What Is The Best Way To Retire Early? The 4% Rule?
The New York Times had a piece on how the FIRE movement is being or is expecting to be impacted by the coronavirus situation and the expectation is that it could have a big negative impact on people who were retired very early and living on the 4% rule and a lot of the reasons why are going to be discussed in this post why the 4% rule might not be sufficient for a very early retiree with sixty year time horizon.
Nobel Laureate William Sharpe has referred to retirement income as the “nastiest, hardest problem in finance.” The 4% rule for retirement spending has gained popularity as a simple answer.
The rule says that you can safely spend 4% of an investment portfolio in the first year of retirement, and then adjust that amount for inflation each year for the rest of your life.
Distilling a complex problem down to a simple rule of thumb can be useful, but the 4% rule for retirement spending has some serious flaws. And even beyond those flaws, the retirement income problem is complex enough to warrant a more involved discussion.
The 4% rule makes retirement math exceptionally easy.
Take your required expenses and divide by 4%. If you want to spend $40,000 per year in retirement, divide $40,000 by 4% and you have a $1m required portfolio to meet your retirement income goal.
In its design, the 4% rule is meant to deal with the main risks that a retiree faces.
Once you have decided to stop working, or stop working for money, you will likely be relying on relatively risky assets, like stocks and bonds, to sustain a steady stream of income. This introduces a unique retirement risk called sequence risk. Funding a steady stream of income using a volatile asset like stocks poses some big challenges.
Sequence risk is based on the fact that repeated negative returns early on in the distribution period can have a meaningful long-term impact on the ability to fund your lifestyle.
As if sequence risk weren’t enough to deal with, we are also planning for an unknown period of time - nobody can predict how long they are going to live.
This risk is called longevity risk. Together, sequence risk and longevity risk make planning for a secure retirement a unique challenge.
Let’s come back to the 4% rule - the supposedly simple solution to the retirement income problem.
The 4% rule was designed to address sequence risk, and it did touch on longevity risk. I’ll explain the genesis of the rule. William Bengen, a financial planner, wrote a paper in 1994 with the creative title Determining Withdrawal Rates Using Historical Data.
Bengen took historical data for US stocks and intermediate term treasuries and tested how long a portfolio of 50% stocks and 50% bonds would be able to sustain various levels of withdrawals.
Bengen modelled withdrawals starting as a percentage of the portfolio and increasing with inflation each year. The result is constant inflation adjusted spending.
He tested withdrawals starting each calendar year from 1926 to 1976 and observed how long the portfolio lasted at each starting point. For a 30-year retirement period with a 50% stock and 50% bond investment portfolio, Bengen found that a 4% withdrawal rate was always sustainable.
This research was further supported by the 1998 paper Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable by Cooley, Hubbard, and Walz, commonly referred to as the Trinity Study. Instead of using intermediate term treasuries as fixed income, the Trinity Study used long-term high-grade corporate bonds. This decision to use riskier bonds ended up resulting in a 4% withdrawal rate being successful only 95% of the time in the historical data.
A 95% success rate sounds great, but this is only true in the historical data. A retiree today should not expect a 95% chance of success with a 4% starting withdrawal rate.
I’ll explain why that’s true in a minute, but first we have to come back to one of the most important assumptions built into the 4% rule: the time frame.
The 4% rule was based on testing a 30-year withdrawal rate in the historical data. This post is about retiring early. How many early retirees are planning for a 30-year horizon?
Bengen based 30 years on a 60-65 year-old investor and added roughly ten years onto normal life expectancy to plan for longevity risk. A 40-year-old retiree would usually expect to live for more than 30 years.
If we repeat the same analysis for a 40-year period, the 4% rule has a success rate of 87% with a 50% stock and 50% bond portfolio. Longer time periods will see more failures.
One consideration might be increasing the weight in stocks to improve the outcome.
Expanding the analysis to longer time periods and more aggressive stock allocations,they found that while anywhere between 50 and 75% stock exposure resulted in great results for a 3.5% spending rule over 30 years, moving out to 60 years required much higher stock exposure in order to reach a comparable success rate.
They found that for 60-year horizons, higher equity weights give better results. Anything below 70% in stocks becomes precarious over such long time frames.
To quote earlyretirementnow, “Over longer horizons, bonds are bad!”
This finding is extremely important for any early retiree. In the historical data bond exposure over very long periods of time is arguably riskier than stock exposure. This looks fine on paper, but investors are myopic. They evaluate long-term decisions with a short-term view.
You still have to live with the volatility of stock exposure in your investment account.
I think that we have established that the 4% rule is probably not useful when it comes to retirement planning in general and especially early retirement planning.
Increasing the weight in stocks might support a 3.5% withdrawal rate in the historical US data, but we are still missing some important pieces.
Everything that we have covered so far centers on historical US stock and bond data. There is a good chance that the historical experience of the US financial markets might not be representative of expected future returns.
There are a couple of different ways that we can think about this. The first one is looking at the 4% rule based on stock market data from other countries.
Wade Pfau documented this analysis in his book How Much Can I Spend in Retirement using data going back to 1900 through 2015 for 20 developed market countries. He found that the only other country that could have historically sustained a 4% withdrawal rate for a 30-year retirement period was Canada. 18 other countries would have had failure rates between 8 and 62%.
