What’s the Safe “Safe Withdrawal Rate” for Retirement Now?
The 4 percent rule, first described by Bengen (1994), is a convenient shorthand both for those who are figuring how much to withdraw from their investments during retirement, and for those who are not yet retired and trying to figure out how much they need to retire. What this rule tells us is that a starting Safe Withdrawal Rate of 4% of the value of the investments, thereafter adjusting annually for inflation, is sufficient to ensure that the retiree will not run out of money over a retirement of 30 years. In fact, in the majority of the cases, the retiree will finish retirement with more money than at the start! But has that changed with the current era of low interest rates and high stock valuations? After all, the “rule” is almost 30 years old! What is the safe Safe Withdrawal Rate for retirement now?
But there have always been doubts about this rule. For example, Finke, Wade and Blanchett (2013) (and also here) argue that the fact the 4 percent rule works using historical US data is an anomaly, in a period where both stock and bond returns were higher than they are today in the US. If future stock and bond returns are lower, then retirees should instead be using a much lower initial withdrawal rate to be assured of retirement sustainability. Work done by “Big ERN” here also support this conclusion. Pfau (2011) also finds that the 4 percent rule hardly works outside of the US. Our own work here and here also finds that success of the 4 percent rule in the past 20 years both in the US and in Singapore is not guaranteed either.
How can we withdraw our investments in retirement safely now?
And the debate about the safe Safe Withdrawal Rate for retirement now continues
After several years of startling returns to stocks in the US in the late 2010s, including the much anticipated crash and rebound during the Covid-19 pandemic in 2020, the debate about the Safe Withdrawal Rate for retirement now continues. The financial markets have evolved since the studies cited above in the early 2010s, with stock valuations reaching new highs, and the US Federal Reserve starting to raise rates. For many observers, these developments may herald a sharp drop, or at the very least, lower levels of stock returns for the next decade.
Morningstar (2021) starts the ball rolling this time round by estimating the standard rule of thumb should be lower at 3.3%, rather than 4%, for the decade ahead. This outcome is based on a series of Monte Carlo simulations with the following assumptions for stock and bond returns and risk:
Return and Volatility Assumptions used by Morningstar (2021)
| Expected Return | Standard Deviation | |
|---|---|---|
| Stocks | 8.01% | 17% |
| Bonds | 3.30% | 4.2% |
| Cash | 1.71% | 1.73% |
| Inflation | 2.21% |
Predictably, this lowball estimate of the safe Safe Withdrawal Rate generated a fair amount of discussion. While 3.3% is not significantly different what ERN recommends, it is lower than what most followers of Financial Independence Retire Early (FIRE) use for their estimates. Even Bengen himself has advocated a 4.7% Safe Withdrawal Rate in recent years!
So, who is right and who is wrong? Bengen (2021) argues that the lower expected returns which Morningstar (2021) assumes neglect mean reversion in stock returns seen historically. In fact, the lack of mean reversion in Monte Carlo models generally result in Safe Withdrawal Rates which are lower than those estimated using historical returns. After all, stock returns (at least in the US) have always rebounded following a crash (for example in 1987, 2000, 2008 and 2020)! Kitces.com (2021) also point to the use of Monte Carlo as a key reason why the Morningstar (2021) Safe Withdrawal Rate is so low, and argues that this actually proves the resilience of the 4 percent rule because the return assumptions correspond to the worst ever outcomes for the US stock market historically.
What do we think is the Safe Withdrawal Rate for retirement now?
There are actually merits with the approach Morningstar (2021) has taken. And while both Bengen (2021) and Kitces.com (2021) correctly point out the lack of mean reversion in the stock and bond returns in the simulations, the problem with Sequence of Returns Risk is that mean reversion does not matter if the poor returns happen immediately after retirement! It doesn’t matter if stock returns mean revert to a higher level in years 11 to 20 of retirement as the damage from sequence risk is already done through the low returns in years 1 to 10. Just imagine how the people who retired right before the NASDAQ crash in 2000, followed by the Global Financial Crisis in 2008, are doing now, despite the huge run-up in US stocks since 2009 (see here)
Hence, given the high stock valuations now, which is a fairly good predictor of returns over the next decade, it seems prudent to target a lower Safe Withdrawal Rate for retirement now. Even William Bengen seems to have changed his views somewhat, calling for a lower Safe Withdrawal Rate of 4.4% for retirement now, instead of his previous view of 4.7% (see here).
And this means that non-US retirees have to reduce their Safe Withdrawal Rates even more. Why is this so? The answer is given in the Morningstar (2021) report:
Safe Withdrawal Rates for Retirement when returns and volatility change
| SWR | 0.0% change in Volatility | -0.5% change in Volatility | -1.0% change in Volatility | -1.5% change in Volatility |
|---|---|---|---|---|
| 0.0% change in Return | 3.3% | 3.4% | 3.5% | 3.6% |
| 0.5% change in Return | 3.5% | 3.6% | 3.7% | 3.8% |
| 1.0% change in Return | 3.7% | 3.8% | 3.9% | 4.0% |
| 1.5% change in Return | 3.9% | 4.0% | 4.1% | 4.2% |
An increase in the volatility or standard deviation of the investment portfolio results in a lower Safe Withdrawal Rate. Much more so than a lower expected return. For various reasons, international stocks tend to have lower returns, and higher volatility than US stocks. That is the main reason why non-US retirees, who have a larger proportion of their investment assets outside the US stock markets will end up with a lower Safe Withdrawal Rate. And that is also why the 4 percent rule does not seem to work well for international portfolios.
What can we do to safeguard our retirement?
Curiously enough, the debate about the safe Safe Withdrawal Rate for retirement now is very much a storm in a teacup. Interesting, perhaps, but not really relevant anymore. How is this so? It is because the debate is really still about financial options available to retirees back in the 1990s. Developments in finance have moved on quite a lot since then, and there are much better ways to handle sequence risk and retirement adequacy now.
For a start, let’s think about how sequence of returns risk affects the health of the investment portfolio and how much we can draw from it. A series of bad investment returns at the start of retirement means that we have to sell stocks to draw from the portfolio when stock prices are falling. This in turn means that the portfolio gets smaller and smaller, and cannot sustain for the tail end of the 30 year retirement horizon. The traditional way of dealing with this, as discussed in Morningstar (2021), is to keep retirement spending flexible, and reduce spending when returns are poor. Portfolio adjustments, e.g. rebalancing, spending from dividends, coupons and interest first etc. may also be used.
But this is like closing the stable door after the horse has bolted! Why do this elaborate system of spending and portfolio adjustments after the poor returns instead of before them? For example, approaches like the equity glide path, bucketing system and dividend investing are all methods which try to hedge the portfolio against sequence risk by avoiding selling stocks in the first few years of retirement, especially when returns are poor, and allow the stock portfolio to grow over a longer period. This addresses sequence of returns risk much more directly, and proactively than doing spending and portfolio adjustments after the returns are poor.
Adjusting spending in retirement after experiencing poor investment returns is like closing the stable door after the horse has bolted!
In longer term, sequence of returns risk manifests itself after 15 to 20 years of retirement when the portfolio has shrunk too much due to initial poor returns and drawdowns to sustain further inflation adjusted spending in the last few years of retirement. Again, a more direct way to address this risk is to purchase a deferred and inflation adjusted/indexed annuity to guarantee a minimum standard of living in the latter years (see here). And also to purchase this annuity at the start of retirement when it is at its cheapest (see here).
Finally, there is the question of how to draw down the investment portfolio during the middle years of retirement, i.e. years 5 to 15. Again, there are many developments here such as selling stocks to purchase inflation index bonds to safeguard and to keep raising the minimum standard of living as proposed by Kotlikoff (2022). We can also take advantage of the fact that retirement spending during this phase (the slow-go years) also seems to grow slower than in the early phase (the go-go years), allowing for more spending flexibility, as Blanchett (2013) shows.
Annual Real Change in Retirement Spending from Blanchett (2013)
In fact, this pattern of spending in retirement argues against the traditional methods of tackling sequence risk by cutting back spending in the face of poor investment returns, precisely because the spending in the early years of retirement is not flexible at all!
Therefore, to confine ourselves to thinking about Safe Withdrawal Rates for retirement by looking through a lens of the financial technology and knowledge which existed back in 1994 is really selling ourselves short! The financial world has moved on, and we deserve to be able to use the latest approaches and thinking about solving these thorny issues of retirement spending.
Conclusions: How to plan for a safe Safe Withdrawal Rate for retirement
So, while there has been much debate about Safe Withdrawal Rates in the face of stretched stock valuations and rising interest rates, much of this debate is futile. Futile, in the sense that it assumes that retirees:
- are stuck with a single asset allocation throughout their retirement
- have flexibility to adjust spending in the face of poor returns
But that ignores everything that we now know about retirement spending, and the methods for tackling sequence of returns risk:
- using dynamic asset allocation such as the equity glide path and bucketing strategies
- keeping spending fixed for the first few years of retirement and flexible after that
- using deferred annuities to hedge against running out of money in the long term
- hedging equities risk directly using derivatives
- planning to sell down stocks in a way to lock in gains for higher living standards
These are all more fruitful directions to pursue, rather than relying on the traditional investment portfolio with a Safe Withdrawal Rate to solve every single problem! To a hammer, every problem looks like a nail. So, let’s not get too concerned about the current Safe Withdrawal Rate debates, because they do not really try to address the issue using all that we know and can do.
References
William Bengen (2021) Is It Now the “3.3% Rule”? Advisor Perspectives
Blanchett, David (2013) Estimating the True Cost of Retirement, Morningstar
Early Retirement Now.com (2021) The Safe Withdrawal Rate Series
Morningstar (2021) The State of Retirement Income: Safe Withdrawal Rates
Wall Street Journal (2022) Cut Your Retirement Spending Now, Says Creator of the 4% Rule
You may be also be interested in our earlier work on this topic:
- The 4 percent rule for Financial Independence
- Does the 4 percent rule work in Singapore?
- How Sequence of Returns Risk Messes Up Retirement and the 4 percent rule
- Does the Yield Shield protect retirement finances?
- The Yield Shield, 4 Percent Rule and CPF Life – A Perfect Retirement Combo?
- A Yield Shield Portfolio for 2020
- Retiring and Surviving in a Pandemic – The Yield Shield 6 Months On
- Retiring in a Pandemic: The Yield Shield, the 4 percent rule, and Sequence of Returns Risk A Year On
- How to Spend your Retirement Savings (1): How much can I spend from my investments?
- How to Spend Your Retirement Savings (2): Combining CPF Life with Investments
- Dividend Investing: When does it work, and when doesn’t it work?
People tend to overestimate doomsday.
Based on the 4% rule, a retiree who retired on 31st Dec 1999 & put $1M into 60:40 using Vanguard total US stock mkt and a Vanguard total US bond mutual funds … will have $1.22M in his portfolio as of 22 Apr 2022. Every year along the way, he withdraws $40K inflation-adjusted — in 2021 he withdrew $66,260 as his inflation-adjusted spending money.
The above includes the worst decade for US stocks (2000-2010). The 2nd worst was 1929-1939. Third worst was 1972-1982. [All based on total returns, which includes dividends.]
During the 3rd worst decade of the stagflationary 1972-1982, the 4% method resulted in $1M becoming $1.04M by Dec 1982. And in 1982, you would have withdrawn $95,000 as inflation-adjusted spending.
For those who have accumulated enough to comfortably use 3.5% or 3% — good for you. It is a big margin of safety & likely your beneficiaries will inherit a few millions dollars 30 or 40 years later.
Otherwise 4% method still works. It’s designed so that you don’t look at your portfolio. Except once each year when you re-balance.
People who keep checking their portfolios will usually sell & cash out whenever there’s a crash or recession.
Thanks for your contribution!
There are a number of simulated results for this retirement cohort from various tools on the web, ranging from disastrous to healthy, so we need to take the results with a pinch of salt.
Assuming you are using Portfolio Visualizer, it seems that there is an adjustment that is needed for inflation to get the comparable result – see here: https://www.reddit.com/r/financialindependence/comments/rwsj4s/charts_year2000_retirees_the_ultimate_survivors/hrq5k0a/
Not an expert on the tool, so make of it what you will.
Thanks for the note.
And nope, the above historical examples are still correct.
The adjustment you referred to simply shows the generated graphs in terms of the starting year inflation-adjusted dollars. i.e. a chart from 2000-2022 is displayed using 2000-value dollars.
So if a person keeps $1M under his mattress from 2012-2022, the graph will display a slowly descending line with ending amount of $817,000 in 2012-value dollars (assuming 2% yearly inflation). You can see how even a low 2% inflation can eat up the purchasing power! Even though there’s still $1M under the mattress.
Most tools dealing with retirement sufficiency & payouts will be inflation adjusted by default (at least the good ones).
In my 2 examples above, the $66,260 withdrawn in 2021 is equal to $40,000 in 2000-value dollars. This was a period of VERY LOW inflation (also lots of financial hiccups in the 2000s & early-2010s).
While the $95,000 in 1982 is equal to $40,000 in 1972-value dollars — more than 2X the original number in just 10 yrs! And this $95K withdrawal is almost 10% of the remaining portfolio! This is Stagflation!
A retiree in Dec 1982 will be sweating & wondering if he needs to clean toilets or work in NTUC at age 80+. Without an appreciation of long term financial history, he may simply give up & cash out his portfolio, or do all sorts of tweaking & trading etc which will be the wrong thing to do.
If he stays the course & simply does nothing except re-balance once-a-year & withdraws his inflation-adjusted spending annually …. by Dec 2002 his portfolio will be $2.75M … or $625K in 1972-value dollars. This is after Sep 11 terrorist attacks and in the midst of -50% dotcom bust drop.
He would have withdrawn $176,000 in 2002 for spending (remember, this is equal to $40K in 1972-value dollars).
By Mar 2022 (if he’s still alive lol), the portfolio would be $3.18M in nominal dollars & withdrawing over $270K yearly for spending.
Thanks for the clarification!
It will be very helpful for others who use the same tool.