By Mark Hulbert
Retirees and near-retirees have known for some time that the famed 4% rule needs to be revised downwards. But by how much?
Would you believe 1.9%? That’s the conclusion of new research that replicates the original research that led to the 4%, but with more realistic assumptions about our life expectancies and a more comprehensive historical data set.
The implications are huge and potentially devastating. Under the 4% rule, a $1 million 401(k) would allow you to spend an inflation-adjusted $40,000 each year in retirement with minimal odds of outliving your money. With the new rule, you would be able to spend an inflation-adjusted amount of just $19,000 per year.
And that’s assuming you have a $1 million retirement portfolio. According to the most recent analysis by Vanguard , only 15% of retirement accounts at Vanguard are worth even $250,000. And according to an analysis of Federal Reserve data by the Boston College Center for Retirement Research , only 12% of workers have any retirement account in the first place.
The infamous 4% rule traces to a study in 1994 by William Bengen, a financial planner, which appeared in the Journal of Financial Planning . He based the rule on calculations showing that a portfolio of 50% stocks/50% bonds would have survived every 30-year period in the U.S. between 1926 and 1991.
The just-completed study that comes up with a much lower spending rule was conducted by Richard Sias and Scott Cederburg, finance professors at the University of Arizona; Michael O’Doherty, a finance professor at the University of Missouri, and Aizhan Anarkulova, a Ph.D. candidate at the University of Arizona. The study is entitled “ The Safe Withdrawal Rate: Evidence from a Broad Sample of Developed Markets .”
Though the researchers made some tweaks to Bengen’s methodology, the much-lower spending rule they arrived at traces to two primary causes. The first is accounting for our expected longevity.
In an interview, Sias pointed out that based on actuarial data from the Social Security Administration (SSA), a 65-year old couple retiring today has nearly a one out of four chance of at least one spouse surviving more than 30 years. To account for this possibility, the researchers replicated Bengen’s approach with SSA data.
The second cause of the much lower spending rate had an even bigger impact: The researchers relied on a much more comprehensive database of historical stock and bond returns and inflation. Specifically, their database reflected returns from 38 developed countries between 1890 and 2019, containing in total nearly 2,500 years of stock, bond, and inflation data.
This had such a big impact on the researchers’ findings because the U.S. has outperformed almost all other developed countries over the last century. Unless you believe in more or less perpetual U.S. markets’ exceptionalism, the market returns of other countries are also relevant when forecasting the future course of our retirements.
Should you believe in this exceptionalism?
Sias believes not, noting that the comprehensive database on which he and his fellow researchers relied included only developed countries. So their results aren’t caused by the often-inferior returns of emerging and frontier country markets. Indeed, he pointed out, some countries had higher GDP per capita than the U.S. in the first years in which they were added to the researchers’ database.
To appreciate what this means, imagine that, without lookahead bias, a century ago you were betting on which countries’ markets would produce the greatest future returns. If you based your bet solely on country size and income, you would have sooner bet on other countries than the U.S. as most likely to produce the best future stock and bond market returns. Sias reminds us that it’s dangerous to base our retirements on lookahead bias.
Sias referred specifically to Japan as an instructive example. By the end of the 1980s, its stock market had the highest market cap of any in the world, significantly greater than that of the U.S. It subsequently crashed, of course, and never fully recovered. It today is a third lower than where it stood at its 1989 peak.
How can we be so sure that this Japanese experience is irrelevant to U.S. workers retiring today? In 1989, many were predicting that Japan would not only continue to dominate the global economy but become even more predominant. How is that different than a similar prediction today that the U.S. stock and bond markets over the next century will outperform the rest of the world by as much as it did over the last century?
We can hope that the U.S. will avoid a post-1989 Japanese-like fate. But hope is not a retirement strategy. And once you substitute hard data for that hope, according to this new research, you will adopt a spending rule far lower than 4%.
Note on size of spending rule
I should point out that the authors of this new study didn’t come up with just one recommended spending rule, notwithstanding my reporting at the beginning of this column that they said it should be 1.9%.
The researchers instead found that the spending rule you choose is dependent on how much risk you’re willing to incur of outliving your money. The 1.9% I report in this column is what the authors calculate what the spending rule would be if you wanted the same probability of “financial ruin” (outliving your money) as the 4% rule had with U.S.-only data.
Mark Hulbert is a regular contributor to MarketWatch. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at firstname.lastname@example.org .