The global stock market as a whole would have sustained a 3.5% withdrawal rate historically.
It’s also important to point out that the 20 countries in Pfau’s data set are the countries that made it into the dataset - there is survivorship bias potentially making things look better than they actually are.
Countries like Argentina, Russia, and China did not make it into the dataset, but you can be sure that they would not have historically sustained a 4% withdrawal rate, and they would pull down that 3.5% safe withdrawal rate for a global portfolio.
Getting our heads out of the US was important, but we are still stuck in history.
Today stock valuations are high relative to the past.
Bond yields are historically low, or even negative depending on where you look. Nobody can predict the future, but stock valuations matter a lot to expected future returns.
In the 2019 edition of Aswath Damodaran’s annually updated paper Equity Risk Premiums,Determinants, Estimation and Implications Damodaran demonstrated that the earnings yield is the best predictor of the future equity risk premium. It’s still far from perfect, but it is the most reliable metric that we have for forecasting future stock returns.
The earnings yield can be found by taking the inverse of the Shiller cyclically adjusted earnings. For example, if the Shiller CAPE for US stocks is currently 29.71, we take one divided by 29.71 to find the earnings yield. This gives us a result of 3.36%, which is the expected real return for US stocks. The geometric average real return for US stocks from 1900 through 2019 was 6.5%.
Interestingly, 6.5% is roughly what the historical average Shiller CAPE ratio would predict.
Think about that for a moment. We are looking back at 119 years of data for US stocks during which the 4% withdrawal rate was sustainable, but over that period valuations were considerably lower, and the average historical returns that we see are commensurate with those lower valuations.
Today’s high stock valuations forecast much lower future returns.
Applying any historical analysis to today’s starting point does not make sense.
Knowing the limitations of historical data due to currently high valuations, one approach to testing spending rules involves using current expected returns and simulating future periods using Monte Carlo simulation.
Monte Carlo involves randomly sampling returns from a defined distribution of potential outcomes. Using Monte Carlo for a 60-year period with current expected returns for a 100% stock portfolio consisting of Canadian, US, and International stocks, and ignoring taxes, I find a 2.5% safe withdrawal rate where safe means a 5% chance of failure. This analysis is interesting for more than observing the safe withdrawal rate.
It also allows us to see the range of outcomes. In the worst 10% of outcomes our 60-year spending period left the investor with $1m adjusted for inflation, while in the best 10%, they were left with around $30m adjusted for inflation.
That massive range of outcomes seems inefficient, and it is. In a 2008 paper titled The 4% Rule—At What Price?, William Sharpe and two co-authors explain: “Supporting a constant spending plan using a volatile investment policy is fundamentally flawed. A retiree using a 4% rule faces spending shortfalls when risky investments underperform,may accumulate wasted surpluses when they outperform, and in any case, could likely purchase exactly the same spending distributions more cheaply.”
In simple terms, retirees who use a fixed spending rule from a portfolio of risky assets to fund their inflation-adjusted lifestyle needs are overpaying for the potential of investment gains that they do not need to meet their retirement income goals.
More efficient solutions to this problem are mathematically dense and often involve complex financial products like options, leverage, and annuities, but there are some spending rules out there that approach a more efficient solution without getting too complicated to implement.
In a 2017 paper, Vanguard explained some potential approaches. The authors explain that there is a spectrum of spending rules that depend on the preferences of the retiree. Constant spending rules like the 4% rule cater to the preference of spending stability while risking premature portfolio depletion or inefficient consumption. At the other extreme, spending a constant percentage of the portfolio each year results in no chance of depleting the portfolio and more efficient consumption, but it also results in potentially wild swings in the annual spending amount.
The Vanguard paper suggests a middle ground where spending is a percentage of the portfolio, but is only allowed to increase up to a ceiling or decrease down to a floor. The ceiling and floor can be tailored to the needs and preferences of any retiree, keeping in mind the trade-offs at each extreme. In their paper they use a 5% ceiling and a 2.5% floor. Keep in mind, though, that in bad markets there could be multiple years of spending reductions required to follow the rule.
Finally, I think that one of the most important considerations for any early retiree is their ability to earn an income. That might sound counterintuitive - are you really retired if you are still earning an income? If you are able to do something that you love and earn a little bit of income doing it, you are effectively introducing a large safe asset to your portfolio.
This changes all of the retirement math - maybe you only need a 2% withdrawal rate to supplement your earned income. Plus, work is important for humans!
Martin Seligman’s PERMA theory of well-being suggests that there are five building blocks that enable flourishing – Positive Emotion, Engagement, Relationships, Meaning, and Accomplishment.
Working at something that you love to do, even if it doesn’t make a ton of money, is a great way to get engagement, meaning, and accomplishment.
For early retirees, the 4% rule is not useful. Based on a longer time period, global data, and current market valuations a more reasonable guideline might be closer to 2%,and even then, constant inflation adjusted spending is both risky and inefficient. Alternative spending strategies like Vanguard’s dynamic approach might be part of the solution, but any early retiree should also keep in mind the possibility of finding ways to turn themselves into a safe asset by continuing to earn a bit of income.