I don't think this study is as revolutionary as you think. 1970 isn't that long ago, in 30-year chunks. More importantly, it misses the year that defined SAFEMAX for Bengen, 1966. I don't think the data is far off, if you shrunk Bengen's study data start o 1970. (See below)
The 4% rule isn't a smooth thing. Your actual safe withdrawal rate could be 8%, if you were lucky in your retirement. It's just that you don't know until the end if you were right. So, the conservative answer is the worst case.
Exactly. This is the problem with Portfolio Charts with its 50 years of data. (So not even a full life-time for an individual.) And also a problem with FireCalc and cFiresim too. People look at 150 years of history. So generalize based on basically a sample size of 2 or 3.
In a sense, the generalization might be "US stocks work great as long as US is on winning side in WWII and doesn't lose Cold War."
It's a perennial observation on this board that click-to-buy stock investing is a today's-generation phenomenon, and retirement itself is only a feature of the last 70-80 years. Middle-class retirement in the U.S. only became a thing after the beginning of Social Security in 1940. Stocks and mutual funds only became available to middle class people in the 1970s, with massive front end loads, commissions, and expense ratios that none of us today would accept, enduring until the 1990s*. Women were generally forced into lifelong financial dependency until the mid 1970s, when they were finally allowed to open their own accounts and work a wider range of jobs.
Yes, there were private pensions in mid-20th century America, but they generally required a "full retirement age" and are fundamentally different than what we talk about when we buy ETFs through our phone apps. Plus, private sector or union pensions are all but gone now, so most of the examples we are aware of (e.g. our grandparents or parents' retirements) are based on a different model than we can follow. And now, the future of Social Security is in doubt.
Another observation: Retirement/FIRE was a fantasy in most countries during most of the 20th century. Argentina, for instance, started the 20th century with a GDP that was 70% of the U.S's, but due to poor political decisions and instability they ended the century in poverty. Similarly, Brazil and Mexico spent the century wallowing in corruption and failing to educate their children or invest in R&D, falling behind on development and growth. Western Europe in the 1900s contained the only countries that would today meet our definition of "developed" but devastating wars and/or communism would end their prosperity for generations - though at least they have underfunded and shaky old-age pensions.
We're left with a picture of early retirement that has only ever been available to people caught within a very specific confluence of culture, politics, and luck. The globalist assumptions underpinning the 4% rule - that world markets were converging on the US-20th-century model of regulated capitalism, honest government, and free trade - seems in doubt. Thus, I'm casting a wary eye on the 50-year projections based on past performance, as U.S. culture shifts toward chaos, as the politics shift toward authoritarian-Idiocracy, and as we press our luck with wars, deficits, and the national debt. Do we really have a brighter future than Argentina circa 1920 or Germany circa 1900 or Mexico circa 1950? It seems like a bold claim, given the history of how fragile the whole concept is.
*I've never heard of a FIRE calculator that factored in the
roughly 1% one-way transaction costs that a person in the mid-20th century would have had to pay if they were periodically investing a fraction of their paycheck into stocks, or living off of their assets in retirement. The commissions themselves lowered stock valuations, which is why PE ratios looked so much better in the past, and which is why stock returns look so much better in the past than what investors actually experienced.
