Details of the 4% Safe Withdrawal Rate

Let’s follow up on questions from last week:

Explain again this 4% – is that 4% withdrawal from the “investment pot” each year in retirement? Wouldn’t that automatically reduce the amount withdrawn each year?

And the 25 times living expenses…we might be unusual in that our children are both college age as we approach 65 so our current expenses might be much higher than other readers. So what is a reasonably robust guess?

Calculators are only as good as the formula(s)/equations behind them – what is the BEST Formula anyone has determined to date?

It’s a gnarly subject, and there’s perpetual controversy over the details.

The 4% system I’m referring to is the “classic” Trinity Study method:  at the beginning of retirement you withdraw 4% (the “Safe Withdrawal Rate”) from your portfolio. Each year afterward you raise that initial withdrawal by the rate of inflation. If you withdrew $40K from your $1M portfolio at the beginning of retirement, and later that year the Consumer Price Index was reported to be 3%, then next year you’d withdraw $41.2K (=$40Kx1.03). Note that it doesn’t matter whether your portfolio went up, down, or sideways during the last year– you had already set the withdrawal rate at the beginning of the retirement and you continue to blissfully raise it each year by the rate of inflation. Depending on the portfolio’s asset allocation, the Trinity Study’s 4% SWR has a very high (but not 100%) probability of surviving for at least 30 years.

Dozens of newer studies have examined all aspects of the Trinity Study, including this one. I’ve surveyed much of that research at this post and there’s a huge archive of further discussion at

A different system is a straight 4% withdrawal every year (no inflation adjustments). Hypothetically it never run out of money. It also depends on the size of the portfolio every year, so the amount of the withdrawal will vary wildly from one year to the next. However, some studies claim that a high-equity portfolio will grow faster than inflation, so a 4% withdrawal every year from such a portfolio would at least keep up with long-term inflation.

Neither method is much consolation for years like 2008. The S&P500’s -37% return could have boosted the next Trinity withdrawal to well over 6% of the portfolio’s actual value– or would have hammered down a straight 4% withdrawal by a drastic 37% spending cut. That’s where a variable withdrawal system can help, like Bob Clyatt’s “4%/95%” plan. It’s described in this post , but briefly the 95% part is an exception to the straight 4% withdrawal.  You spend a bit less money during down markets instead of having to cut all the way back to 4%, but you still reduce your spending to let your portfolio recover more quickly.

“25x living expenses” is just 1 / 0.04, the inverse of 4%. You base your SWR on your own long-term living expenses, and usually that does not include the kid’s college. Most retirees consider college expenses a separate pot of money to be handled apart from the retirement portfolio. That’s not necessarily more rigorous or “better” but it’s certainly easier, especially because you’re adjusting the college fund’s asset allocation as they approach high school. Of course some college degrees only cost about as much as a new roof while others cost more than the whole house, so your method may vary from the norm.

SWR calculations are based on either historical investment returns data or on randomly generated sequences. Two good examples are or Unfortunately history is a poor approximation of the future (let alone the recent proliferation of new asset classes) and investment returns cannot be rigorously analyzed with mathematical formulae– only with probability & statistics distributions. (Even that’s not foolproof.) The Trinity study is frequently criticized for “cherry-picking” the 20th century’s best returns data, and financial pundits are skeptical that history will ever repeat itself.

One aspect that’s become clear from SWR research is that the Trinity Study’s success depends on the sequence of returns. If you have the good fortune to retire at the beginning of a good five-year sequence of returns, then you’re almost guaranteed to have a successful retirement. If you retired in 1999 or 2007… well… maybe it’s not so certain.

Luckily, the dire warnings of the “series of returns” researchers assume that retirees blissfully follow the spending plan. (Their assumption makes the math and the programming much simpler.) Investor psychology has shown that humans just don’t do that. If the first five years of a retirement include returns like 2000-03 or 2008-10, then retirees will reduce their spending. Some might take part-time work or not even retire in the first place so that their portfolio has a chance to recover. I have yet to see a computer simulation of that spending model, but if you reduce your spending in a recession (and raise it back to your SWR after your portfolio recovers) then your “success” rate will also recover to 100%.

I’m skeptical of anecdotal data, but here’s one other point to consider. Almost every study ever performed on retiree spending has noted that spending drops off dramatically when the retirees are in their 80s. More of that anecdotal data comes from thousands of financial planners (and grateful heirs). You still have to watch out for end-of-life expenses, perhaps including long-term care insurance, but when you retire on a 4% SWR then it’s likely that by your 80s you’ll have money left over.

For every pessimistic financial adviser warning against the flaws of the 4% SWR, there are optimists doing more research. Although future returns may be lower, future long-term inflation may also be lower. Although the S&P500 rose rapidly over the last 80 years but has faced dismal returns over the last decade, there are many more asset classes available today. Today’s investors can also easily diversify and rebalance among a number of funds with much lower expenses and lower taxes.

Diehard skeptics have three ways to avoid this uncertainty, although they can be unnecessarily painful and may prolong your working years.

First, you could live off portfolio dividends and never touch the principal. However, instead of a portfolio that’s 25x expenses, you might need one that’s 30-50x expenses. A portfolio heavy with dividend-paying stocks may be more robust over the years, but its volatility can be breathtaking during a bear market (even though its dividend payments might be unchanged). The more conservative (and less volatile) your portfolio is, the lower its dividend rate and the more principal you’ll need to support your budget.

Second, you could use a different withdrawal system like Bud Hebeler’s “Analyze Now!” plan.  It bases next year’s withdrawal on the previous year’s performance and applies a considerable “margin of safety” to spending.  Jim Otar’s “Retirement Optimizer” calculator uses historical data with different recommendations for annuitizing part (or all) of your portfolio.

Finally, you could annuitize a portion of your portfolio to cover a bare-bones spending budget and use the rest to support your lifestyle preferences. In a good year, you’d spend at the 4% SWR without worry. In bad years you’d cut spending to the bare-bones budget and let your portfolio recover. If you’re a military retiree, then most of this annuity is provided by your pension. Even if you’re not a military retiree, you can still get a “safety net” annuity from your Thrift Savings Plan account and Social Security.

The “best” formula? I don’t think there is one– yet. No matter how rigorously logical a computer simulation can be, the resulting retirement system still has to provide “sleep at night” comfort. Bob’s 4%/95% system incorporates both the Trinity Study and the human tendency to reduce spending during recessions, so it’s better than the fixed-withdrawal systems. The 4% SWR plan also gives you a quick way to check your assets against your retirement budget (“25x”) to decide if you’re reasonably able to retire. If your budget is conservative and you’re ready to cut spending when necessary, then you’ll do fine. If you’re not comfortable, then you can keep working longer until you feel that you have enough margin of safety.

Related articles:
Back to the Trinity Study

Other related links from that post:
Military retirement spending: how much will I need?
Military retirement: how much can I really spend?
How many years does it take to become financially independent?
Tailor your investments to your military pay and your pension
Asset allocation considerations for a military pension

Early retirement and the kid’s college fund
How much should you save for college?

Does this post help? Sign up for more free military retirement tips via e-mail, Facebook, or Twitter!

WHAT I DO: I help you reach financial independence. For free. I retired in 2002 after 20 years in the Navy's submarine force. I wrote "The Military Guide to Financial Independence and Retirement" to share the stories of over 50 other financially independent servicemembers, veterans, and families. All of my writing revenue is donated to military-friendly charities.

We will be happy to hear your thoughts

      Comment? Question? What's on your mind?