jdw_fire said:
If you add 5 years to the life expectancy of a 50 yo, as some have suggested, then the W/D rate drops to 4.44% for a time frame of 39 years.
Detailed analysis like this makes me think of the saying "Man plans and God smiles."
We all know old people who won't buy green bananas because they're not sure they (the people, not the fruit) have enough time left. They're estimating their lifespan as too short to risk a bet on a long-term investment. If they're right, then they've saved a little money for their heirs. If they're wrong, there's no penalty-- they just go buy more bananas.
We also know people whose genetics convince them not to expect to live to triple digits. If all four of their grandparents died before their 70s then they probably wouldn't worry about their own RMDs or a Roth IRA conversion. If they're right, their heirs get to deal with the RMDs. If they're wrong, then there's no penalty and every RMD calculation is a gift. No risk, no penalties.
We also know people who'll grab Social Security the minute they're 62 because they don't think they'll live long enough to see the payback from waiting for full eligibility. However Bud Hebeler has pointed out that it permanently reduces the spouse's benefits, too, so this strategy carries some pain. If the SS recipient is right then they got a little money in the hand at the cost of permanently reducing spouse's benefits. (That may or may not be a concern.) If they're wrong then they have to live the rest of their life with a permanent reduction in benefits. This strategy has plenty of penalties, some quite severe, so can the recipient really afford to be "wrong"?
We all know that current demographics say there's little chance of us late Boomers living to triple digits. However today's centurion class is the fastest-growing part of American society, and actuarial data doesn't predict your individual longevity any more than CPI approximates your personal inflation rate. When we run FIRECalc and confidently set our lifespan at the expected lifespan for our birth year, we should realize that 50% of us will be wrong. When we raise our life expectancy by 10 or even 25 years, that still means we could be wrong 5-10% of the time. That's a big issue-- if the penalty for being wrong is the risk of running out of money, do we want to risk being wrong even 1% of the time?!?
Same for choosing an extremely conservative "sleep at night" retirement portfolio at the risk of losing to inflation. If you unexpectedly exceed your life expectancy and get savaged by inflation, can you really afford to be wrong at all? If there's a 25% chance that at least you
or your spouse, let alone you
and your spouse could live into your 90s, is it really responsible to set your portfolio survivability at 30 years?
So when you're tweaking your FIRECalc numbers, think about the consequences of being wrong and having to live the rest of your frail elderly years on just an SS check. Think about long-term care in a Medicaid facility instead of being able to afford long-term care insurance or being able to pay your own way for home care. The consequences of being wrong are far worse than any potential regrets of sitting on a huge pile of your heirs' money feeling maudlin about missed spending opportunities.
No matter how short you expect your life to be or how low a success rate you're willing to risk, if you're pushing the SWR boundaries then you should also be able to vary your withdrawal rate in case you fall into a nasty bear market for a year or two. I don't care whether you start at 4% and raise it for inflation, or use ESRBob's 4%/95% rule, or try for Bernicke's brass balls ring-- if your portfolio drops by 25% when oil hits $250/barrel and you can't work a Wal-Mart shift then you will really wish that you could reduce your spending that year.
I think any SWR over 4% is a tradeoff whose (admittedly minor) risks outweigh any rewards. To survive any SWR over 4% it'd be far better to raise a portfolio's equity allocation than to attempt to sleep soundly at night on a bond mattress. I think any SWR over 5% is downright foolhardy without an all-equity portfolio and two years' expenses in cash, and even then the risks outweigh the putative rewards if the portfolio's success rate isn't at least 80%.
If your ER success nosedives when the length of your ER goes from 30 to 50 years, then I wouldn't risk ER. If the success rate of your ER drops below 80% before your expected lifespan exceeds 100 years of age, then I'd hesitate to ER. Anything above 80% is an unquantifiable crapshoot between your lifespan and Bernstein's "Retirement Calculator from Hell" articles.
Personally our ER numbers are based on an expected lifespan of 120 years, a minimum success rate of 80%, and a 10% margin of potential belt-tightening. We think inflation also merits the aforementioned equity/cash allocation. We don't necessarily expect to be invited to our kid's 90th birthday party, but we want to be able to afford to pay for the trip...