second, you might consider working another year or two, or at least until the stock market cools down a bit and interest rates move towards their historical norm. This might have the effect of a stock market correction. Basing your initial withdrawal rate on a lower portfolio value will have much the same effect as initially withdrawing 3% instead of 4%.

If you’re close to retirement age or thinking about pulling the plug early and starting on early retirement using the 4% rule, you need to reconsider.


Have you thought about your planned withdrawal rate when you retire? Have you been planning on using the 4% rule? Do you think the 4% withdrawal rate still applies in today’s market environment?

9 thoughts on “The death of the 4% rule

  1. I don’t plan to ever touch the principal, but I also don’t plan to ever give up an active income either.

    Although the second statement could change over time.

    We are currently working to build a $10M net worth that kicks off about 6% or $600,000 per year in income. It will be with a combination of real estate, stocks, online business, and other opportunistic investments.

    This may prove difficult if rates stay low forever. But that’s the goal.

    Based on our current spending, that would 87X multiple.

    1. I’m curious – are you planning on actively working your entire life (in your 70’s, 80’s, and 90’s)? Given your financial goals that obviously wouldn’t be necessary, but I’m guessing you’d be doing it because you like what you do (or expect to like whatever you’ll be doing at the time).

      I’m also very curious about the 6% income goal. Stocks usually yield around 2.5%, so you’d have to have a large amount of much higher yielding investments to generate that kind of return.

      Also, what multiple would $600k of income be of your inflation adjusted future spending level?

      1. To be honest, I will likely work for as long as I am able to. Keeping an active income doesn’t necessarily mean a full-time income. The whole early retirement thing just doesn’t appeal to me. But maybe I read into it too much. I generally don’t like the idea of retirement in general.

        My pursuit of financial freedom is solely for the security and optionality it provides.

        Yes, I am aware that stocks only yield 2.5%, so I will definitely be mixing in other higher yielding asset classes like:

        – Rental Real Estate (only have 1 property at the moment). I’ve seen 10-20% cash on cash returns by investing in certain markets.
        – Hard Money Lending (currently getting 8% on average)
        – Online income (it’s only about $500/month now, but goal is $25,000/month longer term)

        So, these blended in with the stock returns is how I am planning to get there. Of course it isn’t going to be easy, and we may get to $10M and fail to reach our income goal. The good news is the $10M goal that I set back in 2014 is only a 20 year goal (I will be 47), so I will have plenty of time to adjust. We could also come up short of both the net worth and income goal. Only time will tell, but that is what I am working towards.

        In 20 years, the $600K in income will be equivalent to $330K in today’s dollars, which would represent a 3X multiple of our spending.

        1. That’s a really interesting take on retirement. Do you love what you’re doing now, or do you expect to change careers at some point? In my recent post about Michael Lewis’ father remember we had an exchange about how you and I both took the “make money so you can do what you love” route rather than the “do what you love and the money will follow” route.

          I did the quick math – if you start from $0 and want to have $10M in 20 years, assuming 8% annual returns, you’d need to save/invest $218,522.09/year. Given your current saving/investing that seems very possible.

          What’s really crazy is what happens from year 20 to 30. Assuming no withdrawals you’d have $21.59M at 30 years. That’s life-changing generational wealth.

          1. Yes, I still agree with what I said, which was:

            “He is essentially saying to follow your passion and the money will follow. I have tended to believe it is better to follow the money and then use that money to fund your passion.”

            Then I followed that up with the following:

            “But I do agree that if you can find an overlap between three things: Your Passions, What your Good At, and What the Market Is willing to pay for…then you are pretty much set in loving what you do.”

            I have been lucky to follow the money, but also land into the magic overlap I described above.

            I agree the $10M seems doable as long as we continue to stay disciplined and things continue to go our way. So far we are into the 3rd year of the 20 year plan and are ahead of schedule. At the end of 2016, we need $428K and ended up with $527K.

            I update the model with my actual income and net worth figures to back into the annual return I need to achieve the $10M goal by the year 2035. The assumptions of the first model had us saving 50%/year, a starting income of $179K, and the annual return assumed was 8.8% (this was plug that got us to $10M over 20 years of compounding and contributions), and annual increases of $20,000/year in income (from any source).

            We have blown the income assumptions out of the water, which is what has helped us get so far ahead. We finished 2016 with the income I hadn’t projected until 2023. And this year we are on track to earn the income I projected in 2028. The savings rate stays static at 50%, but the annual rate now needed is only 5.9%, and I suspect if we can keep growing our income aggressively, that this number will only fall.

            In the most updated version of the model I have $5.4M of the $10M coming from contributions alone.

            This also doesn’t account for any X-Factors, like my company stock being worth 7-figures + during our next liquidity event. A huge windfall from a lucky investment. Or something along those lines.



            1. I think this just goes to show that there’s just no substitute for a high income. Yes, it’s true that income doesn’t help much without a corresponding savings rate, but the reality is that saving 75% of a $50k income isn’t going to get you as far as saving 25% of a $200k income.

              And I agree with being at least relatively conservative in future projections. We have some private equity investments that I’m assuming will never be worth anything, but if they end up being successful that would definitely fall into the windfall category and just increase our income (or level of financial independence).

  2. I’m with you. I never believed that a 4% withdrawal rate is safe for any early retiree (unless they’re invested 100% in stocks and not have a balanced portfolio – which means they’re taking on too much risk). 3% should be the target withdrawal rate. I’m 47, my husband is 52. We currently have $2M in investment accounts and a $2M equity in our home and 10 rental properties. We generate approx $100K in net income from our rentals but I still don’t think we have enough to be able to retire. We have a target spend of $140K during retirement, which means we need about $180K of total income (with $40K going to taxes). This means we need to have about $2.8M in investments in order to stick within the 3% withdrawal rule. I think we’ll finally be able to get there within 6 years, when i will be 53 and my husband will be 58. Love your site. Keep up the good work.

    1. It sounds like you’ve put yourself in an enviable financial position. I’d love to learn more about your rental properties, how you built you investment portfolio, etc. It might make the basis of a great future article.

      1. Hi. Thanks for the interest. We started buying rental properties in 2006 and bought about 1 property per year since. We put in a pretty sizable down payment for each one since we wanted to have a positive cash flow right away. Approximate cost for 8 of the rentals is between $170K – $180K (plus a couple of $300K properties) and our max loan for most of them is $100K, with a couple of loans at about $120K. Our annual gross income for our rentals is about $150K. But, since we only have outstanding mortgages in 5 properties (we paid off a few of them very quickly by using net income from our jobs to prepay mortgages), it’s straight out free cash flow on 5 of them (only costs are property taxes and insurance). Hubby and I both work still for Fortune 500 companies, with gross income of $400K combined. However, we have a huge amount of expenses as both kids are in private schools. We make about $550K per year (including rental income) and get to save about $200K per year. Our yearly expense at this time is about $200K, with about $50K spent on the kids (not including taxes). In 6 years when we retire, both kids will be in college (with fully-funded college funds of about $300K each). That’s how we came up with a retirement budget spend of $140K per year plus $40K of taxes.

Leave a Reply

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