3 ms·
On a longer timeframe you'll need to have a lower withdrawal rate, but not a huge amount more. The 4% withdrawal rate is based on the Trinity Study [1] which w
by froindt 9y ago
On a longer timeframe you'll need to have a lower withdrawal rate, but not a huge amount more. The 4% withdrawal rate is based on the Trinity Study [1] which was looking at 30 year time periods. (edit: added the link)
Per cFIREsim, with "success" defined as having more than 0 dollars at the end of the time period
375k -> 84.62% success
425k -> 96.15% success
450k -> 99.04% success
Additionally, the median portfolio value at the end went from 650k to 1.1M to 1.4M.
375k is definitely cutting things close. But at the same time...
* you're always going to have unpredictable market conditions
* you're likely to get at least some level of social security (which is unaccounted for in the above numbers)
* pulling back the expenses even a couple percent in the particularly bad years has a huge effect on success. This is admittedly difficult at 15k/year expenses, but padding the 375k by 5-10% would do a lot to mitigate that risk as well.
* If things start going downhill in terms of market returns, you retired young enough that you can reasonably go back to work for a short period of time. Even working at McDonald's at 7.25/hour would be approximately sufficient to cover the annual expenses.
[1] https://en.wikipedia.org/wiki/Trinity_study https://en.wikipedia.org/wiki/Trinity_study
- auxym 9y agoTo add on to this, I found the recent reddit AMA with Bill Bengen (first author on that Trinity paper) highly interesting: https://www.reddit.com/r/financialindependence/comments/6vazih/im_bill_bengen_and_i_first_proposed_the_4_safe/ https://www.reddit.com/r/financialindependence/comments/6vaz...