Originally Posted by Scuba
To rayvt, I hadn't thought of it that way. If our investment portfolio can't be counted on to generate 7-8% each year, isn't it better to take from that vs start the pension? What am I missing?
First off, you can never "count" on investments generating any particular gain each year. It may be 0%, it may be +25%, it may be -15%.
Fun fact: from 1950 to 2016, the one year (12 month) returns of a 60/40 (S&p500/10 yr Tbill) portfolio has ranged from low of -27% to high of +40%, with a median of +10.4%.
A 100/0 portfolio has ranged from low of -43% to high of +61%, with median of 13.4%
But the key point, which you are missing, is that your portfolio and your pension are completely independent of one another. Whether the portfolio goes up or down, the math for delaying the pension is the same -- it costs you 100% now to get 7%-8% every year in the future. The math is: 100/7 (or 100/8) to compute how long it takes you to get back to even.
FWIW, I went through all this when I decided when to start taking my pension. Every month that I delayed the start bumped my pension check by around $14, and I wanted to wait until it crossed the next $100 boundary. Then I said, "Wait--I have to forgo $2000 next month in order to get an extra $14 thereafter, so I'm really buying $14/mo for $2000. So for the next 142 months, I'm just getting my own money back."