How have you "proven" that it's notohai wrote: ↑Mon Jan 13, 2020 3:00 pmFrom the 4% study, we can/must reasonably conclude that some 31y portfolios will fail under 4% withdrawal. Otherwise, we'd be assuming 0% volatility between years 30 and 31. Extend that - again this is just math - and the conclusion is declining withdrawal rates by length of retirement. This is both intuitive (others above suggest the same thing without referring to this math) and mathematically necessary if we accept the 4% withdrawal rate conclusion for 30y.EddyB wrote: ↑Mon Jan 13, 2020 1:30 pmI'm not sure what you mean by "much lower," but you haven'r "rationally" proven your point. It's sort of like someone saying that the perpetual withdrawal rate has been X%, and 50 years is short than perpetual, so the SWR must be higher than X%, forty years must be higher still, and 30 years even higher, so the SWR for thirty years must "rationally" be much higher than the perpetual withdrawal rate. Admittedly, "much lower" and "much higher" are loose terms, but there's nothing "rational" about either approach. Sadly, the past SWRs for 30, 40 or 50 year retirements haven't been radically different.ohai wrote: ↑Mon Jan 13, 2020 12:03 pm30x expenses for early retirement does not sound unreasonably conservative at all.

25x expenses (4% withdrawal rate) is based on a study of 30y retirement. This is the level where no 30y retirement portfolios have failed in some 100 year history of US assets.

Now consider this chain of thought:

Since the 4% withdrawal rate is where some 30y portfolios begin to fail, by definition, the marginal 30y portfolio (the first 30y portfolio to fail) reached 0% in year 30. Therefore, the marginal 31y portfolio must have a withdrawal rate of less than 4%

Now, extend this to more and more years. The marginal 31y portfolio must have had a withdrawal rate of less than 4% (from above). The marginal 32y portfolio must have had a withdrawal rate that is lower still. Even further on... the marginal 50y portfolio must have a withdrawal rate substantially lower than 4%. 3.33% even sounds aggressive!

One could argue that the historical test of "no failures in history" is too conservative. However, if you subscribe to the "4% rule" for normal retirement, rationally, you must believe in a much lower withdrawal rate for early retirement.

*just barely*lower? And, if there's any perpetual rate, then the decrease in the rate must eventually stop as you extend the period.

*That's*just math.