...
I think Firecalc is useful for deciding whether to retire. Remember it assumes that you will spend your principal. No one here is actively planning to do that. ...
Not really accurate. FIRECalc assumes you will spend your principal
only in some cases. Roughly half the historical cases end up with a higher portfolio value than they started with (even after adjusting for inflation).
You need a spreadsheet that has 4 lines:
1. Portfolio value, including capital gains, dividends, ROE
2. Income: Pensions, SS, net rental, annuities
3. Expenses
4. Extra-ordinary costs: New cars, new roof...
Grow line 1 by your average return assumption. Line 2 may grow through COLA or not. Line 3 should have an inflater. Line 4 are specific estimates of non-recurring costs.
That's good, but it doesn't take into account "sequence of returns", which can be hard on portfolio survival.
The key assumption is life expectancy. Each year you live your expectancy increases by more than one year. My 90 yo Dad had an expectancy of 3.5 years and he died 5 years later so he was above the 50% median.
And that is why Firecalc loses its value.
I don't agree. If you are to rerun FIRECalc into retirement, you adjust for your current life expectancy. Since I'm conservative with this, I
plan for (not the same as
expecting) a very long joint LE (likely DW outliving me), like living to 100. Again, not they expect this, but on the outside chance it happens, I'd like to be prepared (same reason I wear a seat belt - to be prepared in case a crash happens, not that I expect a crash each time I drive). Unless medicine really advances, I doubt I'll be extending that age 100 planning, so my number of years entry in FIRECalc will only go down. And conservative planning for 100 pretty much gets you to a 'forever' portfolio anyhow, so it probably works for 105 or 110 as well.
If you only plan for your
median LE, you can expect to be wrong ~ 50% of the time (personal results won't be the same as a group - otherwise it would be exactly 50%). That's not very good planning in my book.
-ERD